Vehicle driving condition identification method, electronic equipment and storage medium

By dividing vehicle driving data, feature structure, dimensionality reduction and clustering processing, rich and accurate vehicle driving working conditions are identified, which solves the problem of simple and inaccurate construction of working conditions in the existing technology, and improves new energy Accuracy of automotive design.

CN118940073APending Publication Date: 2024-11-12ZHEJIANG LEAPPOWER TECH CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410833580.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-25
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In the design of new energy vehicles, the construction of vehicle driving conditions based on vehicle speed is simple and inaccurate, and the real kinematic characteristics of the vehicle cannot be effectively restored, resulting in large errors in the design.

Method used

By obtaining vehicle driving data, dividing kinematic segments, constructing and dimensionality reduction feature sets, performing clustering processing, determining the vehicle driving working condition type, and using feature meanings and clustering results to determine the rich and realistic working condition type.

Benefits of technology

It realizes accurate identification of vehicle driving conditions, matches the richness with actual conditions, and improves the accuracy and reliability of the design.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940073A_ABST
    Figure CN118940073A_ABST
Patent Text Reader

Abstract

The invention discloses a vehicle driving condition identification method. The method comprises the following steps: acquiring driving data of a vehicle; dividing the driving data to obtain a plurality of kinematics fragments; performing feature construction on the plurality of kinematics fragments to obtain a plurality of first kinematics feature sets of the plurality of kinematics fragments; performing dimension reduction processing on the plurality of first kinematic feature sets to obtain a plurality of second kinematic feature sets corresponding to the plurality of first kinematic feature sets and feature meanings corresponding to the second kinematic feature sets; performing clustering processing on the plurality of second kinematic feature sets to obtain at least one clustering result of the plurality of kinematic fragments; and determining a vehicle driving condition type corresponding to the driving data by using the feature meanings of the plurality of second kinematic feature sets and the at least one clustering result. The invention further discloses electronic equipment and a storage medium. The vehicle driving working condition types obtained in the application are rich and accord with the actual working condition.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments disclosed in the present application relate to the field of intelligent automobile technology, and more specifically, to a method for identifying vehicle driving conditions, an electronic device, and a storage medium. Background Art

[0002] Vehicle driving conditions refer to the working state of a vehicle during driving. At present, vehicle driving conditions serve as the basis for vehicle energy consumption / emission test methods and limit standards, reflect the kinematic characteristics of vehicle road driving, and are the main benchmark for calibration and optimization of various performance indicators of vehicles. In the design of new energy vehicles, different operating condition data are usually required as reference inputs. However, the vehicle driving condition categories constructed based only on the time domain characteristics of vehicle speed are few and simple, or the vehicle driving condition categories predicted by fixed parameter models are quite different from the actual conditions, and the real situation of the vehicle cannot be restored in the relevant design. Summary of the invention

[0003] According to the embodiments of the present application, the present application proposes a method for identifying vehicle driving conditions, an electronic device and a storage medium to solve the above-mentioned problems.

[0004] The first aspect of the present application discloses a method for identifying a vehicle driving condition, comprising: acquiring driving data of a vehicle; dividing the driving data to obtain a plurality of kinematic segments; performing feature construction on the plurality of kinematic segments to obtain a plurality of first kinematic feature sets of the plurality of kinematic segments; performing dimensionality reduction processing on the plurality of first kinematic feature sets to obtain a plurality of second kinematic feature sets corresponding to the plurality of first kinematic feature sets and their corresponding feature meanings; performing clustering processing on the plurality of second kinematic feature sets to obtain at least one clustering result of the plurality of kinematic segments; and determining the type of vehicle driving condition corresponding to the driving data using the feature meanings of the plurality of second kinematic feature sets and the at least one clustering result.

[0005] In some embodiments, the multiple first kinematic feature sets are subjected to dimensionality reduction processing, including: performing standardization processing on the multiple first kinematic feature sets to obtain multiple processed first kinematic feature sets; performing principal component analysis on the multiple processed first kinematic feature sets to obtain a first number of eigenvalues ​​and the corresponding first number of eigenvectors; obtaining multiple principal component features and their contribution rates based on the first number of eigenvalues ​​and the corresponding first number of eigenvectors, wherein the i-th principal component feature among the multiple principal component features is represented by the i-th eigenvector corresponding to the multiple first kinematic feature sets and the first number of eigenvectors, and the contribution rate of the i-th principal component is represented by the first number of eigenvalues ​​and the i-th eigenvalue corresponding to the i-th principal component feature; and determining the principal component feature that meets the contribution rate threshold among the multiple principal component features as the target principal component feature.

[0006] In some embodiments, the second kinematic feature set includes the target principal component feature, and the feature meaning of the second kinematic feature set is determined based on a feature vector corresponding to the target principal component feature.

[0007] In some embodiments, clustering processing is performed on the multiple second kinematic feature sets to obtain at least one clustering result of the multiple kinematic segments, including: determining the number of target categories; standardizing the multiple second kinematic feature sets to obtain multiple processed second kinematic feature sets; clustering processing is performed on the multiple processed second kinematic feature sets to obtain clustering results of the number of target categories.

[0008] In some embodiments, determining the target category number includes: obtaining a preset category number; based on the preset category number, determining at least one initial class center corresponding to the preset category number using a hierarchical clustering method; based on the at least one initial class center, clustering the multiple second kinematic feature sets using a K-Means clustering algorithm to obtain classes corresponding to the preset category number; calculating the class average diameter of the class corresponding to the preset category number; and judging whether the preset category number is the target category number based on the class average diameter.

[0009] In some embodiments, the type of vehicle driving condition corresponding to the driving data is determined using the feature meanings of the multiple second kinematic feature sets and the at least one clustering result, including: obtaining a kinematic segment corresponding to any clustering result in the at least one clustering result; calculating feature values ​​corresponding to the multiple second kinematic feature sets in the kinematic segment corresponding to the clustering result; and determining the type of vehicle driving condition represented by the clustering result based on the feature values ​​and the feature meanings of the second kinematic feature sets.

