CN113619447B - Method for predicting state of charge of battery of electric automobile - Google Patents
Method for predicting state of charge of battery of electric automobile Download PDFInfo
- Publication number
- CN113619447B CN113619447B CN202110807480.2A CN202110807480A CN113619447B CN 113619447 B CN113619447 B CN 113619447B CN 202110807480 A CN202110807480 A CN 202110807480A CN 113619447 B CN113619447 B CN 113619447B
- Authority
- CN
- China
- Prior art keywords
- battery
- electric vehicle
- neural network
- current state
- vehicle battery
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L58/00—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles
- B60L58/10—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries
- B60L58/12—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries responding to state of charge [SoC]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/60—Other road transportation technologies with climate change mitigation effect
- Y02T10/70—Energy storage systems for electromobility, e.g. batteries
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Power Engineering (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Secondary Cells (AREA)
Abstract
The application discloses a prediction method of the state of charge of a battery of an electric automobile, which comprises the following steps: acquiring current state data of a battery of the electric automobile; according to the obtained current state data, the predicted discharge potential capacity and the predicted charge potential capacity of the battery of the electric automobile, predicting the current performance capacity of the battery by adopting an artificial neural network; according to the current state of the battery and the predicted expressive power, predicting the current potential power of the battery by adopting an artificial neural network; and predicting the current state of charge of the battery of the electric automobile according to the current performance capability and potential capability of the battery. According to the method, the performance capability and the potential capability of the electric vehicle battery are predicted by adopting the artificial neural network, the state of charge of the electric vehicle battery is predicted according to the performance capability and the potential capability, the prediction accuracy of the state of charge of the electric vehicle battery is greatly improved, and an accurate state of charge prediction result can be obtained by the electric vehicle battery in the whole life cycle.
Description
Technical Field
The application relates to the technical field of electric automobiles, in particular to a method for predicting the state of charge of a battery of an electric automobile.
Background
Accurate prediction of the battery SOC (State of Charge) of an electric vehicle has complexity. The endurance mileage of the electric automobile is influenced by various factors such as the state of the battery, the level of the driver, the traffic conditions and the like, so that the accurate prediction of the state of charge of the battery plays a very important role in effectively guiding the driver to improve the endurance mileage under the condition of the current technical level. Improving the accuracy of prediction of the state of charge of an electric vehicle battery is a currently pending problem.
Disclosure of Invention
The purpose of the application is to provide a prediction method for the state of charge of a battery of an electric automobile. The following presents a simplified summary in order to provide a basic understanding of some aspects of the disclosed embodiments. This summary is not an extensive overview and is intended to neither identify key/critical elements nor delineate the scope of such embodiments. Its sole purpose is to present some concepts in a simplified form as a prelude to the more detailed description that is presented later.
According to an aspect of the embodiments of the present application, there is provided a method for predicting a state of charge of a battery of an electric vehicle, including:
acquiring current state data of the electric automobile battery for the nth time; wherein n is more than or equal to 1;
When n=1, according to the current state data acquired for the first time, predicting the current performance of the electric vehicle battery for the first time by adopting a first artificial neural network which is trained in advance;
according to the current state data acquired for the first time, judging the current state of the electric automobile battery for the first time;
according to the current state obtained by the first judgment, predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance;
when n is more than or equal to 2, predicting the current expressive power of the electric vehicle battery by adopting a first artificial neural network which is trained in advance according to the current state data acquired in the nth time and the discharging potential power and the charging potential power of the electric vehicle battery which are obtained in the n-1 th prediction;
judging the current state of the electric automobile battery according to the current state data acquired by the nth time;
predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance according to the current state of the electric vehicle battery and the expression capacity obtained by the n-1 th prediction;
predicting the current state of charge of the electric vehicle battery according to the current expressive power and potential power of the electric vehicle battery;
And updating the value of n when the preset stopping condition is not reached, and turning to the nth to acquire the current state data of the electric automobile battery until the preset stopping condition is reached.
In some embodiments of the present application, the potential capability includes a discharge potential capability and a charge potential capability.
In some embodiments of the present application, the predicting, according to the current state obtained by the first determination, the current potential capability of the electric vehicle battery by using an artificial neural network that is trained in advance includes:
if the current state obtained by the first judgment is in a discharging state, predicting the discharging potential capacity of the electric automobile battery through the second artificial neural network which is trained in advance;
and if the current state obtained by the first judgment is the current charging state, predicting the potential charging capability of the electric vehicle battery through the third artificial neural network which is trained in advance.
In some embodiments of the present application, the predicting the current potential capability of the electric vehicle battery according to the current state of the electric vehicle battery and the expression capability predicted for the n-1 th time by using an artificial neural network that is trained in advance includes:
If the current state of the electric vehicle battery is in a discharging state, predicting the potential discharging capacity of the electric vehicle battery through the second artificial neural network which is trained in advance according to the expression capacity obtained by the n-1 th prediction;
and if the current state of the electric vehicle battery is the current charging state, predicting the potential charging capacity of the electric vehicle battery through the third artificial neural network which is trained in advance according to the expression capacity obtained by the n-1 th prediction.
In some embodiments of the present application, the obtaining current state data of the battery of the electric vehicle includes:
collecting current state parameter data of the electric automobile battery; the state parameter data comprise voltage, current, consumed electric quantity and temperature;
and carrying out normalization processing on the state parameter data to obtain the current state data.
