An AMT gear shifting strategy selection method, system, device and medium based on working condition prediction

Abstract

The application provides an AMT gear shifting strategy selection method, system, device and medium based on working condition prediction, and belongs to the technical field of vehicle gear shifting control. The method steps are as follows: determining signal characteristic parameters and collecting vehicle driving data to construct a sample signal matrix; performing data standardization on the sample signal matrix; using a principal component analysis method to perform parameter dimension reduction on the standardized sample signal matrix to obtain a dimension-reduced sample signal matrix; performing cluster analysis on the dimension-reduced sample signal matrix to output driving working condition classification results, setting AMT gear shifting strategies for each driving working condition classification result; obtaining real-time vehicle driving data of an actual working condition, updating the working condition classification result, predicting the working condition, matching the predicted working condition with the driving working condition classification result, and selecting a corresponding AMT gear shifting strategy for gear shifting control. The application shortens the gear shifting time, reduces the vehicle impact degree to ensure the smoothness of the vehicle, and improves the gear shifting quality.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of vehicle gear shifting control, and particularly relates to an AMT gear shifting strategy selection method, system, device and medium based on working condition prediction. BACKGROUND

[0002] An AMT gearbox is an automated manual gearbox. Compared with a traditional automatic gearbox, the AMT gearbox has lower cost, can provide higher driving experience and comfort, and has lower maintenance cost. The AMT gearbox is the basis for the development of automatic gear vehicles, and in particular, the AMT gearbox with an added automatic electronic control unit has more competitiveness in terms of manufacturing process and cost. In addition, the AMT gearbox can more accurately control the gear shifting time and gear shifting speed, so the AMT gearbox has higher fuel economy than the traditional automatic gearbox without sacrificing power performance.

[0003] As an upgrade of the traditional manual gearbox, the AMT gearbox, although realizing automatic separation and combination of the clutch and automatic switching of the gear through an electronic control system, to a certain extent, simplifies the operation of the driver and improves the convenience of driving. However, first, the traditional AMT gear shifting strategy is mostly based on the current vehicle state for decision-making, lacks the ability to predict future working condition changes, and results in inflexible gear shifting time, affecting the driving experience. Second, during the gear shifting process using the traditional gear shifting strategy, the control accuracy of the clutch and the coordination of engine torque management affect the gear shifting smoothness. If not properly controlled, it will cause large gear shifting impact, affect the comfort of passengers, and even cause unnecessary wear and tear of vehicle parts. Finally, for complex and variable working conditions, the traditional gear shifting strategy often cannot quickly respond to environmental changes, resulting in too long gear shifting time, affecting the acceleration performance and fuel economy of the vehicle.

[0004] Therefore, it is necessary to provide an AMT gear shifting strategy selection method, system, device and medium based on working condition prediction to solve the above technical problems. SUMMARY

[0005] In view of the above-mentioned defects of the prior art that the traditional AMT gear shifting strategy lacks working condition prediction ability, resulting in inflexible gear shifting time, affecting the driving experience, affecting the gear shifting smoothness, affecting the comfort of passengers, and long gear shifting time affecting the acceleration performance and fuel economy of the vehicle, the present application provides an AMT gear shifting strategy selection method, system, device and medium based on working condition prediction to solve the above technical problems.

[0006] In a first aspect, the present application provides an AMT gear shifting strategy selection method based on working condition prediction, comprising the following steps:

[0007] S1. Determine the CAN signal feature parameters and collect the corresponding vehicle driving data to construct a sample signal matrix;

[0008] S2. Data standardization is performed on the sample signal matrix to obtain a standardized sample signal matrix;

[0009] S3. Principal component analysis is used to analyze the standardized sample signal matrix, realizing dimension reduction of the CAN signal feature parameters, and obtaining a dimension-reduced sample signal matrix based on the dimension-reduced CAN signal feature parameters;

[0010] S4. Cluster analysis is performed on the dimension-reduced sample signal matrix, and the driving condition classification result is output, and an AMT shift strategy is set for each driving condition classification result;

[0011] S5. The real-time vehicle driving data corresponding to the dimension-reduced CAN signal feature parameters of the actual working condition are obtained, the working condition classification result is updated, the working condition is predicted, the predicted working condition is matched with the driving condition classification result, and the corresponding AMT shift strategy is selected for shift control.

