Online parameter distinguishing and backtracking method and system for automatic driving system and medium

Abstract

The invention relates to an automatic driving system online parameter distinguishing and backtracking method and system and a medium, and the method comprises the steps: M1, building a running state model of an automatic driving system of a vehicle when the vehicle runs on a road, obtaining and storing the motion state data information of the vehicle and the motion track data information of the vehicle in the operation process of the automatic driving system in real time; and M2, on the basis of the running state model of the automatic driving system of the vehicle, distinguishing the online parameters of the automatic driving system by adopting an improved maximum correlation minimum redundancy feature selection algorithm to obtain data information of the distinguished online parameters of the automatic driving system. According to the method, automatic driving parameters can be distinguished and backtracked, technicians can conveniently check problems, the backtracking process is simple and rapid, the accuracy is high, and all links of the automatic driving system can be accurately reflected.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to an online parameter identification and backtracking method, system and medium for an autonomous driving system. Background Art

[0002] As autonomous driving technology becomes increasingly mature, autonomous driving systems often require a large number of parameter configurations. The numerous software configurations make software development and debugging very complicated. Once a software problem occurs, engineers need to spend a lot of time to troubleshoot the problem.

[0003] The commonly used debugging method at present is to record the program input and output data through ROS (Robot Operating System), and perform debugging and analysis through the data playback tool provided by the ROS system. The main problem with this method is that the vehicle-side parameters are often inconsistent with the software standard parameters after calibration and debugging, resulting in the inability to reproduce the vehicle-side problems when analyzing the problems. Therefore, a parameter backtracking solution is needed to restore the vehicle-side parameters to the engineer's debugging environment to assist engineers in reproducing and analyzing problems. Summary of the invention

[0004] In view of the above problems, the present invention provides an online parameter identification and backtracing method, system and medium for an autonomous driving system, which can not only identify and backtrace the parameters of the autonomous driving, making it convenient for technical personnel to troubleshoot problems, but also the backtracing process is simple and fast with high accuracy, and can accurately reflect each link of the autonomous driving system.

[0005] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows: an online parameter identification and backtracking method for an autonomous driving system, the method comprising:

[0006] M1. When the vehicle is driving on the road, an operation state model of the vehicle's automatic driving system is constructed, and the vehicle's motion state data information and vehicle's operation trajectory data information during the operation of the automatic driving system are acquired and stored in real time;

[0007] M2. Based on the operating state model of the automatic driving system of the vehicle, an improved maximum relevance minimum redundancy feature selection algorithm is used to identify the online parameters of the automatic driving system, and data information of the identified online parameters of the automatic driving system is obtained;

[0008] M3. Based on the data information of the online parameters of the identified automatic driving system, the running trajectory of the vehicle is predicted by using the BP neural network regression prediction algorithm based on the gray wolf optimization to obtain the predicted running trajectory data information of the vehicle;

[0009] M4. Based on the predicted vehicle running trajectory data information and the vehicle running trajectory data information, a vehicle trajectory difference algorithm based on Gaussian mixture estimation is used to calculate the deviation value of the vehicle's trajectory to obtain the deviation value data information of the vehicle's running trajectory.

[0010] Furthermore, the method further comprises:

[0011] M5. Based on the deviation value data information of the vehicle's running trajectory, a preset threshold is set. If the deviation value data of the vehicle's running trajectory is less than the preset threshold, the data information of the online parameters of the identified automatic driving system is output; if the deviation value data of the vehicle's running trajectory is greater than the preset threshold, return to step M2.