In some embodiments of the present application, the input to the pre-trained first artificial neural network includes a rate of change of the voltage and a rate of change of the temperature.
In some embodiments of the present application, before the acquiring the current state data of the electric vehicle battery, the method further includes:
And training the first artificial neural network by adopting the battery data of the electric automobile under the condition of different discharging multiplying powers to obtain the pre-trained first artificial neural network.
In some embodiments of the present application, before the acquiring the current state data of the electric vehicle battery, the method further includes:
and training the second artificial neural network and the third artificial neural network by adopting the battery test data of the electric automobile under different cycle lives to obtain the pre-trained second artificial neural network and the pre-trained third artificial neural network.
In some embodiments of the present application, the electric vehicle battery is replaced with an electric vehicle battery equivalent circuit model, so as to predict the state of charge obtained by the electric vehicle battery equivalent circuit model as the state of charge of the electric vehicle battery.
According to another aspect of the embodiment of the application, an electric automobile battery equivalent circuit model is provided, and the method is used for realizing the method; the electric automobile battery equivalent circuit model comprises a power supply, a first resistor, a second resistor, a third resistor, a fourth resistor, a first capacitor, a second capacitor, a third capacitor, a first diode, a second diode, a positive output end and a negative output end; the first capacitor, the second capacitor and the third capacitor are all polar capacitors; the positive electrode of the power supply is connected with the negative electrode of the third capacitor, the positive electrode of the third capacitor is respectively connected with the first end of the first resistor and the first end of the second resistor, the second end of the first resistor is connected with the positive electrode of the first diode, the second end of the second resistor is connected with the negative electrode of the second diode, and the negative electrode of the first diode and the positive electrode of the second diode are respectively connected with the positive output end; the negative electrode of the power supply, the third resistor, the fourth resistor and the negative output end are sequentially connected; the first capacitor is connected with the third resistor in parallel, and the positive electrode of the first capacitor is connected with the negative electrode of the power supply; the second capacitor is connected with the fourth resistor in parallel, and the negative electrode of the second capacitor is connected with the negative output end.
One of the technical solutions provided in one aspect of the embodiments of the present application may include the following beneficial effects:
according to the method for predicting the state of charge of the electric automobile battery, the performance capability and the potential capability of the electric automobile battery are predicted by adopting the artificial neural network, and the state of charge of the electric automobile battery is predicted according to the performance capability and the potential capability, so that the accuracy of predicting the state of charge of the electric automobile battery is greatly improved, and an accurate state of charge prediction result can be obtained for the electric automobile battery in the whole life cycle.
Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application. The objectives and other advantages of the application will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 shows a flow chart of a method for predicting the state of charge of an electric vehicle battery according to one embodiment of the present application;
FIG. 2 shows an electric vehicle battery equivalent circuit model circuit diagram of one embodiment of the present application;
FIG. 3 illustrates a schematic diagram of a validated simulation model of a battery equivalent model in one embodiment of the present application;
FIG. 4 shows simulation results of a battery pulse discharge process in one embodiment of the present application;
FIG. 5 illustrates a schematic diagram of the manner in which parallel control in one embodiment of the present application works;
FIG. 6 shows a flow chart of predicting battery remaining capacity in one embodiment of the present application;
FIG. 7 shows a schematic diagram of a battery SOC prediction algorithm in one embodiment of the application;
FIG. 8 illustrates SOC predictions under battery constant current discharge test data in one embodiment of the application;
FIG. 9 illustrates a national standard modal condition speed variation curve in one embodiment of the present application;
FIG. 10 shows AVL Cruise calculation results under national standard modal conditions in one embodiment of the present application;
FIG. 11 illustrates SOC predictions under national standard mode operating condition standard battery data in one embodiment of the application;
FIG. 12 illustrates FTP75 operating mode speed change curves in one embodiment of the present application;
FIG. 13 shows AVL Cruise calculation under a single FTP75 operating condition in one embodiment of the present application;
FIG. 14 illustrates SOC predictions under single FTP75 operating mode standard battery data in one embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
It will be understood by those skilled in the art that all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs unless defined otherwise. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
Since the capacity that a battery can discharge is affected by many factors such as the discharge rate, the battery temperature, the number of charge-discharge cycles, etc., the SOC calculation of the battery should sufficiently take into account the influence of these factors. For battery SOC prediction of an electric vehicle, the following two main influencing factors are:
(1) Influence of the Battery itself
The life of the battery actually determines the actual electric quantity of the battery, the parameter is influenced by factors such as the cycle number, the working environment, the depth of discharge and the like of the battery, and the like state variables determine the total electric quantity which can be discharged by the battery, so that the residual electric quantity of the battery can be accurately obtained by determining the total electric quantity of the battery. In addition, for the same type of battery, there is a problem of consistency between the battery packs and between the battery packs, that is, the equivalent parameters for one battery are not necessarily applicable to another battery pack, so that the parameters of the state of each specific battery need to be corrected at the time of prediction.
(2) Influence of the manner of use of the Battery
The change of the discharging capacity caused by the using mode of the battery is not only determined by different discharging multiplying powers of the battery, but also the battery not only comprises a discharging process in the using process of the electric automobile due to the existence of a regenerative braking technology. The chemical reactions of charge and discharge of the battery are the inverse process, and the products generated by the side reactions in the process are different, so that the battery is determined to have different characteristics in the circuit under different states of charge and discharge, and the conclusion is proved by calculation according to the test data of the actual battery. Therefore, the battery on the electric automobile needs to be considered in the charge and discharge states respectively, so that the actual situation of the battery can be accurately simulated.
