Unmanned container truck fault diagnosis method and system based on deep learning, and medium

Abstract

The invention relates to an unmanned container truck fault diagnosis method and system based on deep learning and a medium, and the method comprises the steps: U1, enabling a vehicle to run on a road, and obtaining the motion state data information of the vehicle in real time based on a plurality of groups of vehicle-mounted sensors; u2, on the basis of the motion state data information of the vehicle, a random forest regression prediction algorithm based on sine factors and dynamic learning is adopted to predict operation parameters of the vehicle, and predicted operation parameter data information of the vehicle is obtained; and U3, on the basis of the predicted running parameter data information of the vehicle, constructing a fault diagnosis model of deep learning based on multi-source data fusion, and diagnosing fault points of the vehicle to obtain diagnosis data information of the fault points of the vehicle. According to the invention, the vehicle fault problem of the unmanned container truck can be accurately and rapidly diagnosed, and an operator can rapidly solve some common problems, so that the maintenance and operation efficiency of the unmanned container truck is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned container trucks, and in particular to a deep learning-based fault diagnosis method, system and medium for unmanned container trucks. Background Art

[0002] In the port unmanned container truck project, fault diagnosis technology is indispensable. When the unmanned container truck is in operation, the fault diagnosis system monitors the vehicle's operating status in real time and diagnoses the hardware (sensors, chassis, etc.) and software (algorithm modules) of the self-driving vehicle. The current diagnostic architecture can cover all aspects of the entire system in a relatively comprehensive manner, and can diagnose and output the current fault events of the system in real time and respond. However, since the diagnostic logic is located at the bottom layer of each software module, the fault events (fault codes) output based on data or signal diagnosis are relatively professional, and when a problem occurs in a certain link of the system, it often causes the modules on the entire link to diagnose fault events, and many fault codes are reported. During the operation process, if a container truck fails, the operator may not be able to quickly locate the problem through these relatively professional fault codes. It is necessary to contact professionals to analyze and solve the problem before resuming vehicle operation, which reduces the efficiency of operation and maintenance. Summary of the invention

[0003] In view of the above problems, the present invention provides an unmanned container truck fault diagnosis method, system and medium based on deep learning, which can not only accurately and quickly diagnose vehicle fault problems of unmanned container trucks, but also enable operators to quickly solve some common problems, thereby improving the maintenance and operation efficiency of unmanned container trucks.

[0004] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows: a method for diagnosing faults of unmanned container trucks based on deep learning, the method comprising:

[0005] U1. The vehicle is driving on the road, and the vehicle's motion status data information is obtained in real time based on multiple sets of on-board sensors;

[0006] U2. Based on the motion state data information of the vehicle, the operating parameters of the vehicle are predicted using a random forest regression prediction algorithm based on sinusoidal factors and dynamic learning to obtain the predicted operating parameter data information of the vehicle;

[0007] U3. Based on the predicted operating parameter data information of the vehicle, a fault diagnosis model based on deep learning of multi-source data fusion is constructed to diagnose the fault point of the vehicle and obtain the diagnostic data information of the fault point of the vehicle;

[0008] U4. Based on the diagnostic data information of the fault point of the vehicle, the fault point diagnosis of the vehicle is optimized by using a multi-objective sparrow optimization algorithm based on a refractive reverse learning mechanism and an adaptive control factor to obtain the optimized fault point diagnostic data information of the vehicle.

[0009] Furthermore, in step U2, the use of a random forest regression prediction algorithm based on sinusoidal factors and dynamic learning to predict the operating parameters of the vehicle includes:

[0010] U21. Based on the motion state data information of the vehicle, construct a feature extraction function Q of the motion state of the vehicle,

[0011]

[0012] Among them, x is the vehicle's motion state data information, α1, α2 and α3 are the feature extraction factors of the vehicle's motion state, the feature points of the vehicle's motion state are extracted, and the feature matrix of the vehicle's motion state is constructed to obtain the feature matrix data information of the vehicle's motion state;

[0013] U22. Based on the characteristic matrix data information of the vehicle motion state, a decision tree function W of the vehicle motion state is established,

[0014]

[0015] Among them, y is the characteristic matrix data information of the vehicle motion state, β1, β2 and β3 are the sine factors of the decision tree, and the characteristic matrix of the vehicle motion state is divided into a decision tree to obtain the decision tree data information of the vehicle motion state;

[0016] U23. Based on the decision tree data information of the vehicle motion state, establish the random forest regression prediction function R of the vehicle,

[0017]

[0018] Among them, z is the decision tree data information of the vehicle's motion state, δ1, δ2 and δ3 are dynamic learning factors of the vehicle's motion state, and the vehicle's operating parameters are predicted to obtain the predicted vehicle's operating parameter data information.