[0010] In some embodiments, the driving data includes multiple data segments, and the multiple data segments are represented by the rear motor speed and the longitudinal acceleration of the vehicle; the driving data is divided to obtain multiple kinematic segments, including: dividing the data segments that meet preset conditions in the multiple data segments into one kinematic segment, and the preset condition is that the rear motor speed in the data segment is converted from the first motor speed to the second motor speed, and then from the second motor speed to the first motor speed; using the longitudinal acceleration of the vehicle to divide each kinematic segment in the multiple kinematic segments, and determining the idle segment data, acceleration segment data, uniform speed segment data and deceleration segment data of each kinematic segment.

[0011] In some embodiments, the first kinematic feature set includes a time feature set, a speed feature set and a torque feature set; wherein the time feature set includes at least one of running distance, running time, idling time, acceleration time, deceleration time and uniform speed time; the speed feature set includes at least one of average rear motor speed, average driving rear motor speed, maximum rear motor speed, maximum acceleration, minimum acceleration, average acceleration in acceleration section, average deceleration in deceleration section, standard deviation of rear motor speed and standard deviation of rear motor acceleration; the torque feature set includes at least one of average torque, average torque during driving, maximum torque, average torque in acceleration section, average torque in deceleration section, average torque in uniform speed section, standard deviation of torque, maximum torque change rate, minimum torque change rate, average torque change rate in acceleration section, average torque change rate in deceleration section and average torque change rate in uniform speed section.

[0012] The second aspect of the present application discloses an electronic device, comprising a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory to implement the method for identifying the vehicle driving condition described in the first aspect.

[0013] The third aspect of the present application discloses a non-volatile computer-readable storage medium having program instructions stored thereon, and when the program instructions are executed by a processor, the method for identifying the vehicle driving condition described in the first aspect is implemented.

[0014] The beneficial effects of the present application include: dividing the vehicle driving data to obtain multiple kinematic segments, performing feature construction on the multiple kinematic segments to obtain multiple first kinematic feature sets of the multiple kinematic segments, performing dimensionality reduction processing on the multiple first kinematic feature sets to obtain multiple second kinematic feature sets corresponding to the multiple first kinematic feature sets and their corresponding feature meanings, further, performing clustering processing on the multiple second kinematic feature sets to obtain at least one clustering result of the multiple kinematic segments, and determining multiple vehicle driving condition types using the feature meanings of the second kinematic feature sets and at least one clustering result, so that the obtained vehicle driving condition types are rich and conform to actual conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present application will be further described below in conjunction with the accompanying drawings and implementation methods, in which:

[0016] Figure 1 is a flow chart of a method for identifying a vehicle driving condition according to an embodiment of the present application;

[0017] Figure 2 is a schematic diagram of a kinematic segment of an embodiment of the present application;

[0018] Figure 3 is a line graph showing changes in the class average diameter according to an embodiment of the present application;

[0019] Figure 4 is a schematic diagram of the structure of an electronic device according to an embodiment of the present application;

[0020] Figure 5 It is a schematic diagram of the structure of the non-volatile computer-readable storage medium of an embodiment of the present application. DETAILED DESCRIPTION

[0021] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0022] The term "and / or" in this application is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the objects associated before and after are in an "or" relationship. In addition, "many" in this article means two or more than two. In addition, the term "at least one" in this article means any combination of at least two of any one or more of a plurality of, for example, including at least one of A, B, and C, and can mean including any one or more elements selected from the set consisting of A, B, and C. In addition, the terms "first", "second", and "third" in this application are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features.

[0023] In order to enable those skilled in the art to better understand the technical solution of the present application, the technical solution of the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0024] See also Figure 1 , Figure 1 The flowchart of the vehicle driving condition identification method of the embodiment of the present application is shown in FIG. The execution subject of the method can be an electronic device with computing function, such as a microcomputer, a server, and a mobile device such as a laptop computer and a tablet computer.

[0025] It should be noted that if there are substantially the same results, the method of the present application is not limited to Figure 1 The process sequence shown is limited.

[0026] In some possible implementations, the method may be implemented by a processor calling computer-readable instructions stored in a memory, such as Figure 1 As shown, the method may include the following steps:

[0027] S11: Acquire vehicle driving data.

[0028] The vehicle's driving data may include vehicle computer signals, rear electrode speed, vehicle longitudinal acceleration, torque, vehicle driving time, mileage, etc.

[0029] S12: Divide the driving data to obtain a plurality of kinematic segments.

[0030] By dividing the vehicle's driving data, multiple kinematic segments are obtained. The multiple kinematic segments can be the vehicle's driving data in multiple continuous time periods or the vehicle's driving data in multiple non-continuous time periods. The driving data corresponding to each kinematic segment includes the vehicle computer signal, rear electrode speed, vehicle longitudinal acceleration, torque, vehicle driving time, mileage, etc. in the corresponding time period.

[0031] S13: Perform feature construction on the multiple kinematic segments to obtain multiple first kinematic feature sets of the multiple kinematic segments.

[0032] Feature construction is performed on multiple kinematic segments to obtain multiple first kinematic feature sets of the multiple kinematic segments. Feature construction is performed on each kinematic segment in the multiple kinematic segments to obtain a first kinematic feature set corresponding to each kinematic segment in the multiple kinematic segments. The first kinematic feature set may include multiple first kinematic features, such as multiple features about time in the kinematic segment, multiple features about motor speed, multiple features about torque, etc.

[0033] S14: Performing dimensionality reduction processing on the multiple first kinematic feature sets to obtain multiple second kinematic feature sets corresponding to the multiple first kinematic feature sets and their corresponding feature meanings.