In practical processes, the sources of errors in the predicted result of the SOC can be generally classified into errors caused by aging, errors caused by a usage mode, and errors caused by measurement. The measurement error is caused by the measurement precision of the equipment such as a sensor and the like, and belongs to unavoidable errors; the errors of the battery and the errors of the use mode are the key to consider in the prediction process for correction. According to the analysis, the embodiment of the application designs an algorithm for tracking and predicting the SOC based on the parallel model, corrects the main cause of the error and improves the prediction precision.
The most currently used battery type of the electric automobile is a lithium ion battery, and a model of the lithium ion battery can have a plurality of different modeling modes aiming at different research contents. Regarding the content of battery SOC prediction, the equivalent circuit model of the battery can not only be used for equivalent battery working principle, but also reflect the state condition of the battery by the change of circuit parameters, thereby being beneficial to the elimination of errors in the prediction calculation process.
Referring to fig. 1, an embodiment of the present application provides a method 1 for predicting a state of charge of a battery of an electric vehicle, including the following steps:
Acquiring current state data of the battery of the electric automobile for the nth time; wherein n is more than or equal to 1;
when n=1, according to the current state data obtained for the first time, predicting the current performance of the electric vehicle battery for the first time by adopting a first artificial neural network which is trained in advance;
according to the current state data acquired for the first time, judging the current state of the battery of the electric automobile for the first time;
according to the current state obtained by the first judgment, predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance;
when n is more than or equal to 2, predicting the current performance capacity of the electric vehicle battery by adopting a first artificial neural network which is trained in advance according to the current state data acquired in the nth time and the discharging potential capacity and the charging potential capacity of the electric vehicle battery obtained in the n-1 th prediction;
judging the current state of the battery of the electric automobile according to the current state data acquired in the nth time;
predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance according to the current state of the electric vehicle battery and the expression capacity obtained by the n-1 th prediction;
predicting the current state of charge of the electric vehicle battery according to the current performance capability and potential capability of the electric vehicle battery;
And updating the value of n when the preset stopping condition is not reached, and turning to the nth to acquire the current state data of the electric automobile battery until the preset stopping condition is reached. The preset stop condition may be, for example, that the predicted current state of charge of the battery of the electric vehicle reaches a certain preset threshold value.
In certain embodiments, the potential capability includes a discharge potential capability and a charge potential capability.
According to the current state obtained by the first judgment, predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance, wherein the method comprises the following steps:
if the current state obtained by the first judgment is the current discharging state, predicting the discharging potential capacity of the electric vehicle battery through a second artificial neural network which is trained in advance;
and if the current state obtained by the first judgment is the current charging state, predicting the potential charging capability of the electric vehicle battery through a third artificial neural network which is trained in advance.
According to the current state of the electric vehicle battery and the expression capacity obtained by the n-1 th prediction, predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance, wherein the method comprises the following steps:
If the current state of the electric vehicle battery is in a discharging state, predicting the discharging potential capacity of the electric vehicle battery through a second artificial neural network which is trained in advance according to the expression capacity obtained by the n-1 th prediction;
if the current state of the electric vehicle battery is in the charging state, predicting the potential charging capacity of the electric vehicle battery through a third artificial neural network which is trained in advance according to the expression capacity obtained through the n-1 th prediction.
Acquiring current state data of an electric vehicle battery, comprising:
collecting current state parameter data of an electric automobile battery; the state parameter data comprises voltage, current, consumed electric quantity and temperature;
and carrying out normalization processing on the state parameter data to obtain current state data.
The inputs to the pre-trained first artificial neural network include the rate of change of voltage and the rate of change of temperature.
Before acquiring the current state data of the electric automobile battery, the method further comprises the following steps:
and training the first artificial neural network by adopting the battery data of the electric automobile under the condition of different discharging multiplying powers to obtain a pre-trained first artificial neural network.
Before acquiring the current state data of the electric automobile battery, the method further comprises the following steps:
And training the second artificial neural network and the third artificial neural network by adopting the battery test data of the electric automobile under different cycle lives to obtain a pre-trained second artificial neural network and a pre-trained third artificial neural network.
The other embodiment of the application provides a method 2 for predicting the state of charge of an electric vehicle battery, which replaces the electric vehicle battery with an electric vehicle battery equivalent circuit model based on the method 1 to predict the state of charge obtained by the electric vehicle battery equivalent circuit model as the state of charge of the electric vehicle battery.
Another embodiment of the present application provides an electric vehicle battery equivalent circuit model, configured to implement a method 2 for predicting a state of charge of an electric vehicle battery; the electric automobile battery equivalent circuit model comprises a power supply, a first resistor, a second resistor, a third resistor, a fourth resistor, a first capacitor, a second capacitor, a third capacitor, a first diode, a second diode, a positive output end and a negative output end; the first capacitor, the second capacitor and the third capacitor are all polar capacitors; the positive electrode of the power supply is connected with the negative electrode of the third capacitor, the positive electrode of the third capacitor is respectively connected with the first end of the first resistor and the first end of the second resistor, the second end of the first resistor is connected with the positive electrode of the first diode, the second end of the second resistor is connected with the negative electrode of the second diode, and the negative electrode of the first diode and the positive electrode of the second diode are respectively connected with the positive output end; the negative electrode, the third resistor, the fourth resistor and the negative output end of the power supply are sequentially connected; the first capacitor is connected with the third resistor in parallel, and the positive electrode of the first capacitor is connected with the negative electrode of the power supply; the second capacitor is connected with the fourth resistor in parallel, and the negative electrode of the second capacitor is connected with the negative output end.