[0012] Further, the specific steps of step S1 are as follows:

[0013] S11. Determine the CAN signal feature parameters of the original dimension p;

[0014] S12. Collect vehicle driving data corresponding to the CAN signal feature parameters of the original dimension in a set number n;

[0015] S13. Construct a p╳n sample signal matrix based on the collected vehicle driving data, wherein the elements of the sample signal matrix are x ij , 1≤i≤p, 1≤j≤n.

[0016] Further, the specific steps of step S2 are as follows:

[0017] S21. Find the minimum element min(x i ) in the row vector of the sample signal matrix, calculate the deviation of each element xi in the row vector from the minimum element min(x i ) as the row deviation;

[0018] S22. Find the maximum element max(x i ) and the minimum element min(x i ) in the column vector x j of the sample signal matrix, and calculate the deviation of the two as the column deviation;

[0019] S23. Calculate the ratio of the row deviation and the column deviation for each element in the sample signal matrix to obtain the normalized element y:

[0020] ;

[0021] S24. Take the matrix formed by the normalized elements y as the normalized sample signal matrix.

[0022] Further, the step S3 is specifically as follows:

[0023] S31. Calculate the covariance matrix of the normalized sample signal matrix , wherein

[0024] ;

[0025] S32. Calculate the eigenvalues of the covariance matrix and the corresponding orthogonal eigenvectors ;

[0026] S33. Select the first eigenvalues of the covariance matrix as the variances corresponding to the first principal components according to the size of the eigenvalues ;

[0027] S34. Set the corresponding unit eigenvectors as the coefficients of the principal components with respect to the original variables X, then the th principal component of the original variable is: ;

[0028] S35. Calculate the variance contribution rate of the principal component by the following formula, which represents the size of the signal amount of each CAN signal characteristic parameter: ;

[0029] S36. Calculate the cumulative variance contribution rate by the following formula:

[0030] ;

[0031] S37. Determine whether the cumulative contribution rate is greater than a set threshold value;

[0032] If yes, go to step S38;

[0033] If no, adjust the value of m and return to step S33;

[0034] S38. Determine the value of m through the cumulative variance contribution rate, so that the selected principal component is , realize the dimension reduction of the CAN signal characteristic parameters, and obtain the dimension-reduced sample signal matrix Z based on the dimension-reduced CAN signal characteristic parameters.​​​​​

[0035] Further, the step S4 comprises the following steps:

[0036] S41. Taking the row vectors in the reduced sample signal matrix Z as the working condition signals;

[0037] S42. Optionally, one working condition signal, calculating the cumulative distance of the selected working condition signal and all working condition signals in the reduced sample signal;

[0038] S43. Calculating the number, mean, variance and eccentricity of the cumulative working condition signals;

[0039] S44. Selecting the K-Means clustering method, taking the number, mean, variance and eccentricity of the working condition signals as input to perform preliminary clustering analysis, obtaining micro-clusters;

[0040] S45. Identifying and merging similar working condition signals to form macro-clusters by creating an adjacency matrix and judging the intersection conditions of micro-clusters, and then processing each micro-cluster in the macro-cluster;

[0041] S46. Analyzing each driving condition category and its CAN signal characteristic parameters to extract the characteristics of each driving condition category;

[0042] S47. Setting the corresponding AMT shift strategy based on the characteristics of each driving condition category.

[0043] Further, the step S45 comprises the following steps:

[0044] S451. Creating an adjacency matrix for each micro-cluster, wherein the dimension of the adjacency matrix is the same as the number of micro-clusters;

[0045] S452. Judging whether two micro-clusters intersect according to the elements in the adjacency matrix;

[0046] If yes, it is determined that the two micro-clusters intersect, and step S453 is entered;

[0047] If no, it is determined that the two micro-clusters do not intersect, and step S454 is entered;

[0048] S453. Setting the two intersecting micro-clusters with similar working condition signals as similar micro-clusters;

[0049] S454. Merging similar micro-clusters to obtain macro-clusters;

[0050] S455. Dividing each macro-cluster into regions, calculating the clustering density of each region, and setting the regions with clustering density lower than the threshold as low-density regions, and setting the regions with clustering density higher than the threshold as high-density regions;

[0051] S456. Discard the micro-clusters located in the low density region, and activate the micro-clusters in the high density region;

[0052] S457. Calculate the weight of each activated micro-cluster in each macro-cluster, and perform weighted normalization processing on each micro-cluster.