[0012] Furthermore, in step M2, the use of the improved maximum relevance minimum redundancy feature selection algorithm to identify online parameters of the autonomous driving system includes:

[0013] M21. Based on the operating state model of the automatic driving system of the vehicle, obtaining the running trajectory data information and motion state data information of the vehicle during the operation of the automatic driving system of the vehicle;

[0014] M22. Based on the running trajectory data information and motion state data information of the vehicle during the operation of the automatic driving system of the vehicle, establish a characteristic function Q of the automatic driving system of the vehicle,

[0015]

[0016] Among them, x1 is the running trajectory data information of the vehicle during the operation of the vehicle's automatic driving system, x2 is the motion state data information of the vehicle during the operation of the vehicle's automatic driving system, α1, α2 and α3 are feature extraction factors of the vehicle, and the running feature points of the vehicle's automatic driving system are extracted to obtain the running feature point data information of the vehicle's automatic driving system;

[0017] M23. Based on the operation feature point data information of the vehicle's automatic driving system, establish the mutual information value function W of the vehicle's automatic driving system,

[0018]

[0019] Wherein, y is the operation feature point data information of the vehicle's automatic driving system, β1, β2 and β3 are weight coefficients of the vehicle's automatic driving system, and the mutual information value of each operation feature point of the vehicle's automatic driving system is calculated to obtain the mutual information value data information of the operation feature point of the vehicle's automatic driving system;

[0020] M24. Based on the mutual information value data information of the characteristic points of the operation of the automatic driving system of the vehicle, a parameter discrimination function R of the automatic driving system of the vehicle is established,

[0021]

[0022] Among them, z is the mutual information value data information of the characteristic points of the operation of the vehicle's automatic driving system, δ1, δ2 and δ3 are the identification determination factors of the vehicle's automatic driving system, and the online parameters of the automatic driving system are identified to obtain the data information of the online parameters of the identified automatic driving system.

[0023] Furthermore, the identification determining factors δ1, δ2 and δ3 of the vehicle's automatic driving system are,

[0024]

[0025] Among them, z is the mutual information value data information of the feature points of the operation of the vehicle's automatic driving system.

[0026] Furthermore, the weight coefficients β1, β2 and β3 of the vehicle's automatic driving system are,

[0027]

[0028] Among them, y is the operating characteristic point of the vehicle's automatic driving system.

[0029] Furthermore, in step M3, the prediction of the running trajectory of the vehicle using the BP neural network regression prediction algorithm based on gray wolf optimization includes:

[0030] M31. Inputting the data information of the online parameters of the identified autonomous driving system into the BP neural network for training and learning, initializing the weights and biases of the network, and obtaining the weights and bias data information of the initialized network;

[0031] M32. Based on the weight and bias data information of the initialized network, the gray wolf population is initialized, the population parameters are determined, and the initialized gray wolf population data information is obtained;

[0032] M33. Based on the initialized gray wolf population data information, establish the fitness function U of the population individuals,

[0033]

[0034] Among them, a is the initialized gray wolf population data information, γ1 and γ2 are the fitness determining factors of the gray wolf population individuals, and the fitness values ​​of the gray wolf population individuals are calculated to obtain the fitness value data information of the gray wolf population individuals;

[0035] M34. Based on the fitness value data information of the gray wolf population individuals, establish a target optimization function P,

[0036]

[0037] Among them, b is the fitness value data information of the gray wolf population individuals, η1, η2 and η3 are optimization factors, the weights and biases of the initialized network are optimized to obtain the optimized BP neural network, and the data information of the online parameters of the identified automatic driving system is input to predict the vehicle's running trajectory to obtain the predicted vehicle's running trajectory data information.

[0038] Furthermore, the regression prediction neuron activation function S of the optimized BP neural network is:

[0039]

[0040] Among them, c is the data information of the online parameters of the autonomous driving system after identification, and λ1, λ2 and λ3 are the online parameter regression factors of the autonomous driving system.