In another embodiment of the present application, the equivalent circuit modeling for battery SOC prediction uses a parallel model, so that the equivalent model needs to be able to approach the actual situation as much as possible, and the model needs to have good steady state characteristics and dynamic response. Meanwhile, under the condition that the electric automobile basically has the function of regenerative braking, the model of the battery is required to show two different processes of battery charging and discharging. As shown in fig. 2, an electric vehicle battery equivalent circuit model according to an embodiment of the present application includes a power supply U oc A first resistor R ind A second resistor R inc Third resistor R P1 Fourth resistor R P2 First capacitor C P1 A second capacitor C P2 Third capacitor C b First diode D 1 Second diode D 2 A positive output Out1 and a negative output Out2; wherein, the first capacitor C P1 A second capacitor C P2 Third capacitor C b All are provided with polar capacitors; power supply U oc Positive electrode of (C) and third capacitor C b A third capacitor C connected with the negative electrode of b The positive electrode of (a) is respectively connected with the first resistor R ind And a first end and a second resistance R of (2) inc A first resistor R connected to the first end of ind And the second end of the first diode D 1 The positive electrode of the second resistor R is connected with inc And a second diode D 2 Is connected with the negative pole of the first diode D 1 Is connected with the cathode of the second diode D 2 The positive poles of the two are respectively connected with a positive output end Out 1; power supply U oc Negative electrode of (d), third resistor R P1 Fourth resistor R P2 And a negative output end Out2 are sequentially connected; first capacitor C P1 And a third resistor R P1 Parallel connection, a first capacitor C P1 Positive electrode of (a) and power supply U oc Is connected with the negative electrode of the battery; second capacitor C P2 And a fourth resistor R P2 Parallel connection, a second capacitor C P2 Is connected to the negative output terminal Out 2.
As shown in fig. 2, since the self-discharge condition of the lithium ion battery is relatively small, the self-discharge during use is negligible, so that the resistance simulating the self-discharge condition is eliminated in the model. According to the parameter identification, different electrochemical reactions occur in the charge and discharge process of the battery, so that no matter the ohmic internal resistance, the energy storage capacitance, the polarized internal resistance or the polarized capacitance of the battery, certain errors exist in the charge and discharge parameters under the condition of the same SOC, and therefore, the charge and discharge states need to be distinguished when the battery parameters need to be considered.
After the equivalent model is determined, the circuit parameters inside the equivalent model need to be identified, and according to the rule of the test battery charge-discharge performance experiment in FreedomCAR Battery Test Manual For Power-Assist Hybrid Electric Vehicles, a test method of HPPC (Hybrid Pulse Power Characteristic, hybrid power pulse capability characteristic) is generally adopted for battery test. HPPC testing is typically accomplished using special battery test equipment such as Arbin BT2000, newware BTs4000, and the like. Because the performance of the battery is greatly affected by temperature, the battery needs to be tested in a constant-temperature environment (the temperature of the environment is usually 26 ℃), the test flow is divided into a capacity test part and a pulse capability test part, and parameters in an equivalent circuit can be calculated according to the result of test data.
Parameter identification is carried out according to HPPC test data of a certain type of lithium ion battery, the rated voltage of the battery is 3.7V, the rated capacity is 63Ah, the rated current is 63A, and the test environment temperature is set at 26 ℃. The following table shows the fitting results for the discharge pulse process.
TABLE 1 improved fitting parameters (discharge) of GNL models
According to the parameter identification result under the condition of charge and discharge failure, whether the test is correct or not can be verified in a circuit simulation mode. The simulation circuit built in the Matlab/Simulink environment is shown in FIG. 3.
Taking the fitting result of the discharge pulse process as an example, the simulation result is shown in fig. 4.
From fig. 4, it can be seen that the simulation curve and the actual test curve at the discharge pulse stage are substantially identical, and it can be demonstrated that the equivalent model and the identified parameters can reflect the actual physical condition of the battery. Based on the above model, the prediction method of the battery SOC of the electric vehicle can be further studied.
SOC prediction method based on parallel model
A parallel model is a model that is applied within a parallel control system to describe the actual process. The parallel control refers to a control method for completing tasks in a virtual-real interaction mode, and is a way for computing and controlling science on the basis of big data and data driving. The parallel control is realized based on the ACP method, and the ACP refers to an organic combination of three of artificial society (Artificial societies), calculation experiment (Computational experiments) and parallel execution (Parallel execution). Essentially, the core of the ACP is to build up the "virtual" and "soft" parts of the complex system, and make them "hardened" through quantitative and practical calculation and real-time, so as to really solve the actual complex problem. The core idea of parallel control is: aiming at a complex system, a parallel system of parallel interaction of an actual system and a manual system is constructed, and the aim is to enable the actual system to trend towards the manual system instead of the manual system to approach the actual system, so that the complex problem is simplified by means of the manual system, and control and management of the complex system are realized. The operation mode of the parallel control is schematically shown in fig. 5. Essentially, the content of the parallel control is an embodiment of knowledge automation. The control method realizes the mutual feedback and interactive operation of the entity world and the virtual world.