[0034] A plurality of first kinematic feature sets are subjected to dimensionality reduction processing to obtain a plurality of second kinematic feature sets and their corresponding feature meanings, that is, a plurality of first kinematic feature sets are subjected to dimensionality reduction processing to obtain a second kinematic feature set corresponding to each first kinematic feature set and its corresponding feature meaning, wherein one second kinematic feature set may include a plurality of second kinematic features, that is, feature meanings corresponding to a plurality of second kinematic features may be obtained. The feature meaning corresponding to each second kinematic feature may be used to reflect the changes in different parameters during vehicle driving, and the number of features of the second kinematic feature in the second kinematic feature set is less than the number of features of the first kinematic feature in the first kinematic feature set.

[0035] S15: performing clustering processing on the plurality of second kinematic feature sets to obtain at least one clustering result of the plurality of kinematic segments.

[0036] Clustering is performed on multiple second kinematic feature sets to obtain at least one clustering result of multiple kinematic segments. For example, based on a second kinematic feature set with a smaller number of features, a clustering algorithm is used to cluster similar kinematic segments into one category, thereby obtaining multiple clustering results of multiple kinematic segments.

[0037] S16: Determine the vehicle driving condition type corresponding to the driving data using the feature meanings of the plurality of second kinematic feature sets and at least one clustering result.

[0038] At least one clustering result of multiple kinematic segments is obtained, for example, n clustering results are obtained, and the second kinematic feature set data of the kinematic segments corresponding to the n clustering results are obtained, so as to summarize the n driving condition types corresponding to the n clustering results, that is, the n driving condition types corresponding to the driving data, by using the feature meaning of the second kinematic feature set.

[0039] In this embodiment, the vehicle's driving data is divided to obtain multiple kinematic segments, feature construction is performed on the multiple kinematic segments to obtain multiple first kinematic feature sets of the multiple kinematic segments, dimension reduction processing is performed on the multiple first kinematic feature sets to obtain multiple second kinematic feature sets corresponding to the multiple first kinematic feature sets and their corresponding feature meanings, further, the multiple second kinematic feature sets are clustered to obtain at least one clustering result of the multiple kinematic segments, and multiple vehicle driving condition types are determined using the feature meanings of the second kinematic feature sets and at least one clustering result, so that the obtained vehicle driving condition types are enriched and conform to actual conditions.

[0040] In some embodiments, the driving data includes multiple data segments, and the multiple data segments are represented by the rear motor speed and the longitudinal acceleration of the vehicle, that is, the multiple data segments can be represented by the rear motor speed and the longitudinal acceleration of the vehicle in the driving data, but each data segment in the multiple data segments still includes the corresponding rear electrode speed, longitudinal acceleration of the vehicle, torque, vehicle driving time, mileage, etc.

[0041] At this time, the driving data is divided to obtain multiple kinematic segments, including: dividing the data segments that meet the preset conditions in the multiple data segments into one kinematic segment, the preset condition being that the rear motor speed in the data segment is converted from the first motor speed to the second motor speed, and then from the second motor speed to the first motor speed; using the longitudinal acceleration of the vehicle to divide each of the multiple kinematic segments, and determining the idle segment data, acceleration segment data, uniform speed segment data and deceleration segment data of each kinematic segment.

[0042] The data segments that meet the preset conditions in the multiple data segments are divided into a kinematic segment. The preset condition is that the rear motor speed in the data segment is converted from the first motor speed to the second motor speed, and then from the second motor speed to the first motor speed. For example, the data segment corresponding to the process of the rear motor speed changing from non-0 last time to 0 and then to non-0 again is divided into a kinematic segment, such as Figure 2 As shown, Figure 24 is a schematic diagram of a kinematic segment of an embodiment of the present application, in which the time period of 40s-100s corresponds to a kinematic segment, 40s-60s is an idle part, that is, the motor speed is 0; 60s-100s is a motion part, that is, the motor speed changes from non-0 to 0. Further, each kinematic segment in a plurality of kinematic segments is divided by the longitudinal acceleration of the vehicle, and the idle segment data, acceleration segment data, uniform speed segment data and deceleration segment data of each kinematic segment are determined. For example, it can be subdivided into an idle state, an acceleration state, a uniform speed state and a deceleration state based on whether the longitudinal acceleration signal of the vehicle in each kinematic segment is within a given threshold range. For example, the given threshold of the longitudinal acceleration is 0.2, when the motor speed is 0, it corresponds to the idle segment data; when the longitudinal acceleration is greater than the threshold 0.2, it corresponds to the acceleration segment data; when the longitudinal acceleration is less than -0.2, it corresponds to the deceleration segment data; when the longitudinal acceleration is between -0.2 and 0.2, it corresponds to the uniform speed segment data.

[0043] Among them, before dividing the data segments that meet the preset conditions among multiple data segments into a kinematic segment, they can be screened based on the vehicle computer signal, and the data segments of the power-on state signal can be retained to perform preliminary division of the driving segments.

[0044] In some embodiments, the first kinematic feature set includes a time feature set, a speed feature set and a torque feature set; wherein the time feature set includes at least one of running distance, running time, idling time, acceleration time, deceleration time and uniform speed time; the speed feature set includes at least one of average rear motor speed, average driving rear motor speed, maximum rear motor speed, maximum acceleration, minimum acceleration, average acceleration in acceleration section, average deceleration in deceleration section, standard deviation of rear motor speed and standard deviation of rear motor acceleration; the torque feature set includes at least one of average torque, average torque during driving, maximum torque, average torque in acceleration section, average torque in deceleration section, average torque in uniform speed section, standard deviation of torque, maximum torque change rate, minimum torque change rate, average torque change rate in acceleration section, average torque change rate in deceleration section and average torque change rate in uniform speed section.