[0019] Furthermore, the dynamic learning factors δ1, δ2 and δ3 of the vehicle motion state are:

[0020] Among them, z is the decision tree data information of the vehicle's motion state.

[0021] Furthermore, the sinusoidal factors β1, β2 and β3 of the decision tree are:

[0022]

[0023]

[0024] Among them, y is the characteristic matrix data information of the vehicle's motion state.

[0025] Furthermore, in step U3, the construction of a fault diagnosis model based on deep learning of multi-source data fusion to diagnose the fault point of the vehicle includes:

[0026] U31. Based on the predicted vehicle operating parameter data information, normalization is performed and a vehicle operating parameter data set is constructed;

[0027] U32. Input the vehicle's operating parameter data set into a fault diagnosis model based on deep learning of multi-source data fusion for training and learning, and determine the vehicle's fault point diagnosis function P,

[0028]

[0029] Among them, r is the vehicle operation parameter data set, γ1, γ2 and γ3 are the weight coefficients of the vehicle operation parameters, and the trained fault diagnosis model based on deep learning of multi-source data fusion is obtained;

[0030] U33. Based on the trained fault diagnosis model based on deep learning of multi-source data fusion, the predicted vehicle operating parameter data information is input, the vehicle fault point is diagnosed, and the diagnostic data information of the vehicle fault point is obtained.

[0031] Furthermore, the constraints of the weight coefficients γ1, γ2 and γ3 of the vehicle operating parameters are:

[0032]

[0033] Furthermore, in step U4, the optimization of vehicle fault point diagnosis using a multi-objective sparrow optimization algorithm based on a refraction reverse learning mechanism and an adaptive control factor includes:

[0034] U41. Based on the diagnostic data information of the fault point of the vehicle, the sparrow population is initialized, the population parameters and the maximum number of iterations L are determined, and the data information of the initialized sparrow population is obtained;

[0035] U42. Based on the data information of the initialized sparrow population, a fitness function S of the refraction reverse learning mechanism of the individual sparrow population is established.

[0036]

[0037] Among them, g is the data information of the initialized sparrow population, η1, η2 and η3 are the refraction reverse learning mechanism factors of the sparrow population, and the fitness of the individuals in the sparrow population is calculated to obtain the data information of the fitness of the individuals in the sparrow population;

[0038] U43. Based on the data information of the fitness of the individuals of the sparrow population, establish the optimization function G of the multi-objective sparrow population,

[0039]

[0040] Among them, h is the data information of the fitness of individuals in the sparrow population, λ1, λ2 and λ3 are the adaptive control factors of individuals in the sparrow population, and the fault point diagnosis of the vehicle is optimized to obtain the optimized fault point diagnosis data information of the vehicle.

[0041] Furthermore, the constraint function f of the adaptive control factors λ1, λ2 and λ3 of the individuals in the sparrow population is,

[0042]

[0043] The constraints of the refraction reverse learning mechanism factors η1, η2 and η3 of the sparrow population are:

[0044] The value range of the constraint function f is (0,1).

[0045] In order to achieve the above-mentioned objectives and other related objectives, the present invention also provides an unmanned container truck fault diagnosis system based on deep learning, including a computer device, which is programmed or configured to execute any one of the steps of the unmanned container truck fault diagnosis method based on deep learning.

[0046] In order to achieve the above-mentioned objectives and other related objectives, the present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the unmanned container truck fault diagnosis methods based on deep learning.

[0047] The present invention has the following positive effects:

[0048] 1. The present invention predicts the operating parameters of the vehicle by adopting a random forest regression prediction algorithm based on sinusoidal factors and dynamic learning, and combines it with a fault diagnosis model based on deep learning of multi-source data fusion to diagnose the fault points of the vehicle. It can not only accurately and quickly diagnose the vehicle fault problems of unmanned container trucks, but also enable operators to quickly solve some common problems, thereby improving the maintenance and operation efficiency of unmanned container trucks.