The method of the embodiment of the application aims to realize accurate prediction of the battery SOC, so that the actual change of the entity in the actual process is required to be tracked and accurately simulated, and the model is enabled to approach to the actual state as much as possible. Therefore, the embodiment of the application combines the equivalent model of the battery with the actual use process, the mode not only can embody the ageing effect through the equivalent model parameters of the battery, but also can simulate the change condition in the actual running process according to the different states of the battery charging and discharging, and the prediction mode is consistent with the relationship of considering the use mode from equipment in practice, so that the more accurate description of the actual process can be realized.
Based on the battery equivalent model adopted in the embodiment of the application, the embodiment of the application has performed fitting identification on parameters therein, and the parameters are used as the performance of the potential capacity of the battery in the model for predicting the residual capacity. "potential capacity" means a change in charge-discharge capacity affected by a change in internal processes of the battery itself with an increase in service life, the potential capacity being determined by the state of the battery itself. In terms of parameter selection, in order to ensure accurate description of battery performance, Q is adopted eff ，R in ，τ 1 ，τ 2 Four parameters are taken as parameters representing the potential capacity of the battery. Wherein Q is eff The total electric quantity of the battery in the current state is shown, and the total electric quantity can be corrected according to the attenuation curve of the battery and the total cyclic discharge amount in the actual use process; ohmic internal resistance R in Used for reflecting the influence of steady state change and aging in the use process of the battery, and can be divided into R during charging according to the fitting result inc And R at discharge ind Applied in different use cases; τ 1 ，τ 2 The dynamic response of the battery is shown, and can be considered in different states of charge and discharge. The parameters basically and completely represent the main performance of the battery model, so that the potential capability of the battery can be reasonably represented by adopting the parameters.
The electric quantity discharged by the battery in the actual running process of the electric automobile is closely related to the state of the battery, environmental conditions such as temperature, working conditions such as discharge rate, voltage change rate and other parameters. The use condition of the battery in the running process of the electric automobile directly determines the quantity of the actual discharged electric quantity of the battery, the actual discharged electric quantity is called as the 'expressive ability' of the battery, and Q is used de To represent. That is, Q de Is determined by the 'potential ability' and the actual operation parameters, and adopts Q de The actual maximum discharge amount of the battery can be ensured to be as accurate as possible, so that the accuracy of calculating the residual capacity after subtracting the use amount is far greater than that of the traditional rated capacity ampere-hour method.
For the prediction modes of potential capability and expressive capability, due to the nonlinearity of parameters caused by the characteristics of a battery, the traditional prediction method is difficult to model, and complicated calculation and adjustment of fitting accuracy can exist, so that the embodiment of the application adopts an artificial neural network to realize the prediction of the potential capability and expressive capability. The whole procedure for predicting the remaining capacity includes two or more neural network modules, wherein one neural network predicts the potential capacity of the battery (two neural networks in the case of charge and discharge are considered), the predicted result is provided to the other neural network, the neural network predicts the expressive capacity of the battery by taking the potential capacity information and other acquired information together as input, and the remaining capacity of the battery is calculated according to the expressive capacity result, and the specific prediction flow is shown in fig. 6.
According to the prediction flow shown in fig. 6, the core for determining the prediction accuracy is the calculation result of the neural network, and is also the most important expression of the parallel model reaction actual situation. Through reasonable training of the neural network, the prediction method not only can obtain an accurate residual capacity under the condition of ensuring calculation accuracy, but also can judge sources of errors generated in the actual process through analysis of potential capacity parameters and performance capacity parameters, so that corresponding measures are adopted for guiding or improving the electric automobile.
The neural network module adopted in the embodiment of the application is a structure with the smallest error selected after training according to a large number of data samples, and the training and setting processes of the neural network are not independently accepted. The choice of structure is not unique and is chosen according to the specific training situation.
In some embodiments, the battery SOC prediction program used is mainly structured as shown in fig. 7.
As shown in fig. 7, the data processing part includes collection of input data, normalization of data, and judgment of initial state. The initial state obtains the current capacity of the battery according to the open-circuit voltage measured at the moment before the battery works. The input data mainly comprises four parameters of voltage U, current I, consumed electric quantity (integration of current with time) Q and temperature T. The potential capacity part adopts two neural networks to respectively represent different charge and discharge conditions, judges according to the current direction, and increases the cycle number n of the battery to be used as a characterization parameter of battery aging; the neural network module input with expressive capacity increases the voltage change rate dU/dT and the temperature change rate dT/dT to reflect the change condition in the using process and output the Q of the result de For calculating the remaining capacity of the battery. The output result of the program may include the battery remaining power and the battery SOC value. The data fed back are mainly through Q de Rate of change of (v) versus Q in potential capacity eff And the prediction result is corrected, so that the prediction accuracy of the potential capability is improved.
According to the embodiment of the application, a program for predicting the residual capacity is compiled for a platform by Matlab/Simulink, and the accuracy of an algorithm is verified through battery test data and vehicle simulation data mentioned in the previous section. And writing basic content of data processing by using M files, and calculating the processed data by using a module built in the Simulink.