[0053] Further, the step S5 specifically comprises the following steps:

[0054] S51. Obtain real-time vehicle driving data corresponding to the p-dimensional CAN signal feature parameters of the actual working condition;

[0055] S52. Construct the sample signal matrix according to step S1, standardize the sample signal matrix according to step S2, reduce the dimension according to step S3, and cluster according to step S4, and then update the driving condition category using the clustering result;

[0056] S53. Compare the reduced dimension CAN signal feature parameters of the real-time working condition with each driving condition category to realize working condition prediction;

[0057] S54. Match the predicted working condition with the driving condition classification result, select the matching driving condition category, and output the corresponding AMT shift strategy for shift control.

[0058] In a second aspect, the present application provides an AMT shift strategy selection system based on working condition prediction, comprising:

[0059] A sample signal matrix construction module is configured to determine CAN signal feature parameters and collect corresponding vehicle driving data to construct a sample signal matrix;

[0060] A matrix standardization module is configured to standardize the data of the sample signal matrix to obtain a standardized sample signal matrix;

[0061] A matrix dimension reduction module is configured to analyze the standardized sample signal matrix using a principal component analysis method to realize dimension reduction of the CAN signal feature parameters, and obtain a reduced dimension sample signal matrix based on the reduced dimension CAN signal feature parameters;

[0062] A driving condition clustering module is configured to perform clustering analysis on the reduced dimension sample signal matrix, output a driving condition classification result, and set an AMT shift strategy for each driving condition classification result;

[0063] A shift strategy selection module is configured to obtain real-time vehicle driving data corresponding to the reduced dimension CAN signal feature parameters of the actual working condition, update the working condition classification result, predict the working condition, match the predicted working condition with the driving condition classification result, and select the corresponding AMT shift strategy for shift control.

[0064] In a third aspect, the present application provides a device comprising a processor and a memory;

[0065] The memory is configured to store a computer program, and the processor is configured to call and run the computer program from the memory, so that the device executes the method of the first aspect.

[0066] In a fourth aspect, the present application provides a storage medium,

[0067] The storage medium stores instructions, which, when executed on a computer, cause the computer to execute the method of the first aspect.

[0068] The present application has the following beneficial effects:

[0069] The AMT shift strategy selection method, system, device and medium based on working condition prediction provided by the present application adjust the selection of vehicle shift strategy and improve shift performance by identifying the main influencing factors in the working condition identification process, taking these factors as input data of the prediction model, analyzing the working condition data trend required for working condition prediction, strengthening the model for vehicle control, shortening the shift time, and reducing the vehicle impact to ensure the smoothness of the vehicle.

[0070] In addition, the design principle of the present application is reliable, the structure is simple, and it has very wide application prospects.

[0071] Therefore, compared with the prior art, the present application has outstanding substantial characteristics and significant progress, and the beneficial effects of its implementation are also obvious. BRIEF DESCRIPTION OF DRAWINGS

[0072] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without creative labor.

[0073] Figure 1 is a flowchart of the AMT shift strategy selection method based on working condition prediction of the present application.

[0074] Figure 2 is a schematic diagram of the AMT shift strategy selection system based on working condition prediction of the present application. DETAILED DESCRIPTION

[0075] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work should fall within the protection scope of the present application.

[0076] The AMT gear shifting strategy selection method based on working condition prediction can be applied to one or more devices, which is a device capable of automatically performing numerical calculation and / or information processing according to a pre-set or stored instruction, and the hardware thereof includes but is not limited to a microprocessor, an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc.

[0077] The device can be any electronic product capable of human-computer interaction with a user, for example, a personal computer, a tablet computer, a smart phone, a personal digital assistant (PDA), an interactive Internet Protocol Television (IPTV), etc. The network in which the device is located includes but is not limited to the Internet, a wide area network, a metropolitan area network, a local area network, a virtual private network (VPN), etc.

