Hybrid test method and system for hybrid electric vehicle
By constructing performance evaluation network and aggregated sample working condition data, the problem of high manual participation in hybrid tests of hybrid vehicles is solved, the testing efficiency and evaluation accuracy are improved, and the manual error is reduced.
Patent Information
- Application Number
- CN202411217302.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-02
- Publication Date
- 2024-12-13
- Estimated Expiration
- 2044-09-02
AI Technical Summary
The manual participation in the hybrid test of existing hybrid vehicles is too high, resulting in inconsistent test accuracy, low efficiency and insufficient credibility, affecting performance evaluation.
Through interaction, the target test performance set is obtained, the performance index mapping table is constructed, the sample working condition data is aggregated, the hybrid test sequence is generated, the vehicle sensor network is used for data acquisition, and the performance evaluation network is built based on the convolutional neural network to perform performance evaluation.
It improves the efficiency of hybrid testing and the accuracy and credibility of performance evaluation, and reduces the impact of manual errors.
Smart Images
Figure CN118730574B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to a hybrid test method and system for hybrid electric vehicles. Background Art
[0002] There is a significant problem in the current hybrid vehicle hybrid test process, which is excessive reliance on human participation. This excessive reliance will lead to two problems. First, different operators may produce inconsistent results in the same test phase. This error in human experience may affect the accuracy of the hybrid test. Second, manual participation in the adjustment of hybrid test parameters may also affect the efficiency of the test and the credibility of the hybrid test results, thereby affecting the accurate evaluation of the performance of the hybrid system.
[0003] In summary, in the prior art, the human involvement in the entire hybrid test process of hybrid vehicles is too high, and there are technical problems such as human experience errors affecting the accuracy of hybrid tests, and human participation in hybrid test parameter adjustment affecting the efficiency of hybrid tests and the credibility of hybrid test results. Summary of the invention
[0004] The present application provides a hybrid testing method and system for hybrid vehicles, which are used to solve the technical problems in the prior art that the human participation in the entire hybrid testing process of hybrid vehicles is too high, there are human experience errors affecting the accuracy of hybrid testing, and human participation in hybrid testing parameter adjustment affects the efficiency of hybrid testing and affects the credibility of hybrid testing results.
[0005] In view of the above problems, the present application provides a hybrid test method and system for hybrid electric vehicles.
[0006] In a first aspect of the present application, a hybrid test method for a hybrid vehicle is provided, the method comprising: interactively obtaining a target test performance set for a hybrid test of a vehicle to be tested; pre-building a performance indicator mapping table, and using the target test performance set as a constraint to perform associated indicator calls on the performance indicator mapping table to obtain a target associated indicator set; interactively obtaining multiple sample operating condition sets and multiple sample hybrid test sets of multiple sample vehicles of the same model as the vehicle to be tested, and constructing a hybrid performance evaluation network based on the multiple sample hybrid test sets, the target test performance set, and the target associated indicator set; and The working condition set is aggregated to generate multiple hybrid test working conditions, and working condition feature correlation analysis is performed on the multiple hybrid test working conditions to obtain a hybrid test sequence; based on the hybrid test sequence, continuous hybrid testing is performed on the vehicle to be tested, and data collection of the control process is performed based on the vehicle sensor network of the vehicle to be tested to generate a hybrid test data set, wherein the hybrid test data set includes multiple monitoring time series data of multiple monitoring indicators; the hybrid test data set is divided according to the target correlation indicator set, and the division result is synchronized to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result.
[0007] In a second aspect of the present application, a hybrid test system for a hybrid vehicle is provided, the system comprising: a test performance interaction unit, for interactively obtaining a target test performance set for a hybrid test of a vehicle to be tested; an associated index calling unit, for pre-building a performance index mapping table, and performing associated index calling on the performance index mapping table with the target test performance set as a constraint to obtain a target associated index set; an evaluation network construction unit, for interactively obtaining a plurality of sample operating condition sets and a plurality of sample hybrid test sets of a plurality of sample vehicles of the same model as the vehicle to be tested, and constructing a hybrid performance evaluation network according to the plurality of sample hybrid test sets, the target test performance set and the target associated index set; a test sequence output unit, Used to generate multiple hybrid test conditions by aggregating the conditions of the multiple sample condition sets, perform condition feature correlation analysis on the multiple hybrid test conditions, and obtain a hybrid test sequence; a test data acquisition unit, used to perform continuous hybrid tests on the vehicle to be tested based on the hybrid test sequence, and perform data collection of the control process based on the vehicle sensor network of the vehicle to be tested to generate a hybrid test data set, wherein the hybrid test data set includes multiple monitoring time series data of multiple monitoring indicators; an evaluation result output unit, used to divide the hybrid test data set according to the target correlation indicator set, and synchronize the division result to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result.