[0041] Furthermore, in step M4, the step of calculating the deviation value of the vehicle trajectory using the vehicle trajectory difference algorithm based on Gaussian mixture estimation includes:

[0042] M41. Based on the predicted vehicle running trajectory data information, a Gaussian mixture estimation function F1 of the first running trajectory of the vehicle is established,

[0043]

[0044] Among them, r1 is the predicted running trajectory data information of the vehicle, μ1, μ2 and μ3 are the difference factors of Gaussian estimation, and the Gaussian estimation value of the running trajectory of the vehicle is characterized to obtain the Gaussian mixture estimation data information of the first running trajectory of the vehicle;

[0045] M42. Based on the running trajectory data information of the vehicle, a Gaussian mixture estimation function F2 of the second running trajectory of the vehicle is established,

[0046]

[0047] Characterizing the Gaussian estimation value of the running trajectory of the vehicle to obtain Gaussian mixture estimation data information of the second motion trajectory of the vehicle;

[0048] M43. Establishing an estimated deviation value function G of the vehicle based on the Gaussian mixture estimation data information of the second motion trajectory of the vehicle and the Gaussian mixture estimation data information of the first motion trajectory of the vehicle,

[0049]

[0050] Among them, h1 is the Gaussian mixture estimation data information of the vehicle's first motion trajectory, h2 is the Gaussian mixture estimation data information of the vehicle's second motion trajectory, ρ1, ρ2 and ρ3 are the vehicle's adaptive deviation factors, and the deviation value of the vehicle's trajectory is calculated to obtain the deviation value data information of the vehicle's running trajectory.

[0051] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides an online parameter identification and backtracking system for an autonomous driving system, including a computer device, which is programmed or configured to execute the steps of any one of the online parameter identification and backtracking methods for the autonomous driving system.

[0052] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the online parameter identification and backtracking methods for the autonomous driving system.

[0053] The present invention has the following positive effects:

[0054] 1. The present invention constructs an operating state model of the vehicle's automatic driving system and uses an improved maximum relevance minimum redundancy feature selection algorithm to identify the online parameters of the automatic driving system. It can not only identify and trace the parameters of the automatic driving, making it convenient for technical personnel to troubleshoot problems, but also the traceback process is simple and fast with high accuracy, and can accurately reflect each link of the automatic driving system.

[0055] 2. The present invention predicts the vehicle's running trajectory by adopting a BP neural network regression prediction algorithm based on gray wolf optimization, and calculates the deviation value of the vehicle's trajectory by combining the vehicle's trajectory difference algorithm based on Gaussian mixture estimation. It can not only accurately analyze the automatic driving parameters according to the vehicle's real-time data to ensure the accuracy of the automatic driving parameters and facilitate the work of engineers, but also the whole process is quick and convenient, reducing labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0057] Figure 2 A schematic flow chart of an improved maximum relevance minimum redundancy feature selection algorithm of the present invention;

[0058] Figure 3 It is a schematic diagram of the process of the BP neural network regression prediction algorithm based on gray wolf optimization of the present invention;

[0059] Figure 4The figure is a flow chart of the vehicle trajectory difference algorithm based on Gaussian mixture estimation of the present invention. DETAILED DESCRIPTION

[0060] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0061] Example 1: Figure 1 As shown, an online parameter identification and backtracking method for an automatic driving system, the method comprising:

[0062] M1. When the vehicle is driving on the road, an operation state model of the vehicle's automatic driving system is constructed, and the vehicle's motion state data information and vehicle's operation trajectory data information during the operation of the automatic driving system are acquired and stored in real time;

[0063] M2. Based on the operating state model of the automatic driving system of the vehicle, an improved maximum relevance minimum redundancy feature selection algorithm is used to identify the online parameters of the automatic driving system, and data information of the identified online parameters of the automatic driving system is obtained;

[0064] M3. Based on the data information of the online parameters of the identified automatic driving system, the running trajectory of the vehicle is predicted by using the BP neural network regression prediction algorithm based on the gray wolf optimization to obtain the predicted running trajectory data information of the vehicle;

[0065] M4. Based on the predicted vehicle running trajectory data information and the vehicle running trajectory data information, a vehicle trajectory difference algorithm based on Gaussian mixture estimation is used to calculate the deviation value of the vehicle's trajectory to obtain the deviation value data information of the vehicle's running trajectory.