[0078] Please refer to Figure 1 The AMT gear shifting strategy selection method based on working condition prediction in the embodiment includes the following steps:

[0079] S1. Determine CAN signal characteristic parameters and collect corresponding vehicle driving data to construct a sample signal matrix;

[0080] S2. Perform data standardization on the sample signal matrix to obtain a standardized sample signal matrix;

[0081] S3. Perform analysis on the standardized sample signal matrix using a principal component analysis method, realize dimension reduction of the CAN signal characteristic parameters, and obtain a dimension-reduced sample signal matrix based on the dimension-reduced CAN signal characteristic parameters;

[0082] S4. Perform clustering analysis on the dimension-reduced sample signal matrix, output a driving working condition classification result, and set an AMT gear shifting strategy for each driving working condition classification result;

[0083] S5. Obtain real-time vehicle driving data corresponding to the reduced dimension CAN signal feature parameters of the actual working condition, update the working condition classification result, predict the working condition, match the predicted working condition with the driving working condition classification result, and select the corresponding AMT shift strategy for shift control.

[0084] In this embodiment, the specific steps of step S1 are as follows:

[0085] S11. Determine the CAN signal feature parameters of the original dimension p; for example, select commonly used CAN signal feature parameters: maximum speed , maximum acceleration , maximum deceleration , average speed , average acceleration , average deceleration , high-speed driving time ratio , medium-speed driving time ratio , low-speed driving time ratio , idle speed time ratio , standard deviation of speed , standard deviation of acceleration ;

[0086] S12. Collect vehicle driving data corresponding to the CAN signal feature parameters of the original dimension in a set number n;

[0087] S13. Construct a sample signal matrix of p╳n based on the collected vehicle driving data, wherein the elements of the sample signal matrix are x ij , 1≤i≤p, 1≤j≤n;

[0088] For example, the constructed sample signal matrix Z is as follows:

[0089]

[0090] The specific steps of step S2 are as follows:

[0091] S21. Find the minimum element min(x i ) in the row vector of the sample signal matrix, calculate the deviation of each element xi in the row vector from the minimum element min(x i ) as the row deviation;

[0092] S22. Find the maximum element max(x j ) and the minimum element min(x i ) in the column vector x i of the sample signal matrix, and calculate the deviation of the two as the column deviation;

[0093] S23. Calculate the ratio of row deviation and column deviation of each element in the sample signal matrix to obtain normalized element y:

[0094] ;

[0095] S24. Form a matrix of the normalized element y as the normalized sample signal matrix;

[0096] It should be noted that the data standardization of the sample signal matrix, i.e. the mean value, is first calculated the maximum value and the minimum value in each working condition signal data instance, and the range is calculated; then, each observation value is subtracted by the minimum value, and then divided by the range, to complete the standardization;

[0097] It should be noted that a large number of related indicators, for example indicators are recombined into fewer comprehensive indicators that are not related to each other to replace the original indicators; these comprehensive indicators not only can reflect the information represented by the original variables to the greatest extent, but also can ensure the independence between new indicators;

[0098] The specific steps of step S3 are as follows:

[0099] S31. Calculate the covariance matrix of the normalized sample signal matrix , wherein

[0100] ;

[0101] S32. Calculate the eigenvalue and the corresponding orthogonal eigenvector of the covariance matrix ;

[0102] S33. According to the size of the eigenvalue , select the first eigenvalues of the covariance matrix as the variances corresponding to the first principal components;

[0103] S34. Set the corresponding unit eigenvector as the coefficient of the principal component about the original variable X, then the principal component of the original variable is: ; ;

[0104] S35. Calculate the variance contribution rate of the principal component by the following formula: The size of the signal quantity of each CAN signal characteristic parameter is represented: ;

[0105] S36. The variance cumulative contribution rate is calculated by the following formula:

[0106] ;

[0107] S37. It is determined whether the cumulative contribution rate is greater than a set threshold value;

[0108] If yes, go to step S38;

[0109] If no, adjust the value of m and return to step S33;

[0110] S38. The value of m is determined by the variance cumulative contribution rate, so that the selected principal component is , the dimensionality reduction of the CAN signal characteristic parameter is realized, and the dimensionality reduction sample signal matrix Z based on the dimensionality reduction CAN signal characteristic parameter is obtained;

[0111] For example, after principal component analysis, the original CAN signal matrix changes to: , and the obtained signals are maximum speed, maximum acceleration, maximum deceleration, average speed, average acceleration, average deceleration, high-speed driving time ratio, low-speed driving time ratio, and idling time ratio;