[0045] Feature construction is performed on each of the multiple kinematic segments to obtain a first kinematic feature set corresponding to each of the multiple kinematic segments, where the first kinematic feature set includes a time feature set, a rotation speed feature set, and a torque feature set.

[0046] Among them, the time feature set includes at least one of running distance, running time, idling time, acceleration time, deceleration time and uniform speed time. For example, the time feature set of each kinematic segment includes the corresponding vehicle running distance, vehicle running time, vehicle idling time, vehicle acceleration time, vehicle deceleration time and vehicle uniform speed time.

[0047] The speed feature set includes at least one of the average rear motor speed, the average driving rear motor speed, the maximum rear motor speed, the maximum acceleration, the minimum acceleration, the average acceleration of the acceleration segment, the average deceleration of the deceleration segment, the standard deviation of the rear motor speed, and the standard deviation of the rear motor acceleration. For example, the speed feature set of each kinematic segment includes the average rear motor speed, the average driving rear motor speed, the maximum rear motor speed, the maximum acceleration, the minimum acceleration, the average acceleration of the acceleration segment, the average deceleration of the deceleration segment, the standard deviation of the rear motor speed, and the standard deviation of the rear motor acceleration.

[0048] Among them, the average rear motor speed is the average value of the rear motor speed of the entire kinematic segment (for example, one kinematic segment among multiple kinematic segments); the average driving rear motor speed is the average value of the rear motor speed based on the entire kinematic segment except the idle part; the maximum rear motor speed is the maximum value of the rear motor speed in the entire kinematic segment; the maximum acceleration is the maximum value of the longitudinal acceleration in the entire kinematic segment; the minimum acceleration is the minimum value of the longitudinal acceleration in the entire kinematic segment; the average acceleration of the acceleration segment is the average value of the longitudinal acceleration in the acceleration state in the entire kinematic segment; the average deceleration of the deceleration segment is the average value of the longitudinal acceleration in the deceleration state in the entire kinematic segment; the standard deviation of the rear motor speed is the standard deviation of the rear motor speed in the entire kinematic segment; the standard deviation of the rear motor acceleration is the standard deviation of the longitudinal acceleration of the entire kinematic segment.

[0049] The torque feature set includes at least one of the average torque, the average torque during driving, the maximum torque, the average torque in the acceleration segment, the average torque in the deceleration segment, the average torque in the uniform speed segment, the torque standard deviation, the maximum torque change rate, the minimum torque change rate, the average torque change rate in the acceleration segment, the average torque change rate in the deceleration segment, and the average torque change rate in the uniform speed segment. For example, the torque feature set of each kinematic segment includes the average torque, the average torque during driving, the maximum torque, the average torque in the acceleration segment, the average torque in the deceleration segment, the average torque in the uniform speed segment, the torque standard deviation, the maximum torque change rate, the minimum torque change rate, the average torque change rate in the acceleration segment, the average torque change rate in the deceleration segment, and the average torque change rate in the uniform speed segment.

[0050] Among them, the average torque is the average value of the torque of the entire kinematic segment (for example, one kinematic segment among multiple kinematic segments); the average torque during driving is the average value of the torque in the entire kinematic segment except the idle part; the maximum torque is the maximum value of the torque in the entire kinematic segment; the average torque of the acceleration segment is the average value of the torque in the acceleration state in the entire kinematic segment; the average torque of the deceleration segment is the average value of the torque in the deceleration state in the entire kinematic segment; the average torque of the uniform speed segment is the average value of the torque in the uniform speed state in the entire kinematic segment; the torque standard deviation is the standard deviation of the torque in the entire kinematic segment; the maximum torque change rate is the maximum value of the torque change rate in the entire kinematic segment; the minimum torque change rate is the minimum value of the torque change rate in the entire kinematic segment; the average torque change rate of the acceleration segment is the average value of the torque change rate in the acceleration state in the entire kinematic segment; the average torque change rate of the deceleration segment is the average value of the torque change rate in the deceleration state in the entire kinematic segment; the average torque change rate of the uniform speed segment is the average value of the torque change rate in the uniform speed state in the entire kinematic segment.

[0051] In this embodiment, the first kinematic feature set includes a time feature set, a rotational speed feature set, and a torque feature set, that is, a torque feature is introduced into the processed data, so that the clustered working conditions are more representative and more complex.

[0052] Furthermore, the time domain features derived from the speed and torque can be more complex and detailed, such as statistical features such as kurtosis and skewness; time series features such as autocorrelation coefficient and lagged difference; frequency domain features such as power spectral density and spectral entropy; nonlinear features such as the maximum Lyapunov exponent, and so on.

[0053] In some embodiments, dimensionality reduction processing is performed on multiple first kinematic feature sets, including: standardizing the multiple first kinematic feature sets to obtain multiple processed first kinematic feature sets; performing principal component analysis on the multiple processed first kinematic feature sets to obtain a first number of eigenvalues ​​and a corresponding first number of eigenvectors; obtaining multiple principal component features and their contribution rates based on the first number of eigenvalues ​​and the corresponding first number of eigenvectors, wherein the i-th principal component feature among the multiple principal component features is represented by the i-th eigenvector corresponding to the multiple first kinematic feature sets and the first number of eigenvectors, and the contribution rate of the i-th principal component is represented by the first number of eigenvalues ​​and the i-th eigenvalue corresponding to the i-th principal component feature; and determining the principal component feature that meets the contribution rate threshold among the multiple principal component features as the target principal component feature.

[0054] The plurality of first kinematic feature sets are normalized to obtain processed first kinematic feature sets, for example, the plurality of first kinematic feature sets are normalized using a Z-Score normalization method, wherein: Where x is the first kinematic characteristic, is the mean of the feature, and σ is the standard deviation of the feature. Further, principal component analysis is performed on the processed multiple first kinematic feature sets to obtain a first number of eigenvalues ​​and a corresponding first number of eigenvectors, for example, to obtain n processed first kinematic feature sets X n×p , X n×p Including p first kinematic features, calculate X n×p The covariance matrix Σ p×p , and based on the covariance matrix Σ p×p Find the corresponding p eigenvalues ​​and sort them from large to small to get λ1, λ2, ..., λ p , and the corresponding p eigenvectors T1,T2,...,T p .