In the program designed by the embodiment of the application, the most important is to ensure the calculation accuracy of the neural network, and the accuracy of the neural network is determined according to the situation of the prediction sample and the training situation of the neural network. In the embodiment of the application, battery test data under different cycle lives are adopted for training potential capacity, and data under different discharge multiplying factors are adopted for training expressive capacity. The determination of the number of hidden layers and the number of nodes of the neural network directly influences the prediction effect of the neural network in the training process of the neural network. Taking neural network training in the embodiment of the application as an example, because of limited training data samples, the number of nodes of different hidden layers must be trained, the optimal solution is judged by comparing the errors of calculation results, and meanwhile, whether the parameters calculated by the neural network meet the corresponding change rules is also required to be judged by the prediction results in program use. In the above example of training the potential ability module for discharging, the neural network structure includes 4 input parameters and 4 output parameters, and after comparing the calculation errors of the single hidden layer and the double hidden layer respectively, the embodiment of the application finally selects a mode of adopting the double hidden layer; in order to determine the number of hidden layer nodes, the number of nodes with the smallest error is finally selected as the structure of the neural network used by the program through comparing the calculation errors of different node numbers. The discharging potential capacity part of the embodiment of the application adopts a structure of two layers of 20 nodes, and the number of the nodes of the charging potential capacity module is respectively 11 and 16; the expressive power module nodes are 10 and 20, respectively.
Verification of constant-current discharge test data of battery
According to the data of the lithium ion battery of a certain model, the data of constant-current discharge in the HPPC test is adopted for verification. And firstly, verifying the state of SOC prediction of the complete discharging process of the battery under the condition of no attenuation. And (3) performing verification by adopting a constant-current discharge curve with the SOC variation of 100% to 0%. Because the predicted data are excessively amplified and marked on the partial areas, the predicted data curve and the actual data curve of the program almost coincide, the prediction error slightly increases along with the increase of the discharge depth, the overall prediction accuracy is very high, and the prediction program can well realize the relatively accurate prediction of the battery SOC under the condition of higher SOH. The predicted SOC for the battery constant current discharge test data is shown in fig. 8.
Verification of electric vehicle running standard data
In the actual use process of the electric automobile, the use condition of the battery is not simple constant current discharge, but a dynamic charge-discharge combined process, irregular discharge data is required to be adopted for verifying the program, and the program is ensured to be suitable for the actual use of the electric automobile. And calculating battery data under the actual working condition by adopting AVL Cruise software. The Cruise software is advanced simulation software developed by Austrian AVL company, and can be used for designing and calculating fuel economy, power, transmission ratio, emission performance and braking performance of vehicles (including motorcycles, buses, trucks and the like). The software can generate the change curves of the parameters of the vehicle under different working conditions through setting the parameters of the actual vehicle. Because the accurate simulation of the discharge condition of the battery on the electric automobile is difficult to realize in the off-line test of the battery, the AVL Cruise software can test the battery consumption data generated under the actual working condition as standard data, thereby further verifying the rationality of the design program of the embodiment of the application. Firstly, the energy consumption rate and the driving range test method of the GB/T18386-2005 electric automobile are adopted for testing. Fig. 9 shows a speed change curve of a national standard mode working condition, fig. 10 shows an AVL Cruise calculation result under the national standard mode working condition, and fig. 11 shows an SOC prediction result under standard battery data of the national standard mode working condition.
Fig. 11 shows a comparison of the predicted result of the application program and the standard data in the initial state where the battery SOC is 100%. It can be seen that the SOC curve does not exhibit a monotonically decreasing law of change due to regenerative braking, and there is a brief rise in the SOC curve at some time due to the charging process of regenerative braking. Compared with standard data, the predicted result has higher precision, and can be used for predicting the SOC curve of the irregularly discharged battery, so that the program can realize the SOC prediction under the NEDC working condition.
The NEDC condition reflects the change situation of the vehicle running to a certain extent, but the structural form of the NEDC condition belongs to the mode condition of the composite running, and more possible conditions in the actual process are that the speed and the acceleration change frequently, so that program verification is required for transient conditions like the us FTP75 to verify that the model can predict an accurate result for any condition, and the conditions are shown in fig. 12. Fig. 13 shows AVL Cruise calculation results under a single FTP75 condition.
Based on the above data, verification was performed using the prediction program designed in the embodiment of the present application, and the obtained results are shown in fig. 14.
From the predicted results, it can be seen that the initial SOC of the battery is 50% and the frequency of change in the SOC curve is higher than that of NEDC conditions throughout the prediction process, which is caused by more frequent changes in the speed of FTP75 conditions. The graph can also see a section of gentle SOC descending process, the speed of the process is zero in the FTP75 working condition, namely the electric automobile is in a stop state, but the curve is in a descending trend due to the fact that the additional electric energy consumption on the automobile is arranged in the AVL software, and the section can simulate the traffic jam or intersection waiting condition in the actual process, namely the process that the electric energy is consumed through other equipment except a motor on the electric automobile, so that the electric automobile meets the actual condition.
By analyzing the two different data prediction results, the program designed by the embodiment of the application can realize more accurate prediction of the residual capacity of the battery, has very high prediction precision for the ideal constant current discharge condition of the battery, and the prediction error under the constant current discharge condition is calculated to be less than 1% as a whole, and the method can well realize the prediction of the SOC due to simpler discharge process; and the error maximum value in the prediction process can be obtained after calculation under the condition of irregular charge and discharge in the actual use of the battery, so that the accuracy requirement of the prediction is met.