[0112] The specific steps of step S4 are as follows:

[0113] S41. The row vector in the dimensionality reduction sample signal matrix Z is taken as a working condition signal;

[0114] S42. An optional working condition signal is selected, and the cumulative distance between the selected working condition signal and all working condition signals in the dimensionality reduction sample signal is calculated;

[0115] S43. The number of working condition signal accumulations, the average value, the variance, and the eccentricity are calculated;

[0116] The number of working condition signal accumulations is: ;

[0117] The average value of the signal quantity at the current moment is: ;

[0118] The variance of the signal quantity at the current moment is: ;

[0119] The eccentricity of the signal quantity at the current moment becomes: ;

[0120] S44. The K-Means clustering method is selected, and the number of working condition signals, the average value, the variance, and the eccentricity are taken as inputs for preliminary clustering analysis to obtain micro clusters;

[0121] S45. Identify and merge similar working condition signals to form macro clusters by creating an adjacency matrix and judging the intersection conditions of micro clusters, and then process each micro cluster in the macro cluster; the specific steps of step S45 are as follows:

[0122] S451. Create an adjacency matrix for each micro cluster, wherein the dimension of the adjacency matrix is the same as the number of micro clusters;

[0123] S452. Determine whether two micro clusters intersect according to the elements in the adjacency matrix;

[0124] If yes, it is determined that the two micro clusters intersect, and step S453 is entered;

[0125] If no, it is determined that the two micro clusters do not intersect, and step S454 is entered;

[0126] S453. The two intersecting micro clusters have similar working condition signals, and are set as similar micro clusters;

[0127] S454. Similar micro clusters are merged to obtain macro clusters;

[0128] S455. Divide each macro cluster into regions, calculate the clustering density of each region, and set the region with a clustering density lower than a threshold as a low-density region, and set the region with a clustering density higher than the threshold as a high-density region;

[0129] S456. Discard the micro clusters located in the low-density region, and activate the micro clusters in the high-density region;

[0130] S457. Calculate the weight of each activated micro cluster in each macro cluster, and perform weighted normalization processing on each micro cluster;

[0131] S46. Analyze each driving condition category and its CAN signal feature parameters, and extract the features of each driving condition category;

[0132] S47. Set the corresponding AMT shift strategy based on the features of each driving condition category;

[0133] The specific steps of step S5 are as follows:

[0134] S51. Obtain real-time vehicle driving data corresponding to the p-dimensional CAN signal feature parameters of the actual working condition;

[0135] S52. Construct the sample signal matrix according to step S1, standardize the sample signal matrix according to step S2, reduce the dimension according to step S3, and cluster according to step S4, and then update the driving condition category using the clustering result;

[0136] S53. Compare the reduced dimension CAN signal feature parameters of the real-time working condition with each driving condition category to realize working condition prediction;

[0137] S54. The predicted working condition is matched with the driving condition classification result, a corresponding AMT shift strategy of a matched driving condition category is selected and output, and shift control is performed.

[0138] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution. The execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0139] The following is an embodiment of an AMT shift strategy selection system based on working condition prediction provided by the embodiments of the present disclosure. The system belongs to the same inventive concept as the AMT shift strategy selection based on working condition prediction of each of the above embodiments. Details not described in the embodiment of the AMT shift strategy selection system based on working condition prediction can be referred to the above embodiment of the AMT shift strategy selection method based on working condition prediction.

[0140] As shown in Figure 2 , the system comprises:

[0141] A sample signal matrix construction module for determining CAN signal feature parameters and collecting corresponding vehicle driving data to construct a sample signal matrix;

[0142] A matrix normalization module for data normalization of the sample signal matrix to obtain a normalized sample signal matrix;

[0143] A matrix dimension reduction module for analyzing the normalized sample signal matrix using a principal component analysis method to realize dimension reduction of the CAN signal feature parameters and obtain a dimension-reduced sample signal matrix based on the dimension-reduced CAN signal feature parameters;

[0144] A driving condition clustering module for clustering analysis of the dimension-reduced sample signal matrix to output a driving condition classification result and set an AMT shift strategy for each driving condition classification result;

[0145] A shift strategy selection module for obtaining real-time vehicle driving data corresponding to the dimension-reduced CAN signal feature parameters of the actual working condition, updating the working condition classification result, predicting the working condition, matching the predicted working condition with the driving condition classification result, and selecting a corresponding AMT shift strategy for shift control.