[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0009] The method provided in the embodiment of the present application obtains a target test performance set for hybrid testing of a vehicle to be tested through interaction; pre-constructs a performance indicator mapping table, and uses the target test performance set as a constraint to call the associated indicators of the performance indicator mapping table to obtain a target associated indicator set; interactively obtains multiple sample operating condition sets and multiple sample hybrid test sets of multiple sample vehicles of the same model as the vehicle to be tested, and constructs a hybrid performance evaluation network according to the multiple sample hybrid test sets, the target test performance set, and the target associated indicator set; generates multiple hybrid test operating conditions by operating condition aggregation of the multiple sample operating condition sets, performs operating condition feature association analysis on the multiple hybrid test operating conditions, and obtains a hybrid test sequence; performs continuous hybrid testing on the vehicle to be tested based on the hybrid test sequence, and performs data collection of a control process based on a vehicle sensor network of the vehicle to be tested to generate a hybrid test data set, wherein the hybrid test data set includes multiple monitoring time series data of multiple monitoring indicators; divides the hybrid test data set according to the target associated indicator set, and synchronizes the division result to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result. The technical effect of improving the test efficiency of hybrid vehicle hybrid testing and improving the evaluation accuracy and evaluation credibility of hybrid vehicle performance evaluation based on hybrid testing is achieved. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Figure 1 A flow chart of a hybrid test method for a hybrid vehicle provided in this application;
[0011] Figure 2 A schematic diagram of a flow chart of obtaining a hybrid test sequence in a hybrid test method for a hybrid vehicle provided in the present application;
[0012] Figure 3 A schematic diagram of the structure of a hybrid test system for a hybrid electric vehicle provided in this application.
[0013] Explanation of the accompanying drawings: test performance interaction unit 1, associated indicator calling unit 2, evaluation network construction unit 3, test sequence output unit 4, test data acquisition unit 5, evaluation result output unit 6. DETAILED DESCRIPTION
[0014] The present application provides a hybrid test method and system for hybrid electric vehicles, which is used to solve the technical problems in the prior art that the human involvement in the whole process of hybrid test of hybrid electric vehicles is too high, there are human experience errors affecting the accuracy of hybrid test, and human participation in hybrid test parameter adjustment affects the efficiency of hybrid test and the credibility of hybrid test results. The technical effect of improving the test efficiency of hybrid electric vehicle hybrid test and improving the evaluation accuracy and evaluation credibility of hybrid electric vehicle performance evaluation based on hybrid test is achieved.
[0015] Below, the technical solutions in the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the accompanying drawings, rather than all of them.
[0016] Embodiment 1
[0017] like Figure 1 As shown, the present application provides a hybrid test method for a hybrid electric vehicle, comprising:
[0018] A100: interactively obtain a target test performance set for hybrid testing of a vehicle to be tested;
[0019] Specifically, in this embodiment, the target test performance set is a collection of test objectives or standards set for various performance indicators of the hybrid power system of the automobile to be tested, including but not limited to K hybrid test performances such as fuel economy, emission level, power performance, driving performance, electric driving range, system switching efficiency, battery charging efficiency, reliability and durability.
[0020] A200: pre-build a performance indicator mapping table, and perform associated indicator calls on the performance indicator mapping table with the target test performance set as a constraint to obtain a target associated indicator set;
[0021] Specifically, it should be understood that each hybrid test performance evaluation often requires reference to one or more indicator data. For example, for the power performance evaluation of the hybrid test performance, the indicator data that needs to be referenced include acceleration time, engine power and torque.
[0022] Based on this, this embodiment interactively obtains multiple sample test performances and multiple sample performance-related indicator groups, performs performance indicator mapping and tabulation on the multiple sample test performances and the multiple sample performance-related indicator groups, and obtains a performance indicator mapping table.
[0023] The K hybrid test performances in the target test performance set are used to traverse the performance indicator mapping table one by one to call the sample performance associated indicator set of the sample test performance corresponding to the hybrid test performance, and K groups of associated indicators are obtained to form the target associated indicator set.
[0024] The hybrid test performance in the target test performance set is mapped to the associated indicators in the target associated indicator set, and the target test performance set includes K of the hybrid test performances, and the target associated indicator set includes K groups of the associated indicators.
[0025] The following is an exemplary performance indicator mapping table (partial):
[0026]
[0027] A300: interactively obtaining multiple sample operating condition sets and multiple sample hybrid test sets of multiple sample vehicles of the same model as the vehicle to be tested, and constructing a hybrid performance evaluation network according to the multiple sample hybrid test sets, the target test performance set, and the target associated index set;
[0028] In one embodiment, a plurality of sample operating condition sets and a plurality of sample hybrid test sets of a plurality of sample vehicles of the same model as the vehicle to be tested are interactively obtained, and a hybrid performance evaluation network is constructed according to the plurality of sample hybrid test sets, the target test performance set, and the target associated index set. Step A300 of the method provided in the present application further includes:
[0029] A310: Dividing the plurality of sample hybrid test sets according to the K hybrid test performances and the K groups of associated indicators to obtain K sample hybrid performance data sets and K sample monitoring data sets;
[0030] A320: constructing a standard hybrid performance evaluation branch based on a convolutional neural network, and using the K sample hybrid performance data sets and the K sample monitoring data sets to train the standard hybrid performance evaluation branch to obtain K hybrid performance evaluation branches;
[0031] A330: Connect the K hybrid performance evaluation branches in parallel to construct the hybrid performance evaluation network.