[0055] Based on the first number of eigenvalues ​​and the corresponding first number of eigenvectors, multiple principal component features and their contribution rates are obtained, wherein the i-th principal component feature among the multiple principal component features is represented by the i-th eigenvector corresponding to the multiple first kinematic feature sets and the first number of eigenvectors, and the contribution rate of the i-th principal component is represented by the first number of eigenvalues ​​and the i-th eigenvalue corresponding to the i-th principal component feature; the principal component feature that meets the contribution rate threshold among the multiple principal component features is determined as the target principal component feature.

[0056] Wherein, a plurality of principal component features are obtained according to the first number of eigenvalues ​​and the corresponding first number of eigenvectors, and the i-th principal component feature in the plurality of principal component features is represented by the i-th eigenvector corresponding to the first number of eigenvectors and the plurality of first kinematic feature sets, for example, the i-th principal component Y i For: Y i =X·T i ,1≤i≤p.

[0057] The contribution rates of multiple principal component features include the contribution rate of the kth principal component among the i principal components and the cumulative contribution rates of the first m principal components.

[0058] Among them, the contribution rate of the kth principal component is:

[0059]

[0060] The cumulative contribution rate of the first m principal components is:

[0061]

[0062] It can be understood that the number of principal components is selected by calculating the principal component contribution rate and the cumulative contribution rate, and then the target principal component characteristics are obtained. For example, the contribution rate threshold is that the cumulative contribution rate reaches more than 85%, and the first m principal components with cumulative contribution rates of more than 85% are used as the target principal component characteristics. The original information of the data can be well preserved, and the purpose of dimensionality reduction can be achieved.

[0063] In some embodiments, the second kinematic feature set includes a target principal component feature, and the feature meaning of the second kinematic feature set is determined based on a feature vector corresponding to the target principal component feature.

[0064] For example, the first kinematic feature set includes running distance, running time, idling time, acceleration time, deceleration time, uniform speed time, average rear motor speed, average driving rear motor speed, maximum rear motor speed, maximum acceleration, minimum acceleration, average acceleration in acceleration section, average deceleration in deceleration section, standard deviation of rear motor speed, standard deviation of rear motor acceleration; the torque feature set includes average torque, average torque during driving, maximum torque, average torque in acceleration section, average torque in deceleration section, average torque in uniform speed section, standard deviation of torque, maximum rate of change of torque, minimum rate of change of torque, average rate of change of acceleration section There are 27 features in total, including the average torque change rate, the average torque change rate in the deceleration section, and the average torque change in the uniform speed section. Through principal component analysis of multiple first kinematic feature sets, it is calculated that the cumulative contribution rate of the first 7 principal components is more than 85%, and then the first m principal components are determined as the target principal component features. At this time, the second kinematic feature set includes the target principal component features, that is, the second kinematic feature set includes the obtained m principal component features, wherein the characteristic meaning of each principal component feature is determined based on the feature vector corresponding to the principal component feature, wherein the feature vector corresponding to a principal component feature can be represented by a set of weight coefficients.

[0065] If the first kinematic feature set includes 27 features, the order is the 6 running time related features, 9 speed related features and 12 torque related features. During the model training process, the threshold of the cumulative contribution rate during principal component analysis dimensionality reduction is 85%. When the cumulative contribution rate exceeds 85%, for example, the number of principal components is 7, the cumulative variance contribution rate reaches 0.87314, that is, the second kinematic feature set includes the first principal component feature, the second principal component feature, the third principal component feature, the fourth principal component feature, the fifth principal component feature, the sixth principal component feature and the seventh principal component feature.

[0066] For example, the eigenvalue corresponding to the first principal component feature is 11.98, and the eigenvector corresponding to the first principal component, that is, the first weight coefficient of the first principal component with respect to multiple first kinematic feature sets is [0.109, 0.141, 0.023, 0.153, 0.151, 0.11, 0.238, 0.249, 0.262, 0.263, -0.234, 0.212, -0.178, 0.251, 0.238, 0.068, -0.014, 0.264, 0.226, -0.212, -0.006, 0.247, 0.259, -0.249, 0.114, -0.158, -0.054], the variance contribution rate is 0.4436, and the cumulative contribution rate is 0.4436. Based on the first weight coefficient, it can be seen that the weights of the time domain features related to speed and torque are relatively large, around 0.25, and the weights of their standard deviations are also relatively large. Therefore, the characteristic meaning of the first principal component feature is attributed to reflecting the average level of speed and torque, as well as the extreme cases of their fluctuations.

[0067] Similarly, the characteristic meanings of the second principal component feature, the third principal component feature, the fourth principal component feature, the fifth principal component feature, the sixth principal component feature, and the seventh principal component feature are summarized in turn. For example, the characteristic meaning of the second principal component feature is to reflect the running distance and running time, wherein the larger the characteristic value, the longer the driving distance and the longer the driving time; the characteristic meaning of the third principal component feature is to reflect the average torque level, including the average torque of the entire kinematic segment and the average torque during driving, wherein the larger the characteristic value, the greater the torque; the characteristic meaning of the fourth principal component feature is to reflect the change of the average torque of uniform acceleration The characteristic meaning of the fifth principal component feature is to reflect the idle time and the uniform torque level, wherein the larger the characteristic value is, the longer the idle speed is and the larger the torque is at uniform speed; the characteristic meaning of the sixth principal component feature is to reflect the change of the average torque of uniform deceleration, wherein the larger the characteristic value is, the larger the torque is in the deceleration section, and the smaller the torque is in the uniform speed section; the characteristic meaning of the seventh principal component feature is to reflect the limit of the deceleration section, for example, the average deceleration weight of the deceleration section is much greater than that of other features, wherein the larger the characteristic value is, the milder the deceleration is, and the smaller the value is, the more violent the deceleration is.

