CN113820376A - Comprehensive poison monitoring method of microbial electrochemical sensor based on machine learning model - Google Patents
Comprehensive poison monitoring method of microbial electrochemical sensor based on machine learning model Download PDFInfo
- Publication number
- CN113820376A CN113820376A CN202111071935.5A CN202111071935A CN113820376A CN 113820376 A CN113820376 A CN 113820376A CN 202111071935 A CN202111071935 A CN 202111071935A CN 113820376 A CN113820376 A CN 113820376A
- Authority
- CN
- China
- Prior art keywords
- poison
- machine learning
- dropratio
- current
- microbial electrochemical
- 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.)
- Pending
Links
- 231100000614 poison Toxicity 0.000 title claims abstract description 72
- 239000002574 poison Substances 0.000 title claims abstract description 71
- 238000010801 machine learning Methods 0.000 title claims abstract description 39
- 230000000813 microbial effect Effects 0.000 title claims abstract description 38
- 238000000034 method Methods 0.000 title claims abstract description 22
- 238000012544 monitoring process Methods 0.000 title claims abstract description 21
- 238000012549 training Methods 0.000 claims description 26
- LPXPTNMVRIOKMN-UHFFFAOYSA-M sodium nitrite Chemical compound [Na+].[O-]N=O LPXPTNMVRIOKMN-UHFFFAOYSA-M 0.000 claims description 22
- 230000009467 reduction Effects 0.000 claims description 16
- 229910021380 Manganese Chloride Inorganic materials 0.000 claims description 11
- GLFNIEUTAYBVOC-UHFFFAOYSA-L Manganese chloride Chemical compound Cl[Mn]Cl GLFNIEUTAYBVOC-UHFFFAOYSA-L 0.000 claims description 11
- XMEVHPAGJVLHIG-FMZCEJRJSA-N chembl454950 Chemical compound [Cl-].C1=CC=C2[C@](O)(C)[C@H]3C[C@H]4[C@H]([NH+](C)C)C(O)=C(C(N)=O)C(=O)[C@@]4(O)C(O)=C3C(=O)C2=C1O XMEVHPAGJVLHIG-FMZCEJRJSA-N 0.000 claims description 11
- 229940099607 manganese chloride Drugs 0.000 claims description 11
- 235000002867 manganese chloride Nutrition 0.000 claims description 11
- 239000011565 manganese chloride Substances 0.000 claims description 11
- 235000010288 sodium nitrite Nutrition 0.000 claims description 11
- 229960004989 tetracycline hydrochloride Drugs 0.000 claims description 11
- 239000003440 toxic substance Substances 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 8
- 230000004044 response Effects 0.000 claims description 8
- 231100000167 toxic agent Toxicity 0.000 claims description 8
- 238000012795 verification Methods 0.000 claims description 8
- 229960000819 sodium nitrite Drugs 0.000 claims description 6
- 238000002347 injection Methods 0.000 claims description 4
- 239000007924 injection Substances 0.000 claims description 4
- 238000010200 validation analysis Methods 0.000 claims description 4
- 230000035800 maturation Effects 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 2
- 239000012528 membrane Substances 0.000 claims description 2
- 231100000331 toxic Toxicity 0.000 claims description 2
- 230000002588 toxic effect Effects 0.000 claims description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 abstract description 15
- 238000007405 data analysis Methods 0.000 abstract description 7
- 229910001385 heavy metal Inorganic materials 0.000 abstract description 2
- IOVCWXUNBOPUCH-UHFFFAOYSA-M Nitrite anion Chemical compound [O-]N=O IOVCWXUNBOPUCH-UHFFFAOYSA-M 0.000 abstract 1
- 230000003115 biocidal effect Effects 0.000 abstract 1
- 230000001360 synchronised effect Effects 0.000 abstract 1
- 238000001514 detection method Methods 0.000 description 9
- 231100000419 toxicity Toxicity 0.000 description 5
- 230000001988 toxicity Effects 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 3
- 210000004027 cell Anatomy 0.000 description 3
- 239000003344 environmental pollutant Substances 0.000 description 3
- 229910002804 graphite Inorganic materials 0.000 description 3
- 239000010439 graphite Substances 0.000 description 3
- 231100000719 pollutant Toxicity 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- NBIIXXVUZAFLBC-UHFFFAOYSA-N Phosphoric acid Chemical compound OP(O)(O)=O NBIIXXVUZAFLBC-UHFFFAOYSA-N 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000003153 chemical reaction reagent Substances 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 239000011259 mixed solution Substances 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 229910001220 stainless steel Inorganic materials 0.000 description 2
- 239000010935 stainless steel Substances 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 229910000147 aluminium phosphate Inorganic materials 0.000 description 1
- 239000003242 anti bacterial agent Substances 0.000 description 1
- 229940088710 antibiotic agent Drugs 0.000 description 1
- 239000007853 buffer solution Substances 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005518 electrochemistry Effects 0.000 description 1
- 238000011066 ex-situ storage Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000004060 metabolic process Effects 0.000 description 1
- 244000005700 microbiome Species 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 150000002826 nitrites Chemical class 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000010238 partial least squares regression Methods 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 230000027756 respiratory electron transport chain Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 229920002379 silicone rubber Polymers 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000003911 water pollution Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/26—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating electrochemical variables; by using electrolysis or electrophoresis
- G01N27/28—Electrolytic cell components
- G01N27/30—Electrodes, e.g. test electrodes; Half-cells
- G01N27/327—Biochemical electrodes, e.g. electrical or mechanical details for in vitro measurements
- G01N27/3275—Sensing specific biomolecules, e.g. nucleic acid strands, based on an electrode surface reaction
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/26—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating electrochemical variables; by using electrolysis or electrophoresis
- G01N27/28—Electrolytic cell components
- G01N27/30—Electrodes, e.g. test electrodes; Half-cells
- G01N27/308—Electrodes, e.g. test electrodes; Half-cells at least partially made of carbon
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/26—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating electrochemical variables; by using electrolysis or electrophoresis
- G01N27/416—Systems
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/30—Prediction of properties of chemical compounds, compositions or mixtures
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/70—Machine learning, data mining or chemometrics
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Electrochemistry (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Crystallography & Structural Chemistry (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Investigating Or Analyzing Materials By The Use Of Electric Means (AREA)
Abstract
The invention discloses a method for realizing synchronous monitoring of concentrations of multiple poisons by combining a machine learning model and a microbial electrochemical system. Aiming at the problem that the traditional microbial electrochemical sensor is difficult to identify different types of poisons, the system consists of a microbial electrochemical sensor for detecting different types of comprehensive poisons in a water body and a data acquisition system, wherein data analysis is carried out based on a machine learning regression model, an anode electroactive biomembrane sensing element for a microbial electrolytic cell configuration is constructed, the operation and data acquisition of a reactor are successfully realized through the data acquisition system, and the concentration of different types of poisons such as heavy metal, nitrite and antibiotic is simultaneously quantified by pertinently using different algorithms and characteristic value combinations. The invention provides a scheme for wide application of the microbial electrochemical sensor in water quality monitoring.
Description
Technical Field
The invention belongs to the technical field of environmental detection, and particularly relates to a complex poison detection method for performing data analysis based on machine learning, a microbial electrochemical sensor for implementing the method and a use method of the microbial electrochemical sensor.
Background
With the acceleration of urbanization and industrialization, the problem of water pollution becomes increasingly serious. In order to deal with the pollutants with complex types, the development of a high-precision rapid water quality detection means becomes a problem to be solved urgently. The existing water quality detection means including physical and chemical methods have the problems of high cost, long detection time, need of ex-situ analysis in a laboratory and the like. Therefore, on the basis of realizing high-precision and high-accuracy comprehensive poison water quality detection, the development of an online low-cost and in-situ real-time detection technology is particularly important. In recent years, Microbial electrochemical toxicity sensors based on a Microbial Electrochemical System (MES) have been widely studied because they can realize low-cost monitoring of broad-spectrum poisons in real time. The MES sensor generally takes an anode electroactive biomembrane as a sensing unit, and when the MES sensor is impacted by poison, the metabolism of electroactive microorganisms is influenced, the electron transfer rate is reduced, and the electric signal generated by the system is correspondingly changed. However, since the electroactive biomembrane can completely respond to all toxicant impacts as the change of the electric signal, when dealing with the actual water body with complex pollutants, researchers are difficult to directly obtain the information of each toxicant by analyzing a single electric signal, which hinders the further application of the MES sensor to the early toxicity early warning and water quality monitoring of the actual water body.
Machine learning is a process of deep-level law mining through data, and has been widely applied to water treatment and environmental monitoring in recent years. In the frame of machine learning, a reasonable statistical model is constructed by selecting proper algorithms and parameters, one part of data set is used as sample data (training set), the other part of data set is used as a verification set to test the accuracy of the model, and finally accurate prediction and decision can be made on newly input data. Through regression modeling, deep analysis of the relationship between different types of poisons and response electric signals can be realized through machine learning, and quantification of the MES sensor on the various poisons is finally realized.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a method for a comprehensive poison monitoring system of a microbial electrochemical sensor based on machine learning data analysis.
The technical scheme of the invention is a method of a comprehensive poison monitoring system of a microbial electrochemical sensor for carrying out data analysis based on machine learning, which specifically comprises the following steps:
1. a comprehensive poison monitoring method of a microbial electrochemical sensor based on a machine learning model is realized by the following steps: 1) carrying out data processing on the collected electric signal data of the microbial electrochemical sensors before and after the microbial electrochemical sensors receive toxic impact, and finally obtaining the electric signal data of each microbial electrochemical sensor arranged according to a time sequence;
2) for each reactor, processing experimental data and collected electric signal data to obtain a characteristic value of a descriptive electric signal and a characteristic value of response poison impact;
3) training a machine learning model, selecting a specific algorithm according to the type of a poison, selecting a specific descriptive electric signal characteristic value and a response poison impact characteristic value as input, performing regression model training by taking the concentration of the poison as output, taking one part of a processed electric signal data set as a training set and the other part as a verification set, and predicting the poison concentration of a verification set sample through the model;
4) machine learning model evaluation
And evaluating the precision of the machine learning model, and after the model is accurate enough, the predicted value can be used as the actual poison concentration, so that the comprehensive poison concentration is detected.
2. Optionally, the characteristic values of the descriptive electrical signal are mainly the maximum current (I _ max) generated during the MEC reactor operation, the biofilm acclimation maturation in the reactor, the steady state current (I _ stable) and the time required for the system to reach steady state (t _ stable), the time point for poison injection (t _ inject).
3. Optionally, the characteristic value of the response poison impact is mainly the current and the current reduction rate of every 6 hours after the poison impact, namely, the current (I _6h) and the current reduction rate (DropRatio _6h) of 6 hours, the current (I _12h) and the current reduction rate (DropRatio _12h) of 12 hours, the current (I _18h) and the current reduction rate (DropRatio _18h) of 18 hours, the current (I _24h) of 24 hours and the current reduction rate (DropRatio _24 h).
4. Optionally, the maximum current (I _ max) generated during reactor operation is the third highest current value in the reactor electrical signal data to eliminate possible measurement deviation and outlier caused by accidental disconnection during operation; the reactor current reaches 90% of I _ max, namely entering a steady state, the time point is the time (t _ stable) required for reaching the steady state, and the average current after entering the steady state and before being impacted by the poison is the steady state current (I _ stable); from the time of injection of the poison, the reactor is impacted by the poison; the current reduction rate (DropRatio) is formulated as:
wherein DropRatio is DropRatio _6h, DropRatio _12h, DropRatio _18h, DropRatio _24h, and I isdropDrop _6h, Drop _12h, Drop _18h, Drop _24 h.
5. Optionally, the characteristic group for training the manganese chloride quantitative model is I _ max, I _ stable, t _ stable and DropRatio _12 h; the characteristic group used for training the sodium nitrite quantitative model is I _ max, I _ stable, t _ stable and DropRatio _24 h; the characteristic groups used for training the tetracycline hydrochloride quantitative model are I _ max, I _ stable, t _ stable, DropRatio _6h and DropRatio _12 h.
6. Alternatively, 90% of the data is used for the training set and 20% is used for the validation set.
7. Alternatively, the machine learning models for training manganese chloride, sodium nitrite, and tetracycline hydrochloride are Partial Least Squares (PLS), K Nearest Neighbor (KNN), and Neural Network (NNET), respectively.
Advantageous effects
Compared with the prior art, the invention combines the microbial electrochemical water quality monitoring at the front end with the machine learning data analysis at the rear end, overcomes the defect that the traditional microbial electrochemical sensor cannot simultaneously identify comprehensive toxicants through the operation and the toxicant impact of reactors in the same batch, realizes the quantification of the concentration of various toxicants, and provides a new technology for further applying the microbial electrochemical sensor to the water quality detection and toxicity early warning of the water body with complex pollutants. The machine-learned models are assembled from models that quantify each specific poison, using different algorithms and feature sets for each poison, providing the possibility to customize the model for a specific body of water and to build a broad spectrum poison analysis database. The use of the wireless data acquisition system provides wide application prospect for the intellectualization and the Internet of things of the microbial electrochemical online water quality monitoring and early warning system.
Drawings
FIG. 1 is a water quality detection flow chart of the microbial electrochemical sensor based on machine learning data analysis.
FIG. 2 is a time current plot for a MEC reactor with a manganese chloride concentration of 9mg/L, a sodium nitrite concentration of 9mg/L, and a tetracycline hydrochloride concentration of 6 mg/L.
FIG. 3 is a diagram illustrating the definition of characteristic values of electrical signal data
FIG. 4 is a graph of the results of machine learning model versus the prediction of the concentration of three poisons in the validation set reactor.
Detailed Description
Example 1: microbial electrochemical sensor for simultaneously detecting manganese chloride, sodium nitrite and tetracycline hydrochloride based on machine learning data analysis
Firstly, constructing a microbial electrolytic cell reactor taking an anode electroactive biomembrane as a sensing element and carrying out comprehensive poison monitoring
1) Construction of microbial electrolytic cell reactor with anode electroactive biomembrane as sensing element
The electroactive biofilm is enriched by a two-electrode MEC system. The main body of the reactor is a 100mL reagent bottle with a blue cover (not limited to the size); 1.5X 1.5cm2The stainless steel net is vertical to the bottom surface of the reagent bottle and serves as a cathode; a graphite rod with the diameter of 1cm and the height of 1.5cm is used as an anode, one bottom surface of the graphite rod is parallel to the stainless steel net, and the other surfaces of the graphite rod are covered by silicon rubber, so that the bottom surface is the only surface for enriching the biological membrane. The reactor was applied with an external voltage of 0.7V to enrich the anodic biofilm, and the biofilm status was monitored galvanically.
2) Microbial electrolytic cell sensor using anode electroactive biomembrane as sensing element for toxicity monitoring
After all reactors have reached their steady state current, poisons are added to the reactors as shown in figure 2. Manganese chloride, sodium nitrite, and tetracycline hydrochloride were added to simulate the effect of complex toxicants containing heavy metals, nitrites, antibiotics on electroactive biofilms. In order to ensure that the electric signal data has enough representativeness and universality as sample data of a machine learning model, the concentration ranges of the three poisons are all 1-10 mg/L, and the concentration of each poison in each reactor is randomly generated and is not completely the same. Three poisons and 50mmol/L phosphoric acid buffer solution (PBS) are prepared into a mixed solution of 1mL of poisons, and the mixed solution is injected into the reactor, and the current change of the reactor after being impacted by the poisons is continuously monitored under the condition that the pH value and the substrate concentration of the reactor are not influenced. And finally obtaining 23 data of the MEC reactor after being impacted by poisons with different concentrations and electric signals of the MEC reactor. Secondly, carrying out data processing and machine learning modeling on collected sample data
1) Data cleaning processing is carried out on the acquired electric signal data
The electric signal data stored by the computer is provided with a sampling time stamp, an IP address of a data acquisition system, an equipment number and other labels, and the data is processed into an electric signal data pattern which is arranged according to a time sequence by each reactor so as to facilitate subsequent characteristic value extraction and machine learning modeling.
2) Feature group extraction and feature value selection for electrical signal data
The electric signal data of 23 reactors after data arrangement was used as a database. As in fig. 3, for each MEC reactor, a series of characteristic values are obtained by processing the experimental data and the acquired electrical signal data: characteristic values of the descriptive electrical signals, including the maximum current (I _ max) generated during the MEC reactor operation, the maturation of the biofilm in the reactor, the steady state current (I _ stable) and the time required for the system to reach steady state (t _ stable), the point in time of poison injection (t _ inject); the response poison impact characteristic values comprise current (Drop _6h) and current reduction rate (Drop _6h) 6 hours after the poison impact, current (Drop _12h) and current reduction rate (Drop _12h) 12 hours after the poison impact, current (Drop _18h) and current reduction rate (Drop _18h) 18 hours after the poison impact, current (Drop _24h) 24 hours after the poison impact and current reduction rate (Drop _24h), and the definition and calculation method of the characteristic values are the same as those described above.
All the characteristic value data are normalized, and each characteristic value data set is mapped to a (0, 1) interval. The normalization processing formula is as follows:
and randomly selecting the data sets of 20 reactors as training sets to train the machine learning model, and using the other 3 reactors as training sets to verify that the model can realize concentration prediction and evaluate the prediction accuracy of the model.
The characteristic group used for training the manganese chloride concentration model is I _ max, I _ stable, t _ stable and DropRatio _12 h; the characteristic group used for training the sodium nitrite concentration model is I _ max, I _ stable, t _ stable and DropRatio _24 h; the characteristic group used for training the tetracycline hydrochloride concentration model is I _ max, I _ stable, t stable, DropRatio _6h and DropRatio _12 h.
3) Machine learning modeling using different algorithms for different poisons
The machine learning models for training manganese chloride, sodium nitrite and tetracycline hydrochloride are respectively partial least squares algorithm (PLS), K nearest neighbor algorithm (KNN) and neural network algorithm (NNET). Because the concentration value of the poison is a continuous variable, a regression model is established based on the training current collection signal characteristic value data to carry out supervised machine learning, and the verification centralized electric signal characteristic value data is input into the model to obtain the concentration of each poison for predicting the continuous variable.
The partial least squares regression analysis is a regression analysis method aiming at multiple eigenvalues and multiple output values, and realizes a multivariate regression method for carrying out variable correlation analysis and principal component analysis in combination with typical correlation analysis to carry out data simplification. Through information integration and screening of the electrical signal adjustment data, the PLS algorithm can extract new comprehensive variables from the entire feature set with the best interpretative ability for toxicant concentration for regression modeling. The K-nearest neighbor method is to find a predetermined number (K) of points closest in distance (standard euclidean distance) to a new point from a trained electrical signal data sample and then predict from these points. Based on the electrical signal data feature set of the continuous label, the KNN can realize the training of the regression model. In the regression model, training of the neural network relies on passing through a Multi-layer Perceptron (MLP). The NNET under-framework multilayer perceptron includes an input layer and an output layer, and a non-linear hidden layer. The initial input layer is composed of a group of neurons { x _ i | x _1, x _2, …, x _ m } representing input features, namely a series of electric signal feature values for modeling; neurons in the hidden layer perform weighted linear sum transformation w on the values of the previous layer1x1+w2x2+…+wmxmAnd a non-linear activation function transformation g (·): r → R (identity transform in the regression model); and finally, the output layer receives the continuous variable transformed by the hidden layer. 4) Evaluating machine learning model prediction accuracy
After a regression model for manganese chloride, sodium nitrite and tetracycline hydrochloride was trained by corresponding algorithms and feature sets, the set of feature values for each reactor in the validation set was input to the model to obtain the predicted concentrations of different poisons, as shown in fig. 4. It can be seen that the concentration predicted by the model is very small in difference with the concentration of the actually injected poison, which primarily indicates that the quantitative model of the three poisons has enough accuracy.
Model accuracy was further evaluated by calculating the Root Mean Square0 Error (RMSE) and Mean Absolute Error (MAE) of the model.
The MAE measures the average value of all errors in the prediction process, and the calculation method is as follows:
wherein,
the predicted value output given the feature set of the system sample i, yi is the actual value.
In the RMSE measurement prediction process, the standard deviation of prediction errors is calculated as follows:
wherein,
predicted value, y, output for a given set of features of a system sample iiIs the actual value.
Through the verification of the verification set, the MAE of the concentration quantitative model of the manganese chloride, the sodium nitrite and the tetracycline hydrochloride can be reduced to 0.20, 0.18 and 0.26, and the RMSE can be reduced to 0.21, 0.20 and 0.23.
The electric signal data monitored by a large amount of reaction toxicity is proved, and the quantitative detection of various poisons can be simultaneously realized on the same batch of microbial electrochemical sensors by constructing different machine learning algorithms, so that the foundation is laid for the further popularization of microbial electrochemistry.
The present invention is not limited to the above preferred embodiments, and any modifications, equivalent replacements, improvements, etc. within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (7)
1. A comprehensive poison monitoring method of a microbial electrochemical sensor based on a machine learning model is characterized by comprising the following steps:
1) carrying out data processing on the collected electric signal data of the microbial electrochemical sensors before and after the microbial electrochemical sensors receive toxic impact, and finally obtaining the electric signal data of each microbial electrochemical sensor arranged according to a time sequence;
2) for each reactor, processing experimental data and collected electric signal data to obtain a characteristic value of a descriptive electric signal and a characteristic value of response poison impact;
3) training a machine learning model, selecting a specific algorithm according to the type of a poison, selecting a specific descriptive electric signal characteristic value and a response poison impact characteristic value as input, performing regression model training by taking the concentration of the poison as output, taking one part of a processed electric signal data set as a training set and the other part as a verification set, and predicting the poison concentration of a verification set sample through the model;
4) machine learning model evaluation
And evaluating the precision of the machine learning model, and after the model is accurate enough, the predicted value can be used as the actual poison concentration, so that the comprehensive poison concentration is detected.
2. The method for monitoring synthetic poisons for microbial electrochemical sensors based on machine learning models, according to claim 1, characterized in that the characteristic values of the descriptive electric signals are mainly the maximum current (I _ max) generated during the operation of the MEC reactor, the acclimation and maturation of the biological membranes in the reactor, the steady state current (I _ stable) for the system to reach the steady state and the time (t _ stable) required for reaching the steady state, the time point (t _ inject) for injecting poisons.
3. The method for monitoring the integrated poison of the microbial electrochemical sensor based on the machine learning model is characterized in that the response poison impact characteristic value is mainly the current and the current reduction rate of every 6 hours after the poison impact, namely the current (I _6h) and the current reduction rate (Dropratio _6h) in 6 hours, the current (I _12h) and the current reduction rate (Dropratio _12h) in 12 hours, the current (I _18h) and the current reduction rate (Dropratio _18h) in 18 hours, the current (I _24h) in 24 hours and the current reduction rate (Dropratio _24 h).
4. The method for monitoring synthetic poisons for microbial electrochemical sensors based on machine learning models is characterized in that the maximum current (I _ max) generated during the operation of the reactor is the third highest current value in the electric signal data of the reactor, so as to eliminate possible measurement deviation and outliers caused by accidental disconnection during the operation; the reactor current reaches 90% of I _ max, namely entering a steady state, the time point is the time (t _ stable) required for reaching the steady state, and the average current after entering the steady state and before being impacted by the poison is the steady state current (I _ stable); from the time of injection of the poison, the reactor is impacted by the poison; the current reduction rate (DropRatio) is formulated as:
wherein DropRatio is DropRatio _6h, DropRatio _12h, DropRatio _18h, DropRatio _24h, and I isdropDrop _6h, Drop _12h, Drop _18h, Drop _24 h.
5. The comprehensive poison monitoring method of the microbial electrochemical sensor based on the machine learning model is characterized in that a characteristic group for training a manganese chloride quantitative model is I _ max, I _ stable, t _ stable and DropRatio _12 h; the characteristic group used for training the sodium nitrite quantitative model is I _ max, I _ stable, t _ stable and DropRatio _24 h; the characteristic groups used for training the tetracycline hydrochloride quantitative model are I _ max, I _ stable, t _ stable, DropRatio _6h and DropRatio _12 h.
6. The method for comprehensive toxicant monitoring of the microbial electrochemical sensor based on the machine learning model is characterized in that 90% of data is used in a training set and 20% of data is used in a validation set.
7. The method for monitoring the comprehensive toxic substances of the microbial electrochemical sensor based on the machine learning model is characterized in that the machine learning models for training manganese chloride, sodium nitrite and tetracycline hydrochloride are respectively Partial Least Squares (PLS), K-Nearest Neighbor (KNN) and Neural Network (NNET).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111071935.5A CN113820376A (en) | 2021-09-14 | 2021-09-14 | Comprehensive poison monitoring method of microbial electrochemical sensor based on machine learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111071935.5A CN113820376A (en) | 2021-09-14 | 2021-09-14 | Comprehensive poison monitoring method of microbial electrochemical sensor based on machine learning model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113820376A true CN113820376A (en) | 2021-12-21 |
Family
ID=78914494
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111071935.5A Pending CN113820376A (en) | 2021-09-14 | 2021-09-14 | Comprehensive poison monitoring method of microbial electrochemical sensor based on machine learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113820376A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118503787A (en) * | 2024-06-14 | 2024-08-16 | 广东粤海水务检测技术有限公司 | Data analysis method and device based on electroactive microorganism detection |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995011989A1 (en) * | 1993-10-29 | 1995-05-04 | Invitro International | In vitro fluid toxicity bioassay |
| US6113762A (en) * | 1995-02-22 | 2000-09-05 | Isao Karube | Microbial electrode and microbial sensor |
| CN103868955A (en) * | 2014-04-09 | 2014-06-18 | 西南大学 | Portable in-automobile gas detection system and method |
| CN109642886A (en) * | 2016-06-29 | 2019-04-16 | 新加坡国立大学 | Monitoring of poison system |
| US20190304568A1 (en) * | 2018-03-30 | 2019-10-03 | Board Of Trustees Of Michigan State University | System and methods for machine learning for drug design and discovery |
| US20190389751A1 (en) * | 2016-11-25 | 2019-12-26 | Island Water Technologies Inc. | Bio-electrochemical sensor and method for optimizing performance of a wastewater treatment system |
| CN112432987A (en) * | 2020-11-06 | 2021-03-02 | 南开大学 | Method for constructing toxicity early warning system of photoautotrophic oxygen reduction biological cathode sensor |
-
2021
- 2021-09-14 CN CN202111071935.5A patent/CN113820376A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995011989A1 (en) * | 1993-10-29 | 1995-05-04 | Invitro International | In vitro fluid toxicity bioassay |
| US6113762A (en) * | 1995-02-22 | 2000-09-05 | Isao Karube | Microbial electrode and microbial sensor |
| CN103868955A (en) * | 2014-04-09 | 2014-06-18 | 西南大学 | Portable in-automobile gas detection system and method |
| CN109642886A (en) * | 2016-06-29 | 2019-04-16 | 新加坡国立大学 | Monitoring of poison system |
| US20190389751A1 (en) * | 2016-11-25 | 2019-12-26 | Island Water Technologies Inc. | Bio-electrochemical sensor and method for optimizing performance of a wastewater treatment system |
| US20190304568A1 (en) * | 2018-03-30 | 2019-10-03 | Board Of Trustees Of Michigan State University | System and methods for machine learning for drug design and discovery |
| CN112432987A (en) * | 2020-11-06 | 2021-03-02 | 南开大学 | Method for constructing toxicity early warning system of photoautotrophic oxygen reduction biological cathode sensor |
Non-Patent Citations (1)
| Title |
|---|
| 徐莹;代燕;李超;梅婉欣;杨勇;: "基于支持向量回归的生物膜-电极阻抗模型研究", 航天医学与医学工程, no. 06 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118503787A (en) * | 2024-06-14 | 2024-08-16 | 广东粤海水务检测技术有限公司 | Data analysis method and device based on electroactive microorganism detection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109709057B (en) | Water quality index prediction model construction method and water quality index monitoring method | |
| CN111948272B (en) | Water quality biotoxicity prediction method based on artificial neural network | |
| CN111553468A (en) | Method for accurately predicting effluent quality of sewage treatment plant | |
| Ba-Alawi et al. | Missing data imputation and sensor self-validation towards a sustainable operation of wastewater treatment plants via deep variational residual autoencoders | |
| CN107782857A (en) | Flexible measurement method based on the edible fungus fermented process key parameter for improving CS BPNN | |
| CN111693667A (en) | Water quality detection system and method based on gated recursive array | |
| Gallardo et al. | Use of an electronic tongue based on all‐solid‐state potentiometric sensors for the quantitation of alkaline ions | |
| Bonastre et al. | In-line chemical analysis of wastewater: present and future trends | |
| CN114578011A (en) | Water quality monitoring method based on multi-sensor multi-source data fusion | |
| CN115456245A (en) | Prediction method for dissolved oxygen in tidal river network area | |
| CN113820376A (en) | Comprehensive poison monitoring method of microbial electrochemical sensor based on machine learning model | |
| CN104915563B (en) | The chronic reference prediction method of fresh water based on metal quantitative structure activity relationship | |
| CN118014293A (en) | Industrial park intelligent patrol early warning system based on artificial intelligence | |
| Eo et al. | Deep learning framework with essential pre-processing techniques for improving mixed-gas concentration prediction | |
| CN110222825B (en) | Cement product specific surface area prediction method and system | |
| CN112069621B (en) | Method for predicting residual service life of rolling bearing based on linear reliability index | |
| CN113960090A (en) | LSTM neural network algorithm-based soil Cd element spectrum qualitative analysis method | |
| CN115031794B (en) | Novel gas-solid two-phase flow measuring method based on multi-feature graph convolution | |
| CN113449789A (en) | Quality control method for monitoring water quality by full-spectrum water quality monitoring equipment based on big data | |
| CN112707508A (en) | BP-ANN-based agricultural sewage facility COD treatment effect prediction method, device and platform | |
| Sunori et al. | K-Means and FCM Clustering of Biological Oxygen Demand of Water | |
| Hu et al. | The early warning model of dust concentration in smart construction sites based on long short term memory network | |
| CN116467645B (en) | Pollution source collection monitoring system | |
| CN114878599B (en) | Oil-water two-phase flow water content measuring method based on time-frequency chart convolution | |
| Liu et al. | Polluted gas quantitative detection in multi-gas sensor based on bidirectional long-short term memory network |
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 |