[0049] 2. The present invention optimizes the vehicle's fault point diagnosis by adopting a multi-objective sparrow optimization algorithm based on a refractive reverse learning mechanism and an adaptive control factor, thereby obtaining optimized vehicle fault point diagnosis data information. This can not only further optimize the vehicle's fault point diagnosis data, thereby improving the user's efficiency in eliminating vehicle fault points, but also further improve the operating efficiency of unmanned container trucks. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0051] Figure 2 A schematic diagram of the flow of the random forest regression prediction algorithm based on sinusoidal factors and dynamic learning of the present invention;

[0052] Figure 3 A schematic diagram of a process for constructing a fault diagnosis model based on deep learning of multi-source data fusion according to the present invention;

[0053] Figure 4 It is a flow chart of the multi-objective sparrow optimization algorithm based on the refractive reverse learning mechanism and the adaptive control factor of the present invention. DETAILED DESCRIPTION

[0054] 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.

[0055] Example 1: Figure 1 As shown, a method for fault diagnosis of unmanned container trucks based on deep learning, the method comprising:

[0056] U1. The vehicle is driving on the road, and the vehicle's motion status data information is obtained in real time based on multiple sets of on-board sensors;

[0057] U2. Based on the motion state data information of the vehicle, the operating parameters of the vehicle are predicted using a random forest regression prediction algorithm based on sinusoidal factors and dynamic learning to obtain the predicted operating parameter data information of the vehicle;

[0058] U3. Based on the predicted operating parameter data information of the vehicle, a fault diagnosis model based on deep learning of multi-source data fusion is constructed to diagnose the fault point of the vehicle and obtain the diagnostic data information of the fault point of the vehicle;

[0059] U4. Based on the diagnostic data information of the fault point of the vehicle, the fault point diagnosis of the vehicle is optimized by using a multi-objective sparrow optimization algorithm based on a refractive reverse learning mechanism and an adaptive control factor to obtain the optimized fault point diagnostic data information of the vehicle.

[0060] In this embodiment, if Figure 2 As shown, in step U2, the use of the random forest regression prediction algorithm based on sinusoidal factors and dynamic learning to predict the operating parameters of the vehicle includes:

[0061] U21. Based on the motion state data information of the vehicle, construct a feature extraction function Q of the motion state of the vehicle,

[0062]

[0063] Among them, x is the vehicle's motion state data information, α1, α2 and α3 are the feature extraction factors of the vehicle's motion state, the feature points of the vehicle's motion state are extracted, and the feature matrix of the vehicle's motion state is constructed to obtain the feature matrix data information of the vehicle's motion state;

[0064] U22. Based on the characteristic matrix data information of the vehicle motion state, a decision tree function W of the vehicle motion state is established,

[0065]

[0066] Among them, y is the characteristic matrix data information of the vehicle motion state, β1, β2 and β3 are the sine factors of the decision tree, and the characteristic matrix of the vehicle motion state is divided into a decision tree to obtain the decision tree data information of the vehicle motion state;

[0067] U23. Based on the decision tree data information of the vehicle motion state, establish the random forest regression prediction function R of the vehicle,

[0068]

[0069] Among them, z is the decision tree data information of the vehicle's motion state, δ1, δ2 and δ3 are dynamic learning factors of the vehicle's motion state, and the vehicle's operating parameters are predicted to obtain the predicted vehicle's operating parameter data information.

[0070] In this embodiment, the dynamic learning factors δ1, δ2 and δ3 of the vehicle motion state are,

[0071]

[0072] Among them, z is the decision tree data information of the vehicle's motion state.

[0073] In this embodiment, the sinusoidal factors β1, β2 and β3 of the decision tree are,

[0074]

[0075] Among them, y is the characteristic matrix data information of the vehicle's motion state.

[0076] In this embodiment, if Figure 3 As shown, in step U3, the construction of a fault diagnosis model based on deep learning of multi-source data fusion to diagnose the fault point of the vehicle includes:

[0077] U31. Based on the predicted vehicle operating parameter data information, normalization is performed and a vehicle operating parameter data set is constructed;

[0078] U32. Input the vehicle's operating parameter data set into a fault diagnosis model based on deep learning of multi-source data fusion for training and learning, and determine the vehicle's fault point diagnosis function P,

[0079]

[0080] Among them, r is the vehicle operation parameter data set, γ1, γ2 and γ3 are the weight coefficients of the vehicle operation parameters, and the trained fault diagnosis model based on deep learning of multi-source data fusion is obtained;

[0081] U33. Based on the trained fault diagnosis model based on deep learning of multi-source data fusion, the predicted vehicle operating parameter data information is input, the vehicle fault point is diagnosed, and the diagnostic data information of the vehicle fault point is obtained.

[0082] In this embodiment, the constraints of the weight coefficients γ1, γ2 and γ3 of the vehicle operating parameters are:

[0083]

[0084] Example 2: Based on the unmanned container truck fault diagnosis method based on deep learning in Example 1, the present invention is further illustrated and described below.