[0066] In this embodiment, the method further includes:

[0067] M5. Based on the deviation value data information of the vehicle's running trajectory, a preset threshold is set. If the deviation value data of the vehicle's running trajectory is less than the preset threshold, the data information of the online parameters of the identified automatic driving system is output; if the deviation value data of the vehicle's running trajectory is greater than the preset threshold, return to step M2.

[0068] In this embodiment, if Figure 2 As shown, in step M2, the use of the improved maximum relevance minimum redundancy feature selection algorithm to identify the online parameters of the automatic driving system includes:

[0069] M21. Based on the operating state model of the automatic driving system of the vehicle, obtaining the running trajectory data information and motion state data information of the vehicle during the operation of the automatic driving system of the vehicle;

[0070] M22. Based on the running trajectory data information and motion state data information of the vehicle during the operation of the automatic driving system of the vehicle, establish a characteristic function Q of the automatic driving system of the vehicle,

[0071]

[0072] Among them, x1 is the running trajectory data information of the vehicle during the operation of the vehicle's automatic driving system, x2 is the motion state data information of the vehicle during the operation of the vehicle's automatic driving system, α1, α2 and α3 are feature extraction factors of the vehicle, and the running feature points of the vehicle's automatic driving system are extracted to obtain the running feature point data information of the vehicle's automatic driving system;

[0073] M23. Based on the operation feature point data information of the vehicle's automatic driving system, establish the mutual information value function W of the vehicle's automatic driving system,

[0074]

[0075] Wherein, y is the operation feature point data information of the vehicle's automatic driving system, β1, β2 and β3 are weight coefficients of the vehicle's automatic driving system, and the mutual information value of each operation feature point of the vehicle's automatic driving system is calculated to obtain the mutual information value data information of the operation feature point of the vehicle's automatic driving system;

[0076] M24. Based on the mutual information value data information of the characteristic points of the operation of the automatic driving system of the vehicle, a parameter discrimination function R of the automatic driving system of the vehicle is established,

[0077]

[0078] Among them, z is the mutual information value data information of the characteristic points of the operation of the vehicle's automatic driving system, δ1, δ2 and δ3 are the identification determination factors of the vehicle's automatic driving system, and the online parameters of the automatic driving system are identified to obtain the data information of the online parameters of the identified automatic driving system.

[0079] In this embodiment, the identification determining factors δ1, δ2 and δ3 of the vehicle's automatic driving system are,

[0080]

[0081] Among them, z is the mutual information value data information of the feature points of the operation of the vehicle's automatic driving system.

[0082] In this embodiment, the weight coefficients β1, β2 and β3 of the vehicle's automatic driving system are,

[0083]

[0084] Among them, y is the operating characteristic point of the vehicle's automatic driving system.

[0085] Example 2: Based on the online parameter identification and backtracking method of an automatic driving system in Example 1, the present invention is further illustrated and described below.

[0086] like Figure 1 As shown, an online parameter identification and backtracking method for an automatic driving system, the method comprising:

[0087] M1. When the vehicle is driving on the road, an operation state model of the vehicle's automatic driving system is constructed, and the vehicle's motion state data information and vehicle's operation trajectory data information during the operation of the automatic driving system are acquired and stored in real time;

[0088] M2. Based on the operating state model of the automatic driving system of the vehicle, an improved maximum relevance minimum redundancy feature selection algorithm is used to identify the online parameters of the automatic driving system, and data information of the identified online parameters of the automatic driving system is obtained;

[0089] M3. Based on the data information of the online parameters of the identified automatic driving system, the running trajectory of the vehicle is predicted by using the BP neural network regression prediction algorithm based on the gray wolf optimization to obtain the predicted running trajectory data information of the vehicle;

[0090] M4. Based on the predicted vehicle running trajectory data information and the vehicle running trajectory data information, a vehicle trajectory difference algorithm based on Gaussian mixture estimation is used to calculate the deviation value of the vehicle's trajectory to obtain the deviation value data information of the vehicle's running trajectory.