[0068] It can be understood that the feature meaning of each second kinematic feature in the second kinematic feature set is used to reflect the change of a certain driving parameter. Dimensionality reduction processing can also use compression features such as factor analysis, linear discriminant analysis, and autoencoder.

[0069] In some embodiments, clustering is performed on multiple second kinematic feature sets to obtain at least one clustering result of multiple kinematic segments, including: determining the number of target categories; standardizing the multiple second kinematic feature sets to obtain multiple processed second kinematic feature sets; clustering the processed multiple second kinematic feature sets to obtain clustering results of the number of target categories.

[0070] Determine the number of target categories, which may be an optimal number of clusters, for example, the optimal number of clusters is 6. Perform standardization on the plurality of second kinematic feature sets to obtain a plurality of processed second kinematic feature sets, for example, perform standardization on the plurality of second kinematic feature sets using a maximum-minimum normalization algorithm, wherein x min and x max are the minimum and maximum values ​​of the feature variable, respectively. Clustering is performed on the processed multiple second kinematic feature sets to obtain clustering results of several target categories. For example, clustering is performed on the processed multiple second kinematic feature sets to obtain 6 clustering results, each of which consists of multiple kinematic segments.

[0071] It can be understood that k-means clustering is used to cluster the data after dimensionality reduction. Each class center is only related to the principal component characteristics. It is difficult to obtain the specific meaning of each type of working condition through principal component analysis. That is, the characteristics or meaning of each principal component is summarized through the conversion relationship between the principal component and the original feature, and then the working condition category is summarized according to the size of each principal component in each class center.

[0072] Among them, in some embodiments, determining the target category number includes: obtaining a preset category number; based on the preset category number, determining at least one initial class center corresponding to the preset category number using a hierarchical clustering method; based on at least one initial class center, clustering multiple second kinematic feature sets using a K-Means clustering algorithm to obtain classes corresponding to the preset category number; calculating the class average diameter of the class corresponding to the preset category number; and judging whether the preset category number is the target category number based on the class average diameter.

[0073] Get a preset number of categories, for example, a preset number of categories c, where c can be 1, 2, 3, ..., 9. Based on the preset number of categories, use a hierarchical clustering method to determine at least one initial class center corresponding to the preset number of categories. For example, for the number of categories c, use hierarchical clustering to select the initial center, that is, hierarchical clustering uses the number of categories c to obtain the clustering results, and then calculate c class centers. Based on at least one initial class center, use the K-Means clustering algorithm to cluster multiple second kinematic feature sets to obtain classes corresponding to the preset number of categories, that is, select the sample point closest to the class center of each of the c classes as the initial center, and then use the K-Means algorithm for clustering. Calculate the class average diameter of the class corresponding to the preset number of categories, that is, calculate the class average diameter under the current number of categories. Further, based on the class average diameter, determine whether the preset number of categories is the target number of categories. For example, as the number of clusters increases, the class average diameter gradually decreases until the optimal number of categories is reached, and the class average diameter tends to stabilize, and then the optimal number of categories can be determined, such as Figure 3 As shown, Figure 3 This is a line graph of the changes in the average diameter of the classes according to an embodiment of the present application. When the number of classes is 6, the average diameter of the classes increases, and thus it can be determined that the number of target classes is 6.

[0074] In this embodiment, the K-Means clustering algorithm can also be a density clustering algorithm such as DBSCAN, in which the hierarchical clustering method is used to determine the initial center to avoid differences in clustering results due to different random initial points each time clustering, and the optimal number of clusters is determined by the relationship between the average class diameter and the number of categories, so that the clustering results are reasonable.

[0075] In some embodiments, the type of vehicle driving condition corresponding to driving data is determined using the feature meanings of multiple second kinematic feature sets and at least one clustering result, including: obtaining a kinematic segment corresponding to any clustering result in at least one clustering result; calculating feature values ​​corresponding to multiple second kinematic feature sets in the kinematic segment corresponding to the clustering result; and determining the type of vehicle driving condition represented by the clustering result based on the feature values ​​and the feature meanings of the second kinematic feature sets.

[0076] Obtain a kinematic segment corresponding to any clustering result in at least one clustering result, for example, obtain multiple kinematic segments corresponding to a clustering result, and calculate the eigenvalues ​​corresponding to multiple second kinematic feature sets in the kinematic segment corresponding to the clustering result, for example, including the first principal component eigenvalue, the second principal component eigenvalue, the third principal component eigenvalue, the fourth principal component eigenvalue, the fifth principal component eigenvalue, the sixth principal component eigenvalue, and the seventh principal component eigenvalue. Further, based on the eigenvalue and the characteristic meaning of the second kinematic feature set, determine the type of vehicle driving condition represented by the clustering result, and the characteristic meaning of each feature in the second kinematic feature set is known to reflect the situation, and then based on the specific first principal component eigenvalue, the second principal component eigenvalue, the third principal component eigenvalue, the fourth principal component eigenvalue, the fifth principal component eigenvalue, the sixth principal component eigenvalue, and the seventh principal component eigenvalue, analyze and determine the type of vehicle driving condition of the corresponding category.

[0077] The present application calculates the eigenvalues ​​corresponding to multiple second kinematic feature sets in the kinematic segments corresponding to the clustering results respectively, and then determines the vehicle driving condition type represented by the clustering results based on the eigenvalues ​​and the feature meanings of the second kinematic feature sets. That is, the vehicle driving condition type represented by different clustering results is determined based on the meaning analysis of different principal components in different clustering results, so that the obtained vehicle driving condition types are rich and consistent with actual conditions.