[0032] Specifically, in this embodiment, the driving conditions of the hybrid vehicle include driving speed (km), load (cargo or human load mass), road condition (bumping coefficient), driving cycle and environmental conditions (temperature, humidity, altitude), and the parameter values of the above indicators are different under different conditions.
[0033] This embodiment interactively obtains multiple sample operating condition sets and multiple sample hybrid test sets of multiple sample vehicles of the same model as the vehicle to be tested in hybrid mode, the sample operating condition set corresponding to each sample vehicle includes multiple groups of operating condition index parameters, and the sample hybrid test set corresponding to the sample vehicle includes multiple sample monitoring data and multiple sample hybrid performance data based on association mapping of the multiple groups of operating condition index parameters, the data structure of each group of sample monitoring data-sample hybrid performance data is performance association index group-test performance item, for example, the data structure of the sample monitoring data group is specific values of acceleration time, engine power, and torque, and the corresponding sample hybrid performance data is the evaluation result of the hybrid performance, namely, power performance.
[0034] In this embodiment, the K hybrid test performances of the target test performance set and the K groups of associated indicators of the target associated indicator set divide the multiple sample hybrid test sets to obtain K sample hybrid performance data sets and K sample monitoring data sets based on the operating condition indicator parameter association mapping.
[0035] A standard hybrid performance evaluation branch is constructed based on a convolutional neural network. The input data of the standard hybrid performance evaluation branch is a set of monitoring data of a sample performance-related indicator group that satisfies a performance indicator mapping table, and the output result is an energy evaluation value of the sample test performance corresponding to the sample performance-related indicator group in the performance indicator mapping table.
[0036] This embodiment obtains a first sample hybrid performance dataset and a first sample monitoring dataset by calling the K sample hybrid performance datasets and the K sample monitoring datasets, and then divides the first sample hybrid performance dataset and the first sample monitoring dataset into a training set, a test set, and a validation set according to a data volume of 17:2:1. A conventional convolutional neural network training method is used to train and test a standard hybrid performance evaluation branch using the training set and the test set, and to verify the performance evaluation accuracy of the standard hybrid performance evaluation branch using the validation set, until the performance evaluation accuracy output accuracy of the standard hybrid performance evaluation branch is stably higher than 98.5%, thereby obtaining a first hybrid performance evaluation branch.
[0037] The same method is adopted to train the standard hybrid performance evaluation branch using the remaining K-1 sample hybrid performance data sets and the remaining K-1 sample monitoring data sets to obtain K-1 hybrid performance evaluation branches.
[0038] The K hybrid performance evaluation branches are connected in parallel to construct the hybrid performance evaluation network. This embodiment will elaborate on the application of the hybrid performance evaluation network in hybrid testing of hybrid vehicles in the subsequent description.
[0039] The hybrid performance evaluation network constructed in this embodiment achieves the technical effect of improving the hybrid performance test efficiency of hybrid electric vehicles and improving the hybrid performance evaluation accuracy and evaluation credibility.
[0040] A400: generating a plurality of hybrid test conditions by performing condition aggregation on the plurality of sample condition sets, and performing condition feature correlation analysis on the plurality of hybrid test conditions to obtain a hybrid test sequence;
[0041] In one embodiment, Figure 2 As shown, by performing condition aggregation on the multiple sample condition sets to generate multiple hybrid test conditions, performing condition feature correlation analysis on the multiple hybrid test conditions to obtain a hybrid test sequence, the method step A400 provided in the present application also includes:
[0042] A410: interactively obtaining a standard operating condition parameter threshold and an operating condition parameter deviation threshold, wherein the standard operating condition parameter threshold includes a plurality of operating condition index value thresholds, and the operating condition parameter deviation threshold includes a plurality of operating condition index deviation thresholds;
[0043] A420: traversing the plurality of sample operating condition sets using the standard operating condition parameter threshold to obtain M groups of non-standard operating condition parameters;
[0044] A430: aggregating the M groups of non-standard operating condition parameters by using the operating condition parameter deviation threshold to obtain the multiple hybrid test operating conditions;
[0045] A440: obtaining a first hybrid test operating condition based on the multiple hybrid test operating conditions, and obtaining multiple operating condition similarity coefficients between the first hybrid test operating condition and the multiple hybrid test operating conditions based on Euclidean distance calculation;
[0046] A450: Serialize the multiple operating condition similarity coefficients, and sort the multiple hybrid test operating conditions based on the serialization results to obtain the hybrid test sequence.
[0047] Specifically, based on step A300, it can be known that this embodiment interactively obtains multiple sample operating condition sets of multiple sample vehicles of the same model as the vehicle to be tested in hybrid mode, and the parameter values of the driving speed (Km), load (cargo or human load mass), road condition (bumping coefficient), driving cycle and environmental conditions (temperature, humidity, altitude) of each sample operating condition in the multiple sample operating condition sets are different.