[0146] The AMT shift strategy selection method based on working condition prediction provided by the present application is described in combination with the modules and algorithm steps of each example disclosed in the present application, and can be realized by electronic hardware, computer software or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been described in the above description in general terms. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0147] Those skilled in the art can understand that each aspect of the AMT shift strategy selection method based on working condition prediction provided by the present application can be implemented as a system, a method or a program product. Therefore, each aspect of the present disclosure can be embodied as a whole hardware embodiment, a whole software embodiment (including firmware, microcode, etc.), or an embodiment combining hardware and software aspects, which can be collectively referred to as "circuitry", "module" or "system" herein.

[0148] In a non-transitory computer-readable storage medium storing the AMT shift strategy selection method based on working condition prediction, it can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0149] Although the present application has been described in detail by referring to the preferred embodiments thereof, it is not limited to the preferred embodiments. Various equivalent modifications or replacements can be made to the embodiments of the present application by those skilled in the art without departing from the spirit and essence of the present application, and such modifications or replacements should be within the scope of the present application. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, and such changes or replacements should be within the protection scope of the present application.

Claims

1. A method for AMT shift strategy selection based on operating condition prediction, characterized in that, Comprising the following steps: S1. Determine the CAN signal characteristic parameters and collect the corresponding vehicle driving data to construct a sample signal matrix; S2. Standardize the sample signal matrix to obtain a standardized sample signal matrix; S3. Use principal component analysis method to analyze the standardized sample signal matrix, realize dimension reduction of the CAN signal characteristic parameters, and obtain a dimension-reduced sample signal matrix based on the dimension-reduced CAN signal characteristic parameters; S4. Perform clustering analysis on the dimension-reduced sample signal matrix, and output the driving condition classification result, set the AMT shift strategy for each driving condition classification result; S5. Obtain the real-time vehicle driving data corresponding to the dimension-reduced CAN signal characteristic parameters of the actual condition, update the condition classification result, predict the condition, match the predicted condition with the driving condition classification result, and select the corresponding AMT shift strategy for shift control; Step S4 comprises the following steps: S41. Take the row vector in the dimension-reduced sample signal matrix Z as the condition signal; S42. Select a condition signal, and calculate the cumulative distance between the selected condition signal and all condition signals in the dimension-reduced sample signal matrix; S43. Calculate the number, average, variance, and eccentricity of the cumulative condition signals; S44. Select the K-Means clustering method, and use the number, average, variance, and eccentricity of the condition signals as inputs for preliminary clustering analysis to obtain micro-clusters; S45. Identify and merge similar condition signals to form macro-clusters by creating an adjacency matrix and judging the intersection condition of the micro-clusters, and then process each micro-cluster in the macro-cluster; S46. Analyze each driving condition category and its CAN signal characteristic parameters, and extract the features of each driving condition category; S47. Set the corresponding AMT shift strategy based on the features of each driving condition category; Step S45 comprises the following steps: S451. Create an adjacency matrix for each micro-cluster, wherein the dimension of the adjacency matrix is the same as the number of micro-clusters; S452. Determine whether two micro-clusters intersect according to the elements in the adjacency matrix; If yes, determine that the two micro-clusters intersect, and go to step S453; If no, determine that the two micro-clusters do not intersect, and go to step S454; S453. Set the two intersecting micro-clusters with similar condition signals as similar micro-clusters; S454. Merge the similar micro-clusters to obtain macro-clusters; S455. Divide each macro-cluster into regions, calculate the clustering density of each region, set the region with a clustering density lower than a threshold as a low-density region, and set the region with a clustering density higher than the threshold as a high-density region; S456. Discard the micro-clusters located in the low-density region, and activate the micro-clusters in the high-density region; S457. Calculate the weight of each activated micro-cluster in each macro-cluster, and perform weighted normalization processing on each micro-cluster; Step S5 comprises the following steps: S51. Obtain the real-time vehicle driving data corresponding to the p-dimensional CAN signal characteristic parameters of the actual condition; S52. Construct the sample signal matrix according to step S1, standardize the sample signal matrix according to step S2, perform dimension reduction according to step S3, and perform clustering according to step S4, and then use the clustering result to update the driving condition category; S53. Compare the reduced dimension CAN signal feature parameters of the real-time working condition with each driving working condition category to realize working condition prediction; S54. Match the predicted working condition with the driving working condition classification result, select the AMT shift strategy corresponding to the matched driving working condition category for output, and perform shift control.