[0078] For ease of understanding, the identification method of the vehicle driving condition of the embodiment of the present application is illustrated by example. For example, the historical driving data of 1,000 vehicles for one day is randomly selected. Through data preprocessing and kinematic segment division, 67,181 kinematic segments are obtained. By dividing the kinematic segments, 27 first kinematic features are derived, such as the order of running distance, running time, idling time, acceleration time, deceleration time, uniform speed time, average rear motor speed, average driving rear motor speed, maximum rear motor speed, maximum acceleration, minimum acceleration, average acceleration of the acceleration segment, average deceleration of the deceleration segment, standard deviation of rear motor speed, standard deviation of rear motor acceleration, torque feature set including average torque, average torque during driving, maximum torque, average torque of the acceleration segment, average torque of the deceleration segment, average torque of the uniform speed segment, torque standard deviation, maximum torque change rate, minimum torque change rate, average torque change rate of the acceleration segment, average torque change rate of the deceleration segment, and average torque change of the uniform speed segment.

[0079] When principal component analysis and dimensionality reduction are performed on the 27 first kinematic features of multiple kinematic segments, the threshold of the cumulative contribution rate is 85%. When the cumulative contribution rate exceeds 85%, the number of principal components is 7. Through analysis, it is determined that the characteristic meaning of the first principal component feature is attributed to reflecting the average level of speed and torque, as well as the extreme conditions of their fluctuations; the characteristic meaning of the second principal component feature is to reflect the running distance and running time; the characteristic meaning of the third principal component feature is to reflect the average torque level, including the average torque of the entire kinematic segment and the average torque during driving; the characteristic meaning of the fourth principal component feature is to reflect the change of the average torque of uniform acceleration; the characteristic meaning of the fifth principal component feature is to reflect the idle time and the uniform torque level; the characteristic meaning of the sixth principal component feature is to reflect the change of the average torque of uniform deceleration; the characteristic meaning of the seventh principal component feature is to reflect the limit of the deceleration section.

[0080] Furthermore, clustering is performed on the processed multiple second kinematic feature sets to obtain 6 clustering results (0, 1, 2, 3, 4, 5), as shown in the following table:

[0081] pc1 pc2 pc3 pc4 pc5 pc6 pc7 Number of fragments 0 8.758 10.524 2.289 -2.456 -0.040 -0.664 -0.702 875 1 -0.149 -0.148 -0.458 -0.299 2.252 0.508 -0.315 7941 2 -3.630 0.041 2.560 0.341 -0.016 0.361 0.290 7814 3 0.196 -0.935 0.075 -0.127 -0.653 -0.124 -1.412 8956 4 -3.500 0.492 -0.851 -0.419 -0.402 -0.394 0.269 14753 5 2.676 -0.267 -0.240 0.341 -0.221 0.025 0.355 26842

[0082] Among them, pc1, pc2, pc3, pc4, pc5, pc6, and pc7 correspond to the relevant characteristic values ​​of the first principal component feature, the second principal component feature, the third principal component feature, the fourth principal component feature, the fifth principal component feature, the sixth principal component feature, and the seventh principal component feature, respectively. For example, pc1 (8.758) is the characteristic average value of the first principal component feature of 875 kinematic segments corresponding to the first category (0) clustering results.

[0083] Taking the analysis of the first type of working condition (0) as an example, it can be seen from the table that the first principal component value (pc1-8.758), the second principal component value (pc2-10.524) and the fourth principal component value (pc4-2.456) are significantly different from the other types of working conditions. The larger pc1 value indicates that the speed and torque are larger, the larger pc2 value indicates that the running distance and running time are larger, and the smaller pc4 value indicates that the torque in the acceleration stage is larger and the torque in the uniform speed stage is smaller. Therefore, the first type of working condition can be summarized as sustained The second condition (1) can be summarized as a start-stop condition with a long idling time and then a large torque climbing condition; the third condition (2) can be summarized as a relatively stable low-speed and high-torque climbing condition; the fourth condition (3) can be summarized as a sudden deceleration and emergency braking condition in a very short time; the fifth condition (4) can be summarized as a low-speed and small-torque condition in a relatively short time, and an urban traffic light condition; the sixth condition (5) can be summarized as a slight condition with smooth acceleration and deceleration.

[0084] It should be noted that the vehicle driving condition identification method of the embodiment of the present application can be used to perform relevant model training of the vehicle driving condition clustering model to obtain the vehicle driving condition clustering model. The vehicle driving condition clustering model is used to identify the vehicle driving condition, wherein the vehicle driving condition types are rich and consistent with the actual conditions. For example, by inputting the vehicle driving data into the trained vehicle driving condition clustering model, the vehicle driving condition clustering model outputs the corresponding classification data and the vehicle driving condition type.

[0085] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0086] See also Figure 4 , Figure 4 4 is a schematic diagram of the structure of an electronic device of an embodiment of the present application. The electronic device 40 includes a memory 41 and a processor 42 coupled to each other, and the processor 42 is used to execute program instructions stored in the memory 41 to implement the steps of the above-mentioned vehicle driving condition identification method embodiment. In a specific implementation scenario, the electronic device 40 may include but is not limited to: a microcomputer, a server, which is not limited here.

[0087] Specifically, the processor 42 is used to control itself and the memory 41 to implement the steps of the above-mentioned vehicle driving condition identification method embodiment. The processor 42 can also be called a CPU (Central Processing Unit), and the processor 42 may be an integrated circuit chip with signal processing capabilities. The processor 42 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 42 can be implemented by an integrated circuit chip.

[0088] See also Figure 5 , Figure 5 Schematic diagram of the structure of a non-volatile computer-readable storage medium according to an embodiment of the present application. The non-volatile computer-readable storage medium 50 is used to store a computer program 501. When the computer program 501 is executed by a processor, for example, Figure 4When the processor 42 in the embodiment is executed, it is used to implement the steps of the above-mentioned embodiment of the method for identifying vehicle driving conditions.