[0091] In this embodiment, if Figure 3 As shown, in step M3, the use of the BP neural network regression prediction algorithm based on gray wolf optimization to predict the running trajectory of the vehicle includes:

[0092] M31. Inputting the data information of the online parameters of the identified autonomous driving system into the BP neural network for training and learning, initializing the weights and biases of the network, and obtaining the weights and bias data information of the initialized network;

[0093] M32. Based on the weight and bias data information of the initialized network, the gray wolf population is initialized, the population parameters are determined, and the initialized gray wolf population data information is obtained;

[0094] M33. Based on the initialized gray wolf population data information, establish the fitness function U of the population individuals,

[0095]

[0096] Among them, a is the initialized gray wolf population data information, γ1 and γ2 are the fitness determining factors of the gray wolf population individuals, and the fitness values ​​of the gray wolf population individuals are calculated to obtain the fitness value data information of the gray wolf population individuals;

[0097] M34. Based on the fitness value data information of the gray wolf population individuals, establish a target optimization function P,

[0098]

[0099] Among them, b is the fitness value data information of the gray wolf population individuals, η1, η2 and η3 are optimization factors, the weights and biases of the initialized network are optimized to obtain the optimized BP neural network, and the data information of the online parameters of the identified automatic driving system is input to predict the vehicle's running trajectory to obtain the predicted vehicle's running trajectory data information.

[0100] In this embodiment, the regression prediction neuron activation function S of the optimized BP neural network is:

[0101]

[0102] Among them, c is the data information of the online parameters of the autonomous driving system after identification, and λ1, λ2 and λ3 are the online parameter regression factors of the autonomous driving system.

[0103] In this embodiment, if Figure 4 As shown, in step M4, the use of the vehicle trajectory difference algorithm based on Gaussian mixture estimation to calculate the deviation value of the vehicle trajectory includes:

[0104] M41. Based on the predicted vehicle running trajectory data information, a Gaussian mixture estimation function F1 of the first running trajectory of the vehicle is established,

[0105]

[0106] Among them, r1 is the predicted running trajectory data information of the vehicle, μ1, μ2 and μ3 are the difference factors of Gaussian estimation, and the Gaussian estimation value of the running trajectory of the vehicle is characterized to obtain the Gaussian mixture estimation data information of the first running trajectory of the vehicle;

[0107] M42. Based on the running trajectory data information of the vehicle, a Gaussian mixture estimation function F2 of the second running trajectory of the vehicle is established,

[0108]

[0109] Characterizing the Gaussian estimation value of the running trajectory of the vehicle to obtain Gaussian mixture estimation data information of the second motion trajectory of the vehicle;

[0110] M43. Establishing an estimated deviation value function G of the vehicle based on the Gaussian mixture estimation data information of the second motion trajectory of the vehicle and the Gaussian mixture estimation data information of the first motion trajectory of the vehicle,

[0111]

[0112] Among them, h1 is the Gaussian mixture estimation data information of the vehicle's first motion trajectory, h2 is the Gaussian mixture estimation data information of the vehicle's second motion trajectory, ρ1, ρ2 and ρ3 are the vehicle's adaptive deviation factors, and the deviation value of the vehicle's trajectory is calculated to obtain the deviation value data information of the vehicle's running trajectory.

[0113] In this embodiment, the present invention provides an online parameter identification and backtracking system for an autonomous driving system, comprising a computer device programmed or configured to execute the steps of any one of the online parameter identification and backtracking methods for an autonomous driving system.