[0048] The standard operating condition parameter threshold is the value range of the constituent parameters of the hybrid vehicle under ideal operating conditions. The standard operating condition parameter threshold includes multiple operating condition index value thresholds. The multiple operating condition index value thresholds specifically include the driving speed range, the weight range of the goods or people carried by the vehicle, the road surface roughness (bump coefficient) range, the driving cycle range under the same operating condition, and the environmental condition value range (temperature, humidity, altitude value range).
[0049] The operating condition parameter deviation threshold is used to evaluate whether two or more groups of operating condition parameters are in the same operating condition. The operating condition parameter deviation threshold includes multiple operating condition index deviation thresholds, which are parameter deviations of driving speed (Km), load (cargo or human load mass), road condition (bumping coefficient), driving cycle and environmental conditions (temperature, humidity, altitude).
[0050] In this embodiment, the standard operating condition parameter threshold is used to traverse the multiple sample operating condition sets to obtain M groups of non-standard operating condition parameters, and the parameters of one or more operating condition indicators in each group of non-standard operating condition parameters in the M groups of non-standard operating condition parameters do not meet the corresponding operating condition indicator value threshold of the standard operating condition parameter threshold.
[0051] The M groups of non-standard operating condition parameters are aggregated using the operating condition parameter deviation threshold to obtain the multiple hybrid test conditions, and the values of the two groups of operating condition indicators of any two hybrid test conditions in the multiple hybrid test conditions do not meet the operating condition parameter deviation threshold, that is, the multiple hybrid test conditions are different multiple hybrid test conditions, and the data structure of each hybrid test condition is driving speed (Km), load (cargo or human load mass), road condition (bump coefficient), driving cycle and environmental conditions (temperature, humidity, altitude).
[0052] This embodiment obtains a first hybrid test condition based on the multiple hybrid test conditions, and obtains multiple condition similarity coefficients between the first hybrid test condition and the multiple hybrid test conditions based on Euclidean distance calculation.
[0053] The multiple operating condition similarity coefficients are serialized, and the multiple hybrid test conditions are sorted based on the serialization results to obtain the hybrid test sequence. The multiple hybrid test conditions are simulated based on the hybrid test sequence. This can reduce the complexity of test scenario fitting and scenario adjustment when performing hybrid testing on hybrid vehicles, and improve the test efficiency of hybrid testing on hybrid vehicles.
[0054] A500: performing continuous hybrid testing on the vehicle to be tested based on the hybrid testing sequence, and performing data collection of a control process based on a vehicle sensor network of the vehicle to be tested to generate a hybrid testing data set, wherein the hybrid testing data set includes a plurality of monitoring time series data of a plurality of monitoring indicators;
[0055] In one embodiment, before performing continuous hybrid testing on the vehicle to be tested based on the hybrid test sequence, the method steps provided by the present application further include:
[0056] A511: mapping the hybrid test performance in the target test performance set to the associated indicators in the target associated indicator set, wherein the target test performance set includes K hybrid test performances, and the target associated indicator set includes K groups of associated indicators;
[0057] A512: performing indicator aggregation on the K groups of related indicators to obtain multiple related indicators;
[0058] A513: using the multiple associated indicators as constraints, calling monitoring components on the multiple sample hybrid test sets to obtain multiple groups of sample monitoring components, wherein each sample hybrid test set in the multiple sample hybrid test sets includes multiple sample monitoring data and multiple sample hybrid performance data, and each sample monitoring data has a monitoring component identifier;
[0059] A514: Performing associated component analysis according to the multiple associated indicators and the multiple groups of sample monitoring components to generate a test associated component set;
[0060] A515: Perform component status detection on the vehicle to be tested with reference to the test-associated component set. If the status detection of the vehicle to be tested passes, perform continuous hybrid testing on the vehicle to be tested based on the hybrid test sequence.
[0061] In one embodiment, the method step A514 provided in the present application further includes: performing an associated component analysis based on the multiple associated indicators and the multiple groups of sample monitoring components to generate a test associated component set.
[0062] A514-1: Aggregate the monitoring components of the multiple groups of sample monitoring components to obtain multiple monitoring components;
[0063] A514-2: construct and generate multiple topological isolated vertices based on the multiple associated indicators and multiple monitoring components;
[0064] A514-3: Connect the multiple topological isolated vertices according to the multiple groups of sample monitoring components to generate an indicator component topological network;
[0065] A514-4: Performing a topological connection frequency analysis on the indicator component topological network to obtain a plurality of association frequency sequences of the plurality of association indicators;
[0066] A514-5: Perform extreme value calls on the multiple associated frequency sequences to obtain multiple test associated components, and the multiple test associated components constitute the test associated component set.