[0089] The above description of various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other, and for the sake of brevity, they will not be repeated herein.

[0090] In the several embodiments provided in the present application, it should be understood that the disclosed methods and related devices can be implemented in other ways. For example, the above-described related device implementations are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication disconnection shown or discussed can be through some interfaces, indirect coupling or communication disconnection of devices or units, which can be electrical, mechanical or other forms.

[0091] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0092] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0093] It is easy for a person skilled in the art to know that many modifications and changes can be made to the device and method while maintaining the teaching content of the present application. Therefore, the above disclosure should be considered as limited only by the scope of the appended claims.

Claims

1. A method for identifying vehicle driving conditions, characterized in that: include: Obtain vehicle driving data; Dividing the driving data to obtain a plurality of kinematic segments; Performing feature construction on the multiple kinematic segments to obtain multiple first kinematic feature sets of the multiple kinematic segments; Performing dimensionality reduction processing on the multiple first kinematic feature sets to obtain multiple second kinematic feature sets corresponding to the multiple first kinematic feature sets and their corresponding feature meanings; Performing clustering processing on the plurality of second kinematic feature sets to obtain at least one clustering result of the plurality of kinematic segments; The vehicle driving condition type corresponding to the driving data is determined by utilizing the feature meanings of the plurality of second kinematic feature sets and the at least one clustering result.

2. The method according to claim 1, characterized in that Performing dimensionality reduction processing on the plurality of first kinematic feature sets comprises: Performing standardization processing on the plurality of first kinematic feature sets to obtain a plurality of processed first kinematic feature sets; Performing principal component analysis on the processed plurality of first kinematic feature sets to obtain a first number of eigenvalues ​​and corresponding first number of eigenvectors; According to the first number of eigenvalues ​​and the corresponding first number of eigenvectors, a plurality of principal component features and their contribution rates are obtained, wherein an i-th principal component feature among the plurality of principal component features is represented by the i-th eigenvector corresponding to the plurality of first kinematic feature sets and the first number of eigenvectors, and the contribution rate of the i-th principal component is represented by the i-th eigenvalue corresponding to the first number of eigenvalues ​​and the i-th principal component feature; A principal component feature that meets a contribution rate threshold among the plurality of principal component features is determined as a target principal component feature.

3. The method according to claim 2, characterized in that The second kinematic feature set includes the target principal component feature, and the feature meaning of the second kinematic feature set is determined based on the feature vector corresponding to the target principal component feature.

4. The method according to claim 1, characterized in that: Clustering the plurality of second kinematic feature sets to obtain at least one clustering result of the plurality of kinematic segments includes: Determine the number of target categories; Performing standardization processing on the plurality of second kinematic feature sets to obtain a plurality of processed second kinematic feature sets; Clustering is performed on the processed multiple second kinematic feature sets to obtain clustering results of several target categories.

5. The method according to claim 4, characterized in that Determine the number of target categories, including: Get the number of preset categories; Based on the preset number of categories, determining at least one initial class center corresponding to the preset number of categories using a hierarchical clustering method; Based on the at least one initial class center, clustering the plurality of second kinematic feature sets using a K-Means clustering algorithm to obtain classes corresponding to the preset number of classes; Calculating the average diameter of the classes corresponding to the preset number of classes; Based on the class average diameter, it is determined whether the preset number of classes is a target number of classes.

6. The method according to claim 1, characterized in that Determining the vehicle driving condition type corresponding to the driving data by using the feature meanings of the plurality of second kinematic feature sets and the at least one clustering result includes: Obtaining a kinematic segment corresponding to any clustering result among the at least one clustering result; Calculating feature values ​​corresponding to the plurality of second kinematic feature sets in the kinematic segment corresponding to the clustering result; Based on the characteristic values ​​and the characteristic meanings of the second kinematic characteristic set, the type of vehicle driving condition represented by the clustering result is determined.

7. The method according to claim 1, characterized in that The driving data includes a plurality of data segments, wherein the plurality of data segments are represented by a rear motor speed and a longitudinal acceleration of the vehicle; The driving data is divided to obtain a plurality of kinematic segments, including: Dividing a data segment that meets a preset condition among the multiple data segments into one of the kinematic segments, wherein the preset condition is that the rear motor speed in the data segment is converted from the first motor speed to the second motor speed, and then from the second motor speed to the first motor speed; Each kinematic segment of the plurality of kinematic segments is divided by using the longitudinal acceleration of the vehicle, and idling segment data, acceleration segment data, uniform speed segment data and deceleration segment data of each kinematic segment are determined.

8. The method according to claim 7, characterized in that The first kinematic feature set includes a time feature set, a speed feature set and a torque feature set; The time feature set includes at least one of running distance, running time, idling time, acceleration time, deceleration time and constant speed time; The speed feature set includes at least one of an average rear motor speed, an average driving rear motor speed, a maximum rear motor speed, a maximum acceleration, a minimum acceleration, an average acceleration of an acceleration section, an average deceleration of a deceleration section, a rear motor speed standard deviation, and a rear motor acceleration standard deviation; The torque feature set includes at least one of average torque, average torque during driving, maximum torque, average torque in acceleration section, average torque in deceleration section, average torque in uniform speed section, torque standard deviation, maximum torque change rate, minimum torque change rate, average torque change rate in acceleration section, average torque change rate in deceleration section, and average torque change rate in uniform speed section.

9. An electronic device, characterized in that: It comprises a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory to implement the method for identifying the vehicle driving condition as described in any one of claims 1 to 8.

10. A non-volatile computer-readable storage medium having program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, the method for identifying the vehicle driving condition according to any one of claims 1 to 8 is implemented.