[0114] In this embodiment, the present invention provides a computer-readable storage medium, on which is stored a computer program programmed or configured to execute any one of the online parameter identification and backtracking methods for the autonomous driving system.

[0115] Any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0116] In summary, the present invention can not only identify and trace back the parameters of autonomous driving, making it convenient for technicians to troubleshoot problems, but also the traceback process is simple and fast with high accuracy, and can accurately reflect each link of the autonomous driving system.

[0117] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. An online parameter identification and backtracking method for an automatic driving system, characterized in that: The method comprises: M1. When a vehicle is driving on a road, an operation state model of the vehicle's automatic driving system is constructed, and the vehicle's motion state data information and vehicle's operation trajectory data information during the operation of the automatic driving system are acquired and stored in real time; M2. Based on the operating state model of the automatic driving system of the vehicle, an improved maximum relevance minimum redundancy feature selection algorithm is used to identify the online parameters of the automatic driving system, and data information of the identified online parameters of the automatic driving system is obtained; M3. Based on the data information of the online parameters of the identified automatic driving system, the running trajectory of the vehicle is predicted by using the BP neural network regression prediction algorithm based on the gray wolf optimization to obtain the predicted running trajectory data information of the vehicle; M4. Based on the predicted vehicle running trajectory data information and the vehicle running trajectory data information, a vehicle trajectory difference algorithm based on Gaussian mixture estimation is used to calculate the deviation value of the vehicle's trajectory to obtain the deviation value data information of the vehicle's running trajectory.

2. The method for online parameter identification and backtracking of an automatic driving system according to claim 1, characterized in that: The method further comprises: M5. Based on the deviation value data information of the vehicle's running trajectory, a preset threshold is set. If the deviation value data of the vehicle's running trajectory is less than the preset threshold, the data information of the online parameters of the identified automatic driving system is output; if the deviation value data of the vehicle's running trajectory is greater than the preset threshold, return to step M2.

3. The method for online parameter identification and backtracking of an automatic driving system according to claim 1, characterized in that: In step M2, the use of the improved maximum relevance minimum redundancy feature selection algorithm to identify online parameters of the automatic driving system includes: M21. Based on the operating state model of the automatic driving system of the vehicle, obtaining the running trajectory data information and motion state data information of the vehicle during the operation of the automatic driving system of the vehicle; M22. Based on the running trajectory data information and motion state data information of the vehicle during the operation of the automatic driving system of the vehicle, establish a characteristic function Q of the automatic driving system of the vehicle, Among them, x1 is the running trajectory data information of the vehicle during the operation of the vehicle's automatic driving system, x2 is the motion state data information of the vehicle during the operation of the vehicle's automatic driving system, α1, α2 and α3 are feature extraction factors of the vehicle, and the running feature points of the vehicle's automatic driving system are extracted to obtain the running feature point data information of the vehicle's automatic driving system; M23. Based on the operation feature point data information of the vehicle's automatic driving system, establish the mutual information value function W of the vehicle's automatic driving system, Wherein, y is the operation feature point data information of the vehicle's automatic driving system, β1, β2 and β3 are weight coefficients of the vehicle's automatic driving system, and the mutual information value of each operation feature point of the vehicle's automatic driving system is calculated to obtain the mutual information value data information of the operation feature point of the vehicle's automatic driving system; M24. Based on the mutual information value data information of the characteristic points of the operation of the automatic driving system of the vehicle, a parameter discrimination function R of the automatic driving system of the vehicle is established, Among them, z is the mutual information value data information of the characteristic points of the operation of the vehicle's automatic driving system, δ1, δ2 and δ3 are the identification determination factors of the vehicle's automatic driving system, and the online parameters of the automatic driving system are identified to obtain the data information of the online parameters of the identified automatic driving system.

4. The method for online parameter identification and backtracking of an automatic driving system according to claim 3, characterized in that: The identification determining factors δ1, δ2 and δ3 of the vehicle's automatic driving system are, Among them, z is the mutual information value data information of the feature points of the operation of the vehicle's automatic driving system.