[0085] like Figure 1 As shown, a method for fault diagnosis of unmanned container trucks based on deep learning, the method comprising:

[0086] U1. The vehicle is driving on the road, and the vehicle's motion status data information is obtained in real time based on multiple sets of on-board sensors;

[0087] U2. Based on the motion state data information of the vehicle, the operating parameters of the vehicle are predicted using a random forest regression prediction algorithm based on sinusoidal factors and dynamic learning to obtain the predicted operating parameter data information of the vehicle;

[0088] U3. Based on the predicted operating parameter data information of the vehicle, a fault diagnosis model based on deep learning of multi-source data fusion is constructed to diagnose the fault point of the vehicle and obtain the diagnostic data information of the fault point of the vehicle;

[0089] U4. Based on the diagnostic data information of the fault point of the vehicle, the fault point diagnosis of the vehicle is optimized by using a multi-objective sparrow optimization algorithm based on a refractive reverse learning mechanism and an adaptive control factor to obtain the optimized fault point diagnostic data information of the vehicle.

[0090] In this embodiment, if Figure 4 As shown, in step U4, the optimization of vehicle fault point diagnosis using a multi-objective sparrow optimization algorithm based on a refraction reverse learning mechanism and an adaptive control factor includes:

[0091] U41. Based on the diagnostic data information of the fault point of the vehicle, the sparrow population is initialized, the population parameters and the maximum number of iterations L are determined, and the data information of the initialized sparrow population is obtained;

[0092] U42. Based on the data information of the initialized sparrow population, a fitness function S of the refraction reverse learning mechanism of the individual sparrow population is established.

[0093]

[0094] Among them, g is the data information of the initialized sparrow population, η1, η2 and η3 are the refraction reverse learning mechanism factors of the sparrow population, and the fitness of the individuals in the sparrow population is calculated to obtain the data information of the fitness of the individuals in the sparrow population;

[0095] U43. Based on the data information of the fitness of the individuals of the sparrow population, establish the optimization function G of the multi-objective sparrow population,

[0096]

[0097] Among them, h is the data information of the fitness of individuals in the sparrow population, λ1, λ2 and λ3 are the adaptive control factors of individuals in the sparrow population, and the fault point diagnosis of the vehicle is optimized to obtain the optimized fault point diagnosis data information of the vehicle.

[0098] In this embodiment, the constraint function f of the adaptive control factors λ1, λ2 and λ3 of the individuals in the sparrow population is,

[0099]

[0100] The constraints of the refraction reverse learning mechanism factors η1, η2 and η3 of the sparrow population are:

[0101] The value range of the constraint function f is (0,1).

[0102] In this embodiment, the present invention provides an unmanned container truck fault diagnosis system based on deep learning, including a computer device, which is programmed or configured to execute the steps of any one of the unmanned container truck fault diagnosis methods based on deep learning.

[0103] 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 unmanned container truck fault diagnosis methods based on deep learning.

[0104] 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).

[0105] In summary, the present invention can not only accurately and quickly diagnose vehicle fault problems of unmanned container trucks, but also enable operators to quickly solve some common problems, thereby improving the maintenance and operation efficiency of unmanned container trucks.

[0106] 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. A method for fault diagnosis of unmanned container trucks based on deep learning, characterized in that: The method comprises: U1. The vehicle is driving on the road, and the vehicle's motion status data information is obtained in real time based on multiple sets of on-board sensors; U2. Based on the motion state data information of the vehicle, the operating parameters of the vehicle are predicted using a random forest regression prediction algorithm based on sinusoidal factors and dynamic learning to obtain the predicted operating parameter data information of the vehicle; U3. Based on the predicted operating parameter data information of the vehicle, a fault diagnosis model based on deep learning of multi-source data fusion is constructed to diagnose the fault point of the vehicle and obtain the diagnostic data information of the fault point of the vehicle; U4. Based on the diagnostic data information of the fault point of the vehicle, the fault point diagnosis of the vehicle is optimized by using a multi-objective sparrow optimization algorithm based on a refractive reverse learning mechanism and an adaptive control factor to obtain the optimized fault point diagnostic data information of the vehicle.