2. The AMT shift strategy selection method based on operating condition prediction according to claim 1, characterized in that, Step S1 has the following specific steps: S11. Determine the CAN signal feature parameters of the original dimension p; S12. Collect vehicle driving data corresponding to the CAN signal feature parameters of the original dimension in a set number n; S13. Construct a sample signal matrix of p x n based on the collected vehicle driving data, wherein the elements of the sample signal matrix are x ij , 1≤i≤p, 1≤j≤n.

3. The AMT shift strategy selection method based on operating condition prediction according to claim 2, characterized in that, Step S2 has the following specific steps: S21. Find the minimum element min(x i ) in the row vector of the sample signal matrix, and calculate the deviation of each element xi in the row vector from the minimum element min(x i ) as the row deviation. S22. Find the maximum element max(x j ) and the minimum element min(x i ) from the column vectors x i of the sample signal matrix and compute the difference between both as column bias; S23. Calculate the ratio of row deviation to column deviation for each element in the sample signal matrix to obtain the normalized element y: S24. Form the matrix of the normalized element y as the standardized sample signal matrix.

4. The AMT shift strategy selection method based on operating condition prediction according to claim 3, characterized in that, Step S3 has the following specific steps: S31. Calculate the covariance matrix Σ(s) on the normalized sample signal matrix ij ) p×p where S32. Calculate the eigenvalues λ ij ) p×p of the covariance matrix Σ(s i and the corresponding orthonormal eigenvectors a i ; S33. The feature values λ i Size selection covariance matrix Σ(s ij ) p×p The first m eigenvalues λ1≥ λ2... λ m > 0 as the variance corresponding to the first m principal components S34. Set λ i The corresponding unit eigenvector a i is the coefficient of the principal component F i with respect to the original variable X, then the i-th principal component F i of the original variable is: i F i = a′ S35. The principal component F is calculated by the following formula i The variance contribution rate a of the principal component F i The signal amount of each CAN signal characteristic parameter is represented by the size of the signal amount: S36. Calculate the variance cumulative contribution rate by the following formula: S37. Determine whether the cumulative contribution rate is greater than a set threshold value; If yes, go to step S38; If no, adjust the value of m and return to step S33; S38. Determine the value of m by the variance cumulative contribution rate, so the selected principal components are F1, F2...F m , realize the dimensionality reduction of CAN signal feature parameters, and obtain the dimensionality reduction sample signal matrix Z based on the dimensionality reduction of CAN signal feature parameters.

5. A system for AMT shift strategy selection based on operating condition prediction, characterized in that, The method according to any one of claims 1-4, the system comprises: A sample signal matrix construction module for determining CAN signal feature parameters and collecting corresponding vehicle driving data to construct a sample signal matrix; A matrix standardization module for data standardization of the sample signal matrix to obtain a standardized sample signal matrix; A matrix dimension reduction module for analysis of the standardized sample signal matrix using principal component analysis method to realize dimension reduction of the CAN signal feature parameters and obtain a reduced dimension sample signal matrix based on the reduced dimension CAN signal feature parameters; A driving working condition clustering module for clustering analysis of the reduced dimension sample signal matrix to output a driving working condition classification result, and setting an AMT shift strategy for each driving working condition classification result; A shift strategy selection module for obtaining real-time vehicle driving data corresponding to the reduced dimension CAN signal feature parameters of the actual working condition, updating the working condition classification result, predicting the working condition, matching the predicted working condition with the driving working condition classification result, and selecting the corresponding AMT shift strategy for shift control.

6. An apparatus, comprising: It comprises a processor and a memory; The memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the device executes the method of any one of claims 1-4.

7. A storage medium, characterized in that The storage medium stores instructions, which, when executed on a computer, cause the computer to execute the method of any one of claims 1-4.

Images