5. The method for online parameter identification and backtracking of an automatic driving system according to claim 3, characterized in that: The weight coefficients β1, β2 and β3 of the vehicle's automatic driving system are, Among them, y is the operating characteristic point of the vehicle's automatic driving system.

6. The online parameter identification and backtracking method of the automatic driving system according to claim 1, characterized in that: In step M3, the method of using the BP neural network regression prediction algorithm based on gray wolf optimization to predict the running trajectory of the vehicle includes: M31. Inputting the data information of the online parameters of the identified autonomous driving system into the BP neural network for training and learning, initializing the weights and biases of the network, and obtaining the weights and bias data information of the initialized network; M32. Based on the weight and bias data information of the initialized network, the gray wolf population is initialized, the population parameters are determined, and the initialized gray wolf population data information is obtained; M33. Based on the initialized gray wolf population data information, establish the fitness function U of the population individuals, Among them, a is the initialized gray wolf population data information, γ1 and γ2 are the fitness determining factors of the gray wolf population individuals, and the fitness values ​​of the gray wolf population individuals are calculated to obtain the fitness value data information of the gray wolf population individuals; M34. Based on the fitness value data information of the gray wolf population individuals, establish a target optimization function P, Among them, b is the fitness value data information of the gray wolf population individuals, η1, η2 and η3 are optimization factors, the weights and biases of the initialized network are optimized to obtain the optimized BP neural network, and the data information of the online parameters of the identified automatic driving system is input to predict the vehicle's running trajectory to obtain the predicted vehicle's running trajectory data information.

7. The method for online parameter identification and backtracking of an automatic driving system according to claim 6, characterized in that: The regression prediction neuron activation function S of the optimized BP neural network is: Among them, c is the data information of the online parameters of the autonomous driving system after identification, and λ1, λ2 and λ3 are the online parameter regression factors of the autonomous driving system.

8. The method for online parameter identification and backtracking of an automatic driving system according to claim 1, characterized in that: In step M4, the step of calculating the deviation value of the vehicle's trajectory using the vehicle trajectory difference algorithm based on Gaussian mixture estimation includes: M41. Based on the predicted vehicle running trajectory data information, a Gaussian mixture estimation function F1 of the first running trajectory of the vehicle is established, Among them, r1 is the predicted running trajectory data information of the vehicle, μ1, μ2 and μ3 are the difference factors of Gaussian estimation, and the Gaussian estimation value of the running trajectory of the vehicle is characterized to obtain the Gaussian mixture estimation data information of the first running trajectory of the vehicle; M42. Based on the running trajectory data information of the vehicle, a Gaussian mixture estimation function F2 of the second running trajectory of the vehicle is established, Characterizing the Gaussian estimation value of the running trajectory of the vehicle to obtain Gaussian mixture estimation data information of the second motion trajectory of the vehicle; M43. Establishing an estimated deviation value function G of the vehicle based on the Gaussian mixture estimation data information of the second motion trajectory of the vehicle and the Gaussian mixture estimation data information of the first motion trajectory of the vehicle, Among them, h1 is the Gaussian mixture estimation data information of the vehicle's first motion trajectory, h2 is the Gaussian mixture estimation data information of the vehicle's second motion trajectory, ρ1, ρ2 and ρ3 are the vehicle's adaptive deviation factors, and the deviation value of the vehicle's trajectory is calculated to obtain the deviation value data information of the vehicle's running trajectory.

9. An online parameter identification and backtracking system for an automatic driving system, comprising a computer device, characterized in that: The computer device is programmed or configured to execute the steps of the online parameter identification and backtracking method of the autonomous driving system as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that is programmed or configured to execute the online parameter identification and backtracking method for the autonomous driving system as described in any one of claims 1 to 8.

Images