According to the embodiment of the application, the prediction result of the battery SOC of the electric automobile is achieved by adopting a method of adopting a parallel model and adopting a technical means of a neural network. First, the error and the reason causing the prediction are analyzed, and the identification method of the proper equivalent battery model and parameters is determined. Based on the model, a method flow and an implementation step for predicting the battery SOC are designed, and the accuracy of the method is proved by adopting example data. The method of the embodiment of the application is mainly used for continuously updating and tracking the state and the service condition of the battery of the electric automobile, so that the battery can be ensured to obtain a relatively accurate prediction result in the whole service life period.
It should be noted that:
the term "module" is not intended to be limited to a particular physical form. Depending on the particular application, modules may be implemented as hardware, firmware, software, and/or combinations thereof. Furthermore, different modules may share common components or even be implemented by the same components. There may or may not be clear boundaries between different modules.
The algorithms and displays presented herein are not inherently related to any particular computer, virtual machine, or other apparatus. Various general purpose devices may also be used with the examples herein. The required structure for the construction of such devices is apparent from the description above. In addition, the present application is not directed to any particular programming language. It will be appreciated that a variety of programming languages may be used to implement the teachings of the present application as described herein, and the above description of specific languages is provided for disclosure of preferred embodiments of the present application.
It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited in order and may be performed in other orders, unless explicitly stated herein. Moreover, at least some of the steps in the flowcharts of the figures may include a plurality of sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, the order of their execution not necessarily being sequential, but may be performed in turn or alternately with other steps or at least a portion of the other steps or stages.
The foregoing examples merely represent embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the present application. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application shall be subject to the appended claims.
Claims (8)
1. The method for predicting the state of charge of the battery of the electric automobile is characterized by comprising the following steps of:
acquiring current state data of the electric automobile battery for the nth time; wherein n is more than or equal to 1;
when n=1, according to the current state data acquired for the first time, predicting the current performance of the electric vehicle battery for the first time by adopting a first artificial neural network which is trained in advance;
according to the current state data acquired for the first time, judging the current state of the electric automobile battery for the first time;
according to the current state obtained by the first judgment, predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance;
when n is more than or equal to 2, predicting the current expressive power of the electric vehicle battery by adopting a first artificial neural network which is trained in advance according to the current state data acquired in the nth time and the discharging potential power and the charging potential power of the electric vehicle battery which are obtained in the n-1 th prediction;
judging the current state of the electric automobile battery according to the current state data acquired by the nth time;
predicting the current potential capacity of the electric vehicle battery by adopting an artificial neural network which is trained in advance according to the current state of the electric vehicle battery and the expression capacity obtained by the n-1 th prediction;
Predicting the current state of charge of the electric vehicle battery according to the current expressive power and potential power of the electric vehicle battery;
updating the value of n when the preset stopping condition is not reached, and turning to the nth to acquire the current state data of the electric automobile battery until the preset stopping condition is reached;
the obtaining the current state data of the electric automobile battery includes:
collecting current state parameter data of the electric automobile battery; the state parameter data comprise voltage, current, consumed electric quantity and temperature;
normalizing the state parameter data to obtain the current state data;
before the current state data of the electric automobile battery is acquired, the method further comprises the following steps:
and training the first artificial neural network by adopting the battery data of the electric automobile under the condition of different discharging multiplying powers to obtain the pre-trained first artificial neural network.
2. The prediction method according to claim 1, wherein the potential capability includes a discharging potential capability and a charging potential capability.
3. The prediction method according to claim 2, wherein the predicting the current potential capability of the battery of the electric vehicle by using the artificial neural network trained in advance according to the current state obtained by the first determination includes:
If the current state obtained by the first judgment is in a discharging state, predicting the discharging potential capacity of the electric automobile battery through the second artificial neural network which is trained in advance;
and if the current state obtained by the first judgment is the current charging state, predicting the potential charging capability of the electric vehicle battery through the third artificial neural network which is trained in advance.
4. The prediction method according to claim 2, wherein predicting the current potential capability of the electric vehicle battery using an artificial neural network trained in advance according to the current state of the electric vehicle battery and the performance capability predicted for the n-1 th time comprises:
if the current state of the electric vehicle battery is in a discharging state, predicting the potential discharging capacity of the electric vehicle battery through the second artificial neural network which is trained in advance according to the expression capacity obtained by the n-1 th prediction;
and if the current state of the electric vehicle battery is the current charging state, predicting the potential charging capacity of the electric vehicle battery through the third artificial neural network which is trained in advance according to the expression capacity obtained by the n-1 th prediction.
5. The method of claim 1, wherein the input to the pre-trained first artificial neural network comprises a rate of change of the voltage and a rate of change of the temperature.
6. The prediction method according to claim 1, wherein before the acquiring the current state data of the electric vehicle battery, the method further comprises:
and training the second artificial neural network and the third artificial neural network by adopting the battery test data of the electric automobile under different cycle lives to obtain the pre-trained second artificial neural network and the pre-trained third artificial neural network.
7. The prediction method according to any one of claims 1 to 6, wherein the electric vehicle battery is replaced with an electric vehicle battery equivalent circuit model to predict a state of charge obtained by the electric vehicle battery equivalent circuit model as the state of charge of the electric vehicle battery.