[0067] In one embodiment, a continuous hybrid test is performed on the vehicle to be tested based on the hybrid test sequence, and data collection of a control process is performed based on a vehicle sensor network of the vehicle to be tested to generate a hybrid test data set. Step A500 of the method provided in the present application further includes:
[0068] A521: pre-constructing an indicator sensor mapping table, and performing associated sensor calls on the multiple associated indicators based on the indicator sensor mapping table as a constraint to obtain a target sensor call set;
[0069] A522: interactively obtain sensor composition information of the vehicle sensor network, and perform sensor missing analysis on the sensor composition information with the target sensor call set as a constraint to generate a sensor optimization strategy;
[0070] A523: completing the sensor deployment of the vehicle sensor network based on the sensor optimization strategy and the test-related component set to generate an optimized sensor network;
[0071] A524: Activate sensor mapping on the optimized sensor network with the target sensor call set as a constraint to obtain a hybrid test sensor network;
[0072] A525: Based on the data collection of the control process performed by the dynamic test sensor network, the hybrid test data set is generated.
[0073] Specifically, it should be understood that there are situations where the related indicators of two hybrid test performances overlap, for example, the related indicators of the performance evaluation of acceleration performance and braking performance both include the vehicle's driving speed.
[0074] Therefore, this embodiment performs indicator aggregation on the K groups of associated indicators of the target associated indicator set to obtain multiple non-repeated associated indicators, and uses the multiple associated indicators as constraints to call monitoring components on the multiple sample hybrid test sets to obtain multiple groups of sample monitoring components. Each sample hybrid test set in the multiple sample hybrid test sets includes multiple sample monitoring data and multiple sample hybrid performance data, and each sample monitoring data has a monitoring component identifier, and the monitoring components include but are not limited to internal combustion engines, hybrid battery packs, transmission systems, and control units.
[0075] The monitoring components of the multiple groups of sample monitoring components are aggregated to obtain multiple monitoring components without duplication, and multiple topological isolated vertices are generated based on the multiple associated indicators and the multiple monitoring components; the multiple topological isolated vertices are connected according to the multiple groups of sample monitoring components to generate an indicator component topological network; the indicator component topological network is subjected to a topological connection frequency analysis, specifically, the topological isolated node of each associated indicator is connected to the topological isolated nodes corresponding to one or more monitoring components, and the multiple monitoring components connected to one associated indicator are sorted according to the connection frequency between the associated indicator and the topological isolated node of the monitoring component to obtain multiple associated frequency sequences of the multiple associated indicators.
[0076] The multiple associated frequency sequences are called for extreme values to obtain multiple test associated components, wherein the test associated components are collection objects for collecting data of one indicator in a hybrid test, and the multiple test associated components constitute the test associated component set.
[0077] It should be understood that in a hybrid test, the test-related components in the test-related component set need to be ensured to be in a non-faulty state, so as to avoid component errors of the hybrid vehicle causing test errors in the hybrid test.
[0078] This embodiment performs component status detection on the vehicle to be tested with reference to the test-associated component set. If the status detection of the vehicle to be tested passes (all test-associated components in the test-associated component set are in a non-fault state), a continuous hybrid test is performed on the vehicle to be tested based on the hybrid test sequence.
[0079] Interactively obtain multiple sensors for obtaining indicator parameter values of multiple sample monitoring indicators in hybrid testing, and obtain an indicator sensor mapping table based on the mapping relationship between multiple sample monitoring indicators and multiple sensors. Use the indicator sensor mapping table as a constraint to perform associated sensor calls on the multiple associated indicators to obtain a target sensor call set.
[0080] Interactively obtain sensor composition information of the vehicle sensor network, and perform sensor missing analysis on the sensor composition information with the target sensor call set as a constraint to obtain a number of sensors included in the target sensor call set but not included in the sensor composition information as the sensor optimization strategy;
[0081] Based on the sensor optimization strategy and the test-related component set, the vehicle sensor network is supplemented with sensor layout to generate an optimized sensor network. Based on the optimized sensor network, effective collection of component operating parameters of any component in the test-related component set during hybrid testing can be achieved.
[0082] The optimized sensor network is activated by sensor mapping with the target sensor call set as a constraint to obtain a hybrid test sensor network, wherein the sensors in the hybrid test sensor network are operating parameter collection sensors of multiple test-associated components of the test-associated component set.
[0083] Based on the data collection of the control process executed by the dynamic test sensor network, the hybrid test data set is generated, and the hybrid test data set includes multiple indicator change sequences of multiple related indicators corresponding to multiple test-related components, and each indicator change sequence corresponds to multiple hybrid test conditions of the hybrid test sequence, that is, each indicator change sequence is the indicator data change situation under multiple hybrid test conditions.
[0084] This embodiment realizes the effective collection of parameters for hybrid test indicators and provides effective raw data for the evaluation of test results of hybrid tests.
[0085] A600: Divide the hybrid test data set according to the target associated indicator set, and synchronize the division results to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result.
[0086] In one embodiment, the hybrid test data set is divided according to the target associated indicator set, and the division result is synchronized to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result. The method step A600 provided in the present application also includes:
[0087] A610: Dividing the hybrid test data set according to the target associated indicator set to obtain K groups of hybrid operating condition test sequences;
[0088] A620: Synchronize the K groups of hybrid operating condition test sequences to the K hybrid performance evaluation branches of the hybrid performance evaluation network to obtain K groups of local performance evaluation results;
[0089] A630: Based on the multiple hybrid test conditions, the K groups of local performance evaluation results are divided into the same conditions to obtain multiple global performance evaluation results, and the multiple global performance evaluation results constitute the hybrid performance evaluation results.
[0090] Specifically, in this embodiment, the hybrid test data set is divided according to the target associated indicator set to obtain K groups of hybrid operating condition test sequences, and then the K groups of hybrid operating condition test sequences are split based on multiple hybrid test conditions of the hybrid test sequence to obtain instantaneous K groups of hybrid operating condition test parameters when the hybrid test conditions are stable.
[0091] According to the mapping relationship of multiple hybrid test conditions, K groups of hybrid condition test parameters are divided into a first group of K hybrid condition test parameters to an Nth group of hybrid condition test parameters, and the first group to the Nth group correspond to the order of multiple hybrid test conditions in the hybrid test sequence.
[0092] The first group of K hybrid operating condition test parameters to the Nth group of hybrid operating condition test parameters are synchronized group by group to the K hybrid performance evaluation branches of the hybrid performance evaluation network to obtain K groups of local performance evaluation results, where the local performance evaluation results are evaluation results of a certain performance under a certain hybrid test condition.
[0093] Based on the multiple hybrid test conditions, the K groups of local performance evaluation results are divided into the same conditions to obtain multiple global performance evaluation results, each global performance evaluation result is composed of performance parameters corresponding to the target test performance set under the planned hybrid test conditions, and the multiple global performance evaluation results constitute the hybrid performance evaluation result.
[0094] This embodiment achieves the technical effect of improving the test efficiency of hybrid vehicle hybrid testing and improving the evaluation accuracy and evaluation credibility of hybrid vehicle performance evaluation based on hybrid testing.
[0095] Embodiment 2
[0096] Based on the same inventive concept as the hybrid test method for hybrid vehicles in the aforementioned embodiment, Figure 3 As shown, the present application provides a hybrid test system for a hybrid vehicle, wherein the system comprises:
[0097] A test performance interaction unit 1, used for interactively obtaining a target test performance set for a hybrid test of a vehicle to be tested;
[0098] The associated indicator calling unit 2 is used to pre-build a performance indicator mapping table, and perform associated indicator calling on the performance indicator mapping table with the target test performance set as a constraint to obtain a target associated indicator set;
[0099] An evaluation network construction unit 3 is used to interactively obtain multiple sample operating condition sets and multiple sample hybrid test sets of multiple sample vehicles of the same model as the vehicle to be tested, and to construct a hybrid performance evaluation network according to the multiple sample hybrid test sets, the target test performance set and the target associated index set;
[0100] A test sequence output unit 4 is used to generate a plurality of hybrid test conditions by performing condition aggregation on the plurality of sample condition sets, and to perform condition feature correlation analysis on the plurality of hybrid test conditions to obtain a hybrid test sequence;
[0101] A test data acquisition unit 5 is used to perform continuous hybrid tests on the vehicle to be tested based on the hybrid test sequence, and to collect data of a control process based on a vehicle sensor network of the vehicle to be tested, so as to generate a hybrid test data set, wherein the hybrid test data set includes a plurality of monitoring time series data of a plurality of monitoring indicators;
[0102] The evaluation result output unit 6 is used to divide the hybrid test data set according to the target associated indicator set, and synchronize the division result to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result.
[0103] In one embodiment, the test data obtaining unit 5 further includes:
[0104] The hybrid test performance in the target test performance set is mapped to the associated index in the target associated index set, wherein the target test performance set includes K hybrid test performances, and the target associated index set includes K groups of associated indexes;
[0105] Performing indicator aggregation on the K groups of related indicators to obtain multiple related indicators;
[0106] Using the multiple associated indicators as constraints, calling monitoring components on the multiple sample hybrid test sets to obtain multiple groups of sample monitoring components, wherein each sample hybrid test set in the multiple sample hybrid test sets includes multiple sample monitoring data and multiple sample hybrid performance data, and each sample monitoring data has a monitoring component identifier;
[0107] Performing associated component analysis according to the plurality of associated indicators and the plurality of groups of sample monitoring components to generate a test associated component set;
[0108] A component status detection is performed on the vehicle to be tested with reference to the test-associated component set, and if the status detection of the vehicle to be tested passes, a continuous hybrid test is performed on the vehicle to be tested based on the hybrid test sequence.
[0109] In one embodiment, the test data obtaining unit 5 further includes:
[0110] Aggregating the monitoring components of the multiple groups of sample monitoring components to obtain multiple monitoring components;
[0111] Constructing and generating a plurality of topological isolated vertices based on the plurality of associated indicators and the plurality of monitoring components;
[0112] Connecting the plurality of topologically isolated vertices according to the plurality of groups of sample monitoring components to generate an indicator component topological network;
[0113] Performing a topological connection frequency analysis on the indicator component topological network to obtain a plurality of association frequency sequences of the plurality of association indicators;
[0114] The multiple associated frequency sequences are called for extreme values to obtain multiple test associated components, and the multiple test associated components constitute the test associated component set.
[0115] In one embodiment, the test data obtaining unit 5 further includes:
[0116] Pre-constructing an indicator sensor mapping table, and using the indicator sensor mapping table as a constraint to perform associated sensor calls on the multiple associated indicators to obtain a target sensor call set;
[0117] Interactively obtain sensor composition information of the vehicle sensor network, and perform sensor missing analysis on the sensor composition information with the target sensor call set as a constraint to generate a sensor optimization strategy;
[0118] Based on the sensor optimization strategy and the test-related component set, the vehicle sensor network is equipped with sensors to generate an optimized sensor network;
[0119] Activating the optimized sensor network through sensor mapping based on the target sensor call set as a constraint to obtain a hybrid test sensor network;
[0120] The hybrid test data set is generated based on data collection of the control process performed by the dynamic test sensor network.
[0121] In one embodiment, the test sequence output unit 4 further includes:
[0122] Interactively obtaining a standard operating condition parameter threshold and an operating condition parameter deviation threshold, wherein the standard operating condition parameter threshold includes a plurality of operating condition index value thresholds, and the operating condition parameter deviation threshold includes a plurality of operating condition index deviation thresholds;
[0123] Using the standard operating condition parameter threshold to traverse the multiple sample operating condition sets to obtain M groups of non-standard operating condition parameters;
[0124] Aggregating the M groups of non-standard operating condition parameters using the operating condition parameter deviation threshold to obtain the multiple hybrid test operating conditions;
[0125] Obtaining a first hybrid test operating condition based on the multiple hybrid test operating conditions, and obtaining multiple operating condition similarity coefficients between the first hybrid test operating condition and the multiple hybrid test operating conditions based on Euclidean distance calculation;
[0126] The plurality of operating condition similarity coefficients are serialized, and the plurality of hybrid test operating conditions are sorted based on the serialization result to obtain the hybrid test sequence.
[0127] In one embodiment, the evaluation network construction unit 3 further includes:
[0128] Dividing the plurality of sample hybrid test sets according to the K hybrid test performances and the K groups of associated indicators to obtain K sample hybrid performance data sets and K sample monitoring data sets;
[0129] constructing a standard hybrid performance evaluation branch based on a convolutional neural network, and using the K sample hybrid performance data sets and the K sample monitoring data sets to train the standard hybrid performance evaluation branch, to obtain K hybrid performance evaluation branches;
[0130] The K hybrid performance evaluation branches are connected in parallel to construct and generate the hybrid performance evaluation network.
[0131] In one embodiment, the evaluation result output unit 6 further includes:
[0132] Dividing the hybrid test data set according to the target associated indicator set to obtain K groups of hybrid operating condition test sequences;
[0133] Synchronizing the K groups of hybrid operating condition test sequences to the K hybrid performance evaluation branches of the hybrid performance evaluation network to obtain K groups of local performance evaluation results;
[0134] The K groups of local performance evaluation results are divided into the same working condition based on the multiple hybrid test working conditions to obtain multiple global performance evaluation results, and the multiple global performance evaluation results constitute the hybrid performance evaluation result.
[0135] Any of the methods or steps described above may be stored as computer instructions or programs in various types of computer memories, and the computer instructions or programs may be recognized by various types of computer processors to implement any of the methods or steps described above.
[0136] Based on the above specific embodiments of the present invention, any improvements and modifications made to the present invention by those skilled in the art without departing from the principles of the present invention shall fall within the patent protection scope of the present invention.
Claims
1. A hybrid test method for a hybrid vehicle, characterized in that: The method comprises: Interactively obtain a target test performance set for hybrid testing of the vehicle to be tested; Pre-constructing a performance indicator mapping table, and performing associated indicator calls on the performance indicator mapping table with the target test performance set as a constraint to obtain a target associated indicator set; Interactively obtaining multiple sample operating condition sets and multiple sample hybrid test sets of multiple sample vehicles of the same model as the vehicle to be tested, and constructing a hybrid performance evaluation network according to the multiple sample hybrid test sets, the target test performance set, and the target associated index set; Generating a plurality of hybrid test conditions by performing condition aggregation on the plurality of sample condition sets, and performing condition feature correlation analysis on the plurality of hybrid test conditions to obtain a hybrid test sequence; Performing continuous hybrid testing on the vehicle to be tested based on the hybrid test sequence, and collecting data of a control process based on a vehicle sensor network of the vehicle to be tested to generate a hybrid test data set, wherein the hybrid test data set includes a plurality of monitoring time series data of a plurality of monitoring indicators; dividing the hybrid test data set according to the target associated indicator set, and synchronizing the division result to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result; Wherein, by performing condition aggregation on the multiple sample condition sets to generate multiple hybrid test conditions, performing condition feature correlation analysis on the multiple hybrid test conditions to obtain a hybrid test sequence, the method further includes: Interactively obtaining a standard operating condition parameter threshold and an operating condition parameter deviation threshold, wherein the standard operating condition parameter threshold includes a plurality of operating condition index value thresholds, and the operating condition parameter deviation threshold includes a plurality of operating condition index deviation thresholds; Using the standard operating condition parameter threshold to traverse the multiple sample operating condition sets to obtain M groups of non-standard operating condition parameters; Aggregating the M groups of non-standard operating condition parameters using the operating condition parameter deviation threshold to obtain the multiple hybrid test operating conditions; Obtaining a first hybrid test operating condition based on the multiple hybrid test operating conditions, and obtaining multiple operating condition similarity coefficients between the first hybrid test operating condition and the multiple hybrid test operating conditions based on Euclidean distance calculation; The plurality of operating condition similarity coefficients are serialized, and the plurality of hybrid test operating conditions are sorted based on the serialization result to obtain the hybrid test sequence.
2. The method according to claim 1, characterized in that Before performing a continuous hybrid test on the vehicle to be tested based on the hybrid test sequence, the method further includes: The hybrid test performance in the target test performance set is mapped to the associated index in the target associated index set, wherein the target test performance set includes K hybrid test performances, and the target associated index set includes K groups of associated indexes; Performing indicator aggregation on the K groups of related indicators to obtain multiple related indicators; Using the multiple associated indicators as constraints, calling monitoring components on the multiple sample hybrid test sets to obtain multiple groups of sample monitoring components, wherein each sample hybrid test set in the multiple sample hybrid test sets includes multiple sample monitoring data and multiple sample hybrid performance data, and each sample monitoring data has a monitoring component identifier; Performing associated component analysis according to the plurality of associated indicators and the plurality of groups of sample monitoring components to generate a test associated component set; A component status detection is performed on the vehicle to be tested with reference to the test-associated component set, and if the status detection of the vehicle to be tested passes, a continuous hybrid test is performed on the vehicle to be tested based on the hybrid test sequence.
3. The method according to claim 2, characterized in that Performing associated component analysis according to the multiple associated indicators and the multiple groups of sample monitoring components to generate a test associated component set, the method further includes: Aggregating the monitoring components of the multiple groups of sample monitoring components to obtain multiple monitoring components; Constructing and generating a plurality of topological isolated vertices based on the plurality of associated indicators and the plurality of monitoring components; Connecting the plurality of topologically isolated vertices according to the plurality of groups of sample monitoring components to generate an indicator component topological network; Performing a topological connection frequency analysis on the indicator component topological network to obtain a plurality of association frequency sequences of the plurality of association indicators; The multiple associated frequency sequences are called for extreme values to obtain multiple test associated components, and the multiple test associated components constitute the test associated component set.
4. The method according to claim 3, characterized in that Based on the hybrid test sequence, a continuous hybrid test is performed on the vehicle to be tested, and data collection of a control process is performed based on a vehicle sensor network of the vehicle to be tested to generate a hybrid test data set. The method further includes: Pre-constructing an indicator sensor mapping table, and using the indicator sensor mapping table as a constraint to perform associated sensor calls on the multiple associated indicators to obtain a target sensor call set; Interactively obtain sensor composition information of the vehicle sensor network, and perform sensor missing analysis on the sensor composition information with the target sensor call set as a constraint to generate a sensor optimization strategy; Based on the sensor optimization strategy and the test-related component set, the vehicle sensor network is equipped with sensors to generate an optimized sensor network; Activating the optimized sensor network through sensor mapping based on the target sensor call set as a constraint to obtain a hybrid test sensor network; The hybrid test data set is generated based on data collection of the control process performed by the dynamic test sensor network.
5. The method according to claim 2, characterized in that Interactively obtaining multiple sample operating condition sets and multiple sample hybrid test sets of multiple sample vehicles of the same model as the vehicle to be tested, and constructing a hybrid performance evaluation network according to the multiple sample hybrid test sets, the target test performance set, and the target associated index set. The method further includes: Dividing the plurality of sample hybrid test sets according to the K hybrid test performances and the K groups of associated indicators to obtain K sample hybrid performance data sets and K sample monitoring data sets; constructing a standard hybrid performance evaluation branch based on a convolutional neural network, and using the K sample hybrid performance data sets and the K sample monitoring data sets to train the standard hybrid performance evaluation branch, to obtain K hybrid performance evaluation branches; The K hybrid performance evaluation branches are connected in parallel to construct and generate the hybrid performance evaluation network.
6. The method according to claim 5, characterized in that Dividing the hybrid test data set according to the target associated indicator set, and synchronizing the division result to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result, the method further includes: Dividing the hybrid test data set according to the target associated indicator set to obtain K groups of hybrid operating condition test sequences; Synchronizing the K groups of hybrid operating condition test sequences to the K hybrid performance evaluation branches of the hybrid performance evaluation network to obtain K groups of local performance evaluation results; The K groups of local performance evaluation results are divided into the same working condition based on the multiple hybrid test working conditions to obtain multiple global performance evaluation results, and the multiple global performance evaluation results constitute