2. The unmanned container truck fault diagnosis method based on deep learning according to claim 1 is characterized in that: In step U2, the use of the random forest regression prediction algorithm based on sinusoidal factors and dynamic learning to predict the operating parameters of the vehicle includes: U21. Based on the motion state data information of the vehicle, construct a feature extraction function Q of the motion state of the vehicle, Among them, x is the vehicle's motion state data information, α1, α2 and α3 are the feature extraction factors of the vehicle's motion state, the feature points of the vehicle's motion state are extracted, and the feature matrix of the vehicle's motion state is constructed to obtain the feature matrix data information of the vehicle's motion state; U22. Based on the characteristic matrix data information of the vehicle motion state, a decision tree function W of the vehicle motion state is established, Among them, y is the characteristic matrix data information of the vehicle motion state, β1, β2 and β3 are the sine factors of the decision tree, and the characteristic matrix of the vehicle motion state is divided into a decision tree to obtain the decision tree data information of the vehicle motion state; U23. Based on the decision tree data information of the vehicle motion state, establish the random forest regression prediction function R of the vehicle, Among them, z is the decision tree data information of the vehicle's motion state, δ1, δ2 and δ3 are dynamic learning factors of the vehicle's motion state, and the vehicle's operating parameters are predicted to obtain the predicted vehicle's operating parameter data information.

3. The unmanned container truck fault diagnosis method based on deep learning according to claim 2 is characterized in that: The dynamic learning factors δ1, δ2 and δ3 of the vehicle motion state are, Among them, z is the decision tree data information of the vehicle's motion state.

4. The unmanned container truck fault diagnosis method based on deep learning according to claim 2 is characterized in that: The sinusoidal factors β1, β2 and β3 of the decision tree are, Among them, y is the characteristic matrix data information of the vehicle's motion state.

5. The unmanned container truck fault diagnosis method based on deep learning according to claim 1 is characterized in that: In step U3, the construction of a fault diagnosis model based on deep learning of multi-source data fusion to diagnose the fault point of the vehicle includes: U31. Based on the predicted vehicle operating parameter data information, normalization is performed and a vehicle operating parameter data set is constructed; U32. Input the vehicle's operating parameter data set into a fault diagnosis model based on deep learning of multi-source data fusion for training and learning, and determine the vehicle's fault point diagnosis function P, Among them, r is the vehicle operation parameter data set, γ1, γ2 and γ3 are the weight coefficients of the vehicle operation parameters, and the trained fault diagnosis model based on deep learning of multi-source data fusion is obtained; U33. Based on the trained fault diagnosis model based on deep learning of multi-source data fusion, the predicted vehicle operating parameter data information is input, the vehicle fault point is diagnosed, and the diagnostic data information of the vehicle fault point is obtained.

6. The unmanned container truck fault diagnosis method based on deep learning according to claim 5 is characterized in that: The constraints of the weight coefficients γ1, γ2 and γ3 of the vehicle operating parameters are:

7. The unmanned container truck fault diagnosis method based on deep learning according to claim 1 is characterized in that: In step U4, the optimization of vehicle fault point diagnosis using a multi-objective sparrow optimization algorithm based on a refraction reverse learning mechanism and an adaptive control factor includes: U41. Based on the diagnostic data information of the fault point of the vehicle, the sparrow population is initialized, the population parameters and the maximum number of iterations L are determined, and the data information of the initialized sparrow population is obtained; U42. Based on the data information of the initialized sparrow population, a fitness function S of the refraction reverse learning mechanism of the individual sparrow population is established. Among them, g is the data information of the initialized sparrow population, η1, η2 and η3 are the refraction reverse learning mechanism factors of the sparrow population, and the fitness of the individuals in the sparrow population is calculated to obtain the data information of the fitness of the individuals in the sparrow population; U43. Based on the data information of the fitness of the individuals of the sparrow population, establish the optimization function G of the multi-objective sparrow population, Among them, h is the data information of the fitness of individuals in the sparrow population, λ1, λ2 and λ3 are the adaptive control factors of individuals in the sparrow population, and the fault point diagnosis of the vehicle is optimized to obtain the optimized fault point diagnosis data information of the vehicle.

8. The unmanned container truck fault diagnosis method based on deep learning according to claim 7 is characterized in that: The constraint function f of the adaptive control factors λ1, λ2 and λ3 of the individual sparrow population is, The constraints of the refraction reverse learning mechanism factors η1, η2 and η3 of the sparrow population are: The value range of the constraint function f is (0,1).

9. An unmanned container truck fault diagnosis system based on deep learning, including a computer device, characterized in that: The computer device is programmed or configured to execute the steps of the unmanned container truck fault diagnosis method based on deep learning 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 unmanned container truck fault diagnosis method based on deep learning as described in any one of claims 1 to 8.

Images