8. An electric vehicle battery equivalent circuit model for implementing the method of claim 7; the electric automobile battery equivalent circuit model comprises a power supply, a first resistor, a second resistor, a third resistor, a fourth resistor, a first capacitor, a second capacitor, a third capacitor, a first diode, a second diode, a positive output end and a negative output end; the first capacitor, the second capacitor and the third capacitor are all polar capacitors; the positive electrode of the power supply is connected with the negative electrode of the third capacitor, the positive electrode of the third capacitor is respectively connected with the first end of the first resistor and the first end of the second resistor, the second end of the first resistor is connected with the positive electrode of the first diode, the second end of the second resistor is connected with the negative electrode of the second diode, and the negative electrode of the first diode and the positive electrode of the second diode are respectively connected with the positive output end; the negative electrode of the power supply, the third resistor, the fourth resistor and the negative output end are sequentially connected; the first capacitor is connected with the third resistor in parallel, and the positive electrode of the first capacitor is connected with the negative electrode of the power supply; the second capacitor is connected with the fourth resistor in parallel, and the negative electrode of the second capacitor is connected with the negative output end.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110807480.2A CN113619447B (en) | 2021-07-16 | 2021-07-16 | Method for predicting state of charge of battery of electric automobile |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110807480.2A CN113619447B (en) | 2021-07-16 | 2021-07-16 | Method for predicting state of charge of battery of electric automobile |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113619447A CN113619447A (en) | 2021-11-09 |
| CN113619447B true CN113619447B (en) | 2023-05-16 |
Family
ID=78380024
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110807480.2A Active CN113619447B (en) | 2021-07-16 | 2021-07-16 | Method for predicting state of charge of battery of electric automobile |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113619447B (en) |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106383315A (en) * | 2016-08-29 | 2017-02-08 | 丹阳亿豪电子科技有限公司 | New energy automobile battery state of charge (SOC) prediction method |
| CN107069122A (en) * | 2017-04-01 | 2017-08-18 | 山东省科学院自动化研究所 | A kind of Forecasting Methodology of electrokinetic cell remaining life |
| CN109143105A (en) * | 2018-09-05 | 2019-01-04 | 上海海事大学 | A kind of state-of-charge calculation method of lithium ion battery of electric automobile |
| KR20200140093A (en) * | 2019-06-05 | 2020-12-15 | 삼성에스디아이 주식회사 | Prediction Method and Prediction System for predicting Capacity Change according to Charging / Discharging Cycle of Battery |
-
2021
- 2021-07-16 CN CN202110807480.2A patent/CN113619447B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113619447A (en) | 2021-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Modeling and state-of-charge prediction of lithium-ion battery and ultracapacitor hybrids with a co-estimator | |
| Li et al. | Big data driven lithium-ion battery modeling method based on SDAE-ELM algorithm and data pre-processing technology | |
| Li et al. | Battery life estimation based on cloud data for electric vehicles | |
| Chen et al. | A new state-of-health estimation method for lithium-ion batteries through the intrinsic relationship between ohmic internal resistance and capacity | |
| Zhang et al. | Battery modelling methods for electric vehicles-A review | |
| Ahmed et al. | Model-based parameter identification of healthy and aged li-ion batteries for electric vehicle applications | |
| Wang et al. | A method for state-of-charge estimation of Li-ion batteries based on multi-model switching strategy | |
| Zhou et al. | A low-complexity state of charge estimation method for series-connected lithium-ion battery pack used in electric vehicles | |
| Eddahech et al. | Behavior and state-of-health monitoring of Li-ion batteries using impedance spectroscopy and recurrent neural networks | |
| CN111007401A (en) | Electric vehicle battery fault diagnosis method and device based on artificial intelligence | |
| Wang et al. | An improved coulomb counting method based on dual open‐circuit voltage and real‐time evaluation of battery dischargeable capacity considering temperature and battery aging | |
| Wang et al. | Estimation of battery open-circuit voltage and state of charge based on dynamic matrix control-extended Kalman filter algorithm | |
| CN102540081B (en) | Method for determining charge state of vehicle-mounted storage battery | |
| KR102259265B1 (en) | How to estimate the state of charge for a battery cell | |
| Sun et al. | Study of parameters identification method of Li-ion battery model for EV power profile based on transient characteristics data | |
| CN115684947A (en) | Battery model construction method and battery degradation prediction device | |
| CN110568360A (en) | lithium battery aging diagnosis method based on fuzzy logic algorithm | |
| CN111257770B (en) | Battery pack power estimation method | |
| Dong et al. | Low complexity state-of-charge estimation for lithium-ion battery pack considering cell inconsistency | |
| CN115389936A (en) | Online prediction method for continuous peak power capability of digital-analog hybrid driven lithium battery | |
| CN113687237B (en) | Lithium battery residual charging time prediction method for guaranteeing electrical safety | |
| Mueller et al. | Analysing the influence of driver behaviour and tuning measures on battery aging and residual value of electric vehicles | |
| Li et al. | Joint estimation of SOC and SOH for lithium-ion batteries based on EKF multiple time scales | |
| Mehta et al. | Selection of three RC branches in equivalent circuit model of lithium-ion batteries for improved accuracy | |
| CN113619447B (en) | Method for predicting state of charge of battery of electric automobile |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |