CN111027014B - TSK fuzzy model particle filter method and system for type-2 intuitionistic fuzzy decision-making - Google Patents

Abstract

The invention discloses a type-2 intuitionistic fuzzy decision-making TSK fuzzy model particle filter method and system, which utilizes a type-2 intuitionistic fuzzy decision-making objective function to select an optimal feature set that can reflect the characteristics of a moving target; constructs a TSK fuzzy model, and The optimal feature set is used as input, and the weight of each rule is calculated based on the antecedent membership function of the TSK fuzzy model, and the state fusion results of multiple fuzzy rules are obtained by using the method of weighted summation as the output of the TSK fuzzy model; The important density function of the particle filter, extracting particles from the important density function for filter update. The present invention uses the type-2 intuitionistic fuzzy decision-making objective function to select the optimal feature set that can reflect the characteristics of the moving target to construct the TSK fuzzy model, which not only reduces the number of rules of the model and the redundancy of the model, but also improves the accuracy of the model. The output of the TSK fuzzy model is used as an important density function of the particle filter algorithm to greatly reduce the degradation of particles.

Description

TSK fuzzy model particle filtering method and system for type-2 intuitionistic fuzzy decision making

Technical Field

The present invention relates to the field of particle filtering, and in particular to a TSK fuzzy model particle filtering method and system for type-2 intuitive fuzzy decision making.

Background Art

Modeling technology is widely used in many fields for accurate modeling of nonlinear systems. Fuzzy systems have a good ability to describe the complex dynamics of nonlinear behaviors of dynamic processes. Fuzzy model identification is an effective tool for high-precision modeling of complex nonlinear systems based on measured data. By utilizing the nonlinear mapping capability of fuzzy logic, complex nonlinear systems defined on a compact set can be consistently approximated to arbitrary precision. In the TSK fuzzy model modeling process, structural identification and parameter identification are both very critical steps that determine the quality of the model. The related work of structural identification includes the determination of the number of rules, the selection of structural parameters and the partitioning of fuzzy space, that is, the decision problem about the antecedent characteristic variables.

A set of features may contain enough information to reflect the movement trend of the target. When a modeling problem is defined by features, the number of features may be quite large, and many of them may be irrelevant or redundant. Because irrelevant information is cached in the whole of features, these irrelevant features may reduce the performance of modeling using all features. When particle filtering, due to the high number of rules and redundancy of the model, the particle degradation caused is high.

Summary of the invention

Therefore, the present invention provides a TSK fuzzy model particle filtering method and system for type-2 intuitive fuzzy decision-making, which overcomes the defects of high particle degradation degree caused by high number of model rules and high redundancy of the model in the prior art.

In the first aspect, an embodiment of the present invention provides a TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision, comprising the following steps: using the type-2 intuitive fuzzy decision objective function to select the optimal feature set that can reflect the characteristics of a moving target; constructing a TSK fuzzy model, taking the optimal feature set as input, calculating the weight of each rule based on the antecedent membership function of the TSK fuzzy model, and using the weighted summation method to obtain the state fusion result of multiple fuzzy rules as the output of the TSK fuzzy model; constructing an important density function of the particle filter based on the state fusion result, and extracting particles from the important density function for filtering update.

In one embodiment, the step of using a type-2 intuitive fuzzy decision objective function to select an optimal feature set that can reflect the characteristics of a moving target includes: calculating the similarity of each feature in a preliminary feature set of the moving target by using a ridge descending function; constructing an objective function based on a type-2 intuitive fuzzy decision according to the similarity, and selecting the optimal feature set that can reflect the characteristics of the moving target.

In one embodiment, the feature similarity calculations include: speed similarity calculations, innovation similarity calculations, heading angle similarity calculations, and time interval similarity calculations.

In one embodiment, the step of constructing an objective function based on type-2 intuitive fuzzy decision according to the similarity and selecting the optimal feature set that can reflect the characteristics of the moving target includes: fusing the similarities of each feature; based on the objective function of type-2 intuitive fuzzy decision, selecting features whose intuitive fuzzy decision scores are greater than a preset threshold to constitute the optimal feature set.

其中，Q+1为初选特征集中所有特征的个数。In one embodiment, the preset threshold

Among them, Q+1 is the number of all features in the initial feature set.

In one embodiment, during the TSK fuzzy model construction process, a fuzzy C regression clustering algorithm based on correlation entropy and spatiotemporal information is used to identify antecedent parameters, and a strong tracking algorithm is used to estimate consequent parameters.

In one embodiment, the step of performing particle filtering according to the important density function includes: sampling from the important density function to obtain a particle state set; calculating the weights of the particles in the particle state set and standardizing the weights; and calculating the state and covariance of the particles based on the standardized weights and the particle state set.

In the second aspect, an embodiment of the present invention provides a TSK fuzzy model particle filter system for type-2 intuitive fuzzy decision, including: an optimal feature set acquisition module, which is used to use the type-2 intuitive fuzzy decision objective function to select the optimal feature set that can reflect the characteristics of the moving target; a TSK fuzzy model construction module, which is used to construct a TSK fuzzy model, taking the optimal feature set as input, calculating the weight of each rule based on the antecedent membership function of the TSK fuzzy model, and using the weighted summation method to obtain the state fusion result of multiple fuzzy rules as the output of the TSK fuzzy model; a particle filtering module, which is used to construct an important density function of the particle filter based on the state fusion result, and extract particles from the important density function for filtering update.

In a third aspect, an embodiment of the present invention provides a terminal, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision-making described in the first aspect of the embodiment of the present invention.

In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision-making described in the first aspect of the embodiment of the present invention.

The technical solution of the present invention has the following advantages:

The present invention provides a TSK fuzzy model particle filtering method and system for type-2 intuitive fuzzy decision making, which uses a type-2 intuitive fuzzy decision objective function to select an optimal feature set that can reflect the characteristics of a moving target; construct a TSK fuzzy model, use the optimal feature set as input, calculate the weight of each rule based on the antecedent membership function of the TSK fuzzy model, and use a weighted summation method to obtain a state fusion result of multiple fuzzy rules as the output of the TSK fuzzy model; construct an important density function of particle filtering based on the state fusion result, and perform particle filtering according to the important density function. The present invention uses a type-2 intuitive fuzzy decision objective function to select an optimal feature set that can reflect the characteristics of a moving target to construct a TSK fuzzy model, which can reduce the number of rules and the redundancy of the model, and improve the accuracy of the model. The output result of the constructed TSK fuzzy model is used as the important density function of the particle filtering algorithm, which greatly reduces the degradation problem of particles.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

FIG1 is a flowchart of a specific example of a TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision-making provided by an embodiment of the present invention;

FIG2 is a structural diagram of a type-2 intuitive fuzzy feature decision method provided by an embodiment of the present invention;

FIG3 is a schematic diagram of a descending ridge distribution provided by an embodiment of the present invention;

FIG4 is a schematic diagram of a TSK fuzzy model provided by an embodiment of the present invention;

FIG5 is a schematic diagram of an estimated trajectory and root mean square error of a target motion provided by an embodiment of the present invention;

FIG6 is a schematic diagram of the composition of a TSK fuzzy model particle filter system for type-2 intuitive fuzzy decision making provided by an embodiment of the present invention;

FIG. 7 is a schematic diagram of the module composition of a terminal provided in an embodiment of the present invention.

DETAILED DESCRIPTION

The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

Example 1

The embodiment of the present invention provides a TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision making, as shown in FIG1 , comprising the following steps:

Step S1: Use the type-2 intuitionistic fuzzy decision objective function to select the optimal feature set that can reflect the characteristics of the moving target.

In practice, the degree of membership of the characteristic attributes of a moving target is a problem that needs to be solved. The characteristic attributes that can best reflect the mobility of the target are selected by adopting the method of evidence theory decision-making. It has no consistency requirements on the feature types and measurement sizes of various different features in the input serial combined feature vector, and there is no need to do any preprocessing on the serial combined feature vector. The fuzzy distribution function is used to process each feature separately, and it can process the combined features formed by simply combining various different features. The embodiment of the present invention uses the type-2 intuitive fuzzy decision objective function to select the optimal feature set that can reflect the characteristics of the moving target as the feature subset of the "optimal" decision for the moving target.

Step S2: Construct a TSK fuzzy model, take the optimal feature set as input, calculate the weight of each rule based on the antecedent membership function of the TSK fuzzy model, and use the weighted summation method to obtain the state fusion result of multiple fuzzy rules as the output of the TSK fuzzy model.

The embodiment of the present invention constructs a TSK fuzzy model based on a feature subset of the "optimal" decision of a moving target, which can reduce the number of rules and the redundancy of the model and improve the accuracy of the model.

Step S3: construct an important density function of particle filtering based on the state fusion result, and perform particle filtering according to the important density function.

The embodiment of the present invention adopts a particle filter algorithm as an important density function, thereby improving the diversity of sampled particles to a certain extent and greatly reducing the particle degradation problem.

In one embodiment, the process of executing step S1 includes: calculating the similarity of each feature in the preliminary feature set of the moving target through the ridge descending function; constructing an objective function based on type-2 intuitive fuzzy decision according to the similarity, and selecting the optimal feature set that can reflect the characteristics of the moving target. Among them, the similarity calculations of each feature include: speed similarity calculation, new information similarity calculation, heading angle similarity calculation and time interval similarity calculation. The specific implementation process is shown in Figure 2. After the target features are similarity calculated, the similarities of each feature are fused, and based on the objective function of type-2 intuitive fuzzy decision, the features with intuitive fuzzy decision scores greater than the preset threshold are selected to form the optimal feature set.

In the embodiment of the present invention, at each moment of the target movement, each feature at that moment is compared with the feature in the real trajectory to obtain the feature difference at that moment. The smaller the feature difference, the higher the similarity of the samples characterized by this dimension of feature. At the same time, in order to reduce the influence of the error generated during feature extraction on the judgment result, the fuzzy distribution needs to satisfy the condition that when the feature difference is small, it decreases slowly, and when the feature difference is large, the similarity decreases rapidly. This embodiment uses the feature difference of velocity v, new information r and heading angle difference θ as independent variables, and adopts descending ridge distribution to calculate the similarity between features on the same dimension. The distribution is expressed by the following formula:

Among them, x represents the size of the feature difference, _a1 represents the minimum value of the feature difference, and _a2 represents the maximum value of the feature difference. The distribution is shown in Figure 3.

The embodiment of the present invention calculates the similarity of the i-th feature by formula (2):

特征差矩阵

df_ji＝|f_ji-f_(j-1)i|,j＝1,2,Λ,K,i＝1,2,Λ,Q，j表示时间。in,

Characteristic Difference Matrix

df _ji = |f _ji -f _(j-1)i |, j = 1, 2, Λ, K, i = 1, 2, Λ, Q, where j represents time.

In one embodiment, assuming that the domain of the time interval T is M, the membership of the time interval T adopts a Gaussian similarity function, which is as follows:

Where σ _T , c _T represent the standard deviation and mean of the time interval respectively. From formula (3), we can see that the closer to the mean, the greater the similarity, and vice versa.

After preprocessing the measurement, the embodiment of the present invention uses the intuitive fuzzy decision scores of factors such as time interval T, speed v, new information r and heading angle difference θ to select features with intuitive fuzzy decision scores greater than a threshold τ, and then uses them as the features that best reflect the maneuverability of the target to construct a TSK fuzzy model. In this way, the motion characteristics of the target can be more fully utilized, and the burden of increased calculation caused by too many fuzzy rules caused by multiple features can be avoided.

Since the intuitionistic fuzzy set can solve the uncertainty of the target motion model, the embodiment of the present invention adopts the fusion of the intuitionistic fuzzy membership algorithm and the Type-2 fuzzy C-means regression algorithm to deal with the impact of nonlinear noise and establish the following objective function:

代表各个特征的相似度，Q+1代表特征的个数，C代表聚类数，μ_ij(M_i)表示在时刻j特征M_i的隶属度。in,

represents the similarity of each feature, Q+1 represents the number of features, C represents the number of clusters, and μ _ij (M _i ) represents the membership of feature M _i at time j.

The embodiment of the present invention minimizes the objective function according to the Lagrangian method to obtain the upper and lower memberships:

为下隶属度，m₁,m₂是模糊权重指数。Among them, μ′ _i is the upper membership degree,

is the lower membership degree, m ₁ and m ₂ are fuzzy weight indexes.

In order to combine the intuitive fuzzy characteristics, the membership function is expanded from the traditional fuzzy set to the intuitive fuzzy set. In the embodiment of the present invention, the membership function is modified by introducing the intuitive index π _A(x) :

分别表示直觉指数π_A(x)中隶属度权值和非隶属度权值。in

They represent the membership weight and non-membership weight in the intuitive index π _A (x) respectively.

The embodiment of the present invention normalizes the intuitionistic fuzzy membership to obtain the final intuitionistic fuzzy membership:

Intuitionistic index, membership and non-membership are three members of intuitionistic fuzzy sets, which have a normalized relationship. If the values of any two of the three members are known, the values of the remaining members can be obtained using the normalized relationship. Intuitionistic fuzzy sets are usually represented by membership function, non-membership function and intuitionistic index. Non-membership can be generated using a specific function. Among them, the intuitionistic index is calculated as follows:

表示如下：and

It is expressed as follows:

_Mi represents the i-th attribute measured. The decision result is:

仅以此举例，不以此为限。Threshold value of the embodiment of the present invention

This is just an example and is not intended to be limiting.

The embodiment of the present invention uses the "optimal" features obtained based on type-2 intuitive fuzzy feature decision to model the system with a TSK fuzzy model, which not only realizes the adaptive selection of features, but also realizes the adaptive adjustment of TSK rules.

The embodiment of the present invention constructs a TSK fuzzy model in a particle filter algorithm based on the features obtained from the above decision results. Generally, the TSK fuzzy model considers that any nonlinear system can be represented by the following N _f fuzzy linear model tables:

Model i:

表示规则的前件参数，

表示模型i中第G个前件参数对应的模糊集，

和

分别表示状态转移矩阵和观测矩阵,，TSK模糊模型可以通过如图4所示的6层网络结构得来，其中d>m。in,

Represents the antecedent parameter of the rule,

represents the fuzzy set corresponding to the Gth antecedent parameter in model i,

and

They represent the state transfer matrix and the observation matrix respectively. The TSK fuzzy model can be obtained through the 6-layer network structure shown in Figure 4, where d>m.

In order to obtain a more accurate TSK fuzzy network model, the embodiment of the present invention identifies and estimates the antecedent parameters and consequent parameters in the model respectively. Since the antecedent parameter membership function in the fuzzy layer is set to a Gaussian function, the embodiment of the present invention uses a fine-tuning method to identify the antecedent parameters by the mean and standard deviation of the Gaussian function, and uses a strong tracking algorithm to estimate the consequent parameters.

以及协方差估计

对于每个粒子，以状态估计

以及协方差估计

构建改进的先验概率密度函数：

In the particle filter framework, the TS fuzzy model mainly generates a prior probability density function for each particle that contains rich spatial information. Given the state estimation of the TS fuzzy model obtained

And the covariance estimate

For each particle, the state is estimated

And the covariance estimate

Construct an improved prior probability density function:

Usually, the prior probability density function is selected as the important density function, that is,

Combining the above formula with the weight update formula, the particle weight is calculated as follows:

In a specific embodiment, using the method provided by the embodiment of the present invention, the state equation and measurement equation of the moving target are obtained as follows:

状态向量，x_k表示目标x轴坐标，y_k表示目标y轴坐标，

和

分别表示目标在x轴和y轴坐标对应的速度。过程噪声e_k假设是服从零均值和协方差为σ_i,e的高斯过程噪声Q，其中Q是一个2×2矩阵(Q_ij＝0,for i≠j,Q＝diag(σ_i,e,σ_i,e))。观测噪声v_k假设为服从非高斯分布噪声

初始状态x₀由目标初始位置决定x₀＝[1km,0.15km/s,6km,0.26km/s]^T，主要描述目标的位置和速度，先验概率密度函数设服从的高斯分，其中x_0|0＝[1km,0.15km/s,6km,0.26km/s]^T，P_0|0＝[0.15²000；00.0100；000.15²0；0000.01]。Where _Nf represents the total number of fuzzy rules,

State vector, x _k represents the target x-axis coordinate, y _k represents the target y-axis coordinate,

and

Represent the speed of the target on the x-axis and y-axis respectively. The process noise e _k is assumed to be a Gaussian process noise Q with zero mean and covariance σ _i,e , where Q is a 2×2 matrix (Q _ij ＝0, for i≠j, Q＝diag(σ _i,e ,σ _i,e )). The observation noise v _k is assumed to be a non-Gaussian distribution noise.

The initial state _x0 is determined by the initial position of the target _x0 = [1km, 0.15km/s, 6km, 0.26km/s] ^T , which mainly describes the position and speed of the target. The prior probability density function is assumed to obey the Gaussian distribution, where x0 _|0 = [1km, 0.15km/s, 6km, 0.26km/s] ^T , P0 _|0 = [0.15 ² 000; 00.0100; 000.15 ² 0; 0000.01].

是状态转移矩阵，

是测量矩阵，它们的表示方法如下：In the embodiment of the invention, the number of particles is set to 200. In order to compare all filtering effects, this embodiment performs 100 Monte Carlo operations, and the position of the sensor is at the origin of the coordinate system.

is the state transition matrix,

are measurement matrices, and they are represented as follows:

In order to verify the effectiveness of the method provided by the embodiment of the invention, a particle filter algorithm (T2IF-TSKPF) of TSK fuzzy modeling for type-2 intuitive fuzzy decision making is implemented. As shown in part (a) of Figure 5, the estimated trajectory of the target motion is shown. It can be seen from the figure that the tracking effect of the ITS-PF algorithm is roughly consistent with the simulation estimate, especially in the case of target maneuvers, it shows good robustness, indicating that the algorithm can efficiently process uncertain information in nonlinear systems. The main reason is that the model set selected when the target maneuvers cannot effectively match the state of the target motion, while the T2IF-TSKPF algorithm can adaptively adjust the weight corresponding to each rule according to the fuzzy membership of the spatial information, thereby obtaining a suitable state estimate of the target maneuvering moment. As shown in part (b) of Figure 5 to part (d) of Figure 5, the target position root mean square error, the root mean square error in the x-axis direction and the root mean square error in the y-axis direction of the interactive particle filter algorithm (IMMPF), the interactive Rao-Blackwellized particle filter algorithm (IMMRBPF), the TSK particle filter algorithm (ITS-PF) and the particle filter algorithm for TSK fuzzy modeling of type-2 intuitive fuzzy decision (T2IF-TSKPF) are respectively described. It can be seen that the particle filter algorithm for TSK fuzzy modeling of type-2 intuitive fuzzy decision (T2IF-TSKPF) provided in the present application has the smallest mean square error and better filtering effect.

Example 2

The embodiment of the present invention provides a TSK fuzzy model particle filter system for type-2 intuitive fuzzy decision making, as shown in FIG6 , including:

The optimal feature set acquisition module 1 is used to use the type-2 intuitive fuzzy decision objective function to select the optimal feature set that can reflect the characteristics of the moving target; this module executes the method described in step S1 of embodiment 1, which will not be repeated here.

TSK fuzzy model construction module 2 is used to construct a TSK fuzzy model, taking the optimal feature set as input, calculating the weight of each rule based on the antecedent membership function of the TSK fuzzy model, and using the weighted summation method to obtain the state fusion result of multiple fuzzy rules as the output of the TSK fuzzy model; this module executes the method described in step S2 in Example 1, which will not be repeated here.

The particle filter module 3 is used to construct an important density function of the particle filter based on the state fusion result, and extract particles from the important density function for filtering update. This module executes the method described in step S3 of embodiment 1, which will not be repeated here.

The TSK fuzzy model particle filter system of type-2 intuitive fuzzy decision-making provided by the embodiment of the present invention uses the type-2 intuitive fuzzy decision objective function to select the optimal feature set that can reflect the characteristics of the moving target; constructs a TSK fuzzy model, takes the optimal feature set as input, calculates the weight of each rule based on the antecedent membership function of the TSK fuzzy model, and uses the weighted summation method to obtain the state fusion result of multiple fuzzy rules as the output of the TSK fuzzy model; constructs an important density function of the particle filter based on the state fusion result, and performs particle filtering according to the important density function. The present invention uses the type-2 intuitive fuzzy decision objective function to select the optimal feature set that can reflect the characteristics of the moving target to construct the TSK fuzzy model, which can not only reduce the number of rules and the redundancy of the model, but also improve the accuracy of the model. The output result of the constructed TSK fuzzy model is used as the important density function of the particle filtering algorithm, which greatly reduces the degradation problem of particles.

Example 3

An embodiment of the present invention provides a terminal, as shown in FIG7, including: at least one processor 401, such as a CPU (Central Processing Unit), at least one communication interface 403, a memory 404, and at least one communication bus 402. The communication bus 402 is used to realize the connection and communication between these components. The communication interface 403 may include a display screen (Display) and a keyboard (Keyboard), and the optional communication interface 403 may also include a standard wired interface and a wireless interface. The memory 404 may be a high-speed RAM memory (Ramdom Access Memory, volatile random access memory) or a non-volatile memory (non-volatile memory), such as at least one disk memory. The memory 404 may also be at least one storage device located away from the aforementioned processor 401. The processor 401 may execute the TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision in Example 1. A set of program codes are stored in the memory 404, and the processor 401 calls the program code stored in the memory 404 to execute the TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision in Example 1. The communication bus 402 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 402 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, FIG7 only uses one line to represent, but does not mean that there is only one bus or one type of bus.

Among them, the memory 404 may include a volatile memory (English: volatile memory), such as a random access memory (English: random-access memory, abbreviated: RAM); the memory may also include a non-volatile memory (English: non-volatile memory), such as a flash memory (English: flash memory), a hard disk drive (English: hard disk drive, abbreviated: HDD) or a solid-state drive (English: solid-state drive, abbreviated: SSD); the memory 404 may also include a combination of the above types of memory.

The processor 401 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and a NP.

The processor 401 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

Optionally, the memory 404 is also used to store program instructions. The processor 401 can call the program instructions to implement the TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision-making in Example 1 of the present application.

The embodiment of the present invention also provides a computer-readable storage medium, on which computer-executable instructions are stored, and the computer-executable instructions can execute the TSK fuzzy model particle filtering method of type-2 intuitive fuzzy decision in Example 1. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated: HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory.

Obviously, the above embodiments are merely examples for clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.

Claims (8)

1. A TSK fuzzy model particle filtering method for type-2 intuitive fuzzy decision is characterized by comprising the following steps:

selecting an optimal feature set capable of representing the characteristics of a moving object by using a type-2 intuitional fuzzy decision objective function, wherein the method comprises the following steps of: and calculating the similarity between features on the same dimension by using the characteristic differences of the speed v, the innovation r and the heading angle difference theta as independent variables and adopting a falling-ridge-shaped distribution, wherein the distribution is represented by the following formula:

wherein x represents the size of the feature difference, a ₁ Representing the minimum value of the characteristic difference, a ₂ Representing the maximum value of the feature difference;

the similarity of the ith feature is calculated by formula (2):

wherein ,

characteristic difference matrix->

df _ji ＝|f _ji -f _(j-1)i I, j=1, 2, Λ, K, i=1, 2, Λ, Q, j represents time;

constructing an objective function based on type-2 intuitionistic fuzzy decision according to the similarity, and selecting an optimal feature set capable of reflecting the characteristics of a moving object;

constructing a TSK fuzzy model, taking the optimal feature set as input, calculating the weight of each rule based on a front part membership function of the TSK fuzzy model, and obtaining a state fusion result of a plurality of fuzzy rules by using a weighted summation method as output of the TSK fuzzy model;

constructing an important density function of particle filtering based on the state fusion result, and extracting particles from the important density function for filtering updating;

the state equation and the measurement equation of the moving object obtained through the steps are as follows:

wherein ,N_f The total number of fuzzy rules is represented,

state vector, x _k Representing the x-axis coordinates, y of the object _k Representing the y-axis coordinates of the object,/->

and

Respectively representing the corresponding speeds of the target in the x-axis and y-axis coordinates, and the process noise e _k Assuming zero mean and covariance obeying sigma _i,e Wherein Q is a 2×2 matrix (Q _ij ＝0,for i≠j,Q＝diag(σ _i,e ,σ _i,e ) A) is provided; observation noise v _k Assuming compliance with non-gaussian distributed noise

Initial state x ₀ Determination of x from initial position of target ₀ ＝[1km,0.15km/s,6km,0.26km/s] ^T Describing the position and velocity of the target, the prior probability density function sets a gaussian score to follow, where x _0|0 ＝[1km,0.15km/s,6km,0.26km/s] ^T ，P _0|0 ＝[0.15 ² 0 0 0；0 0.010 0；00 0.15 ² 0；0 0 0 0.01]；

Is a state transition matrix, ">

Is a measurement matrix, and their representation is as follows: />

2. The TSK fuzzy model particle filtering method of type-2 intuitive fuzzy decision of claim 1, wherein said constructing an objective function based on type-2 intuitive fuzzy decision based on said similarity, selecting an optimal feature set capable of embodying the characteristics of a moving object, comprises:

fusing the similarity of the features;

and selecting the characteristics with the intuitionistic fuzzy decision score larger than a preset threshold value based on the objective function of the type-2 intuitionistic fuzzy decision to form the optimal characteristic set.

3. The TSK fuzzy model particle filtering method of type-2 intuitive fuzzy decision of claim 2, wherein a threshold is preset

Wherein Q+1 is the number of all features in the initial feature set.

4. The method for TSK fuzzy model particle filtering of type-2 intuitive fuzzy decision of claim 1, characterized in that,

in the construction process of the TSK fuzzy model, a fuzzy C regression clustering algorithm based on related entropy and space-time information is used for identifying the parameters of the front part, and the parameters of the back part are estimated by using a strong tracking algorithm.

5. The TSK fuzzy model particle filtering method of type-2 intuitive fuzzy decision of claim 1, wherein said step of particle filtering according to said important density function comprises:

sampling from the important density function to obtain a particle state set;

calculating the weight of the particles in the particle state set, and carrying out standardization processing on the weight;

based on the normalized weights and the set of particle states, states and covariances of particles are calculated.

6. A type-2 intuitive fuzzy decision TSK fuzzy model particle filter system, comprising:

the optimal feature set obtaining module is used for selecting an optimal feature set capable of reflecting the characteristics of a moving object by using a type-2 intuitional fuzzy decision objective function, and comprises the following steps: and calculating the similarity between features on the same dimension by using the characteristic differences of the speed v, the innovation r and the heading angle difference theta as independent variables and adopting a falling-ridge-shaped distribution, wherein the distribution is represented by the following formula:

wherein x represents the size of the feature difference, a ₁ Representing the minimum value of the characteristic difference, a ₂ Representing the maximum value of the feature difference;

the similarity of the ith feature is calculated by formula (2):

wherein ,

characteristic difference matrix->

df _ji ＝|f _ji -f _(j-1)i I, j=1, 2, Λ, K, i=1, 2, Λ, Q, j represents time;

constructing an objective function based on type-2 intuitionistic fuzzy decision according to the similarity, and selecting an optimal feature set capable of reflecting the characteristics of a moving object;

the TSK fuzzy model construction module is used for constructing a TSK fuzzy model, taking the optimal feature set as input, calculating the weight of each rule based on a front piece membership function of the TSK fuzzy model, and obtaining a state fusion result of a plurality of fuzzy rules by using a weighted summation method as output of the TSK fuzzy model;

the particle filtering module is used for constructing an important density function of particle filtering based on the state fusion result, and extracting particles from the important density function for filtering updating;

the state equation and measurement equation acquisition module is used for obtaining the state equation and measurement equation of the moving object through the above modules, and the state equation and measurement equation are as follows:

wherein ,N_f The total number of fuzzy rules is represented,

state vector, x _k Representing the x-axis coordinates, y of the object _k Representing the y-axis coordinates of the object,/->

and

Respectively representing the corresponding speeds of the target in the x-axis and y-axis coordinates, and the process noise e _k Assuming zero mean and covariance obeying sigma _i,e Wherein Q is a 2×2 matrix (Q _ij ＝0,for i≠j,Q＝diag(σ _i,e ,σ _i,e ) A) is provided; observation noise v _k Assuming compliance with non-gaussian distributed noise

Initial state x ₀ Determination of x from initial position of target ₀ ＝[1km,0.15km/s,6km,0.26km/s] ^T Describing the position and velocity of the target, the prior probability density function sets a gaussian score to follow, where x _0|0 ＝[1km,0.15km/s,6km,0.26km/s] ^T ，P _0|0 ＝[0.15 ² 0 0 0；0 0.010 0；00 0.15 ² 0；0 0 0 0.01]；

Is a state transition matrix, ">

Is a measurement matrix, and their representation is as follows:

7. a terminal, comprising: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to perform the type-2 intuitive fuzzy model particle filtering method of any of claims 1-5.

8. A computer readable storage medium storing computer instructions for causing the computer to perform the type-2 intuitive fuzzy model particle filtering method of any one of claims 1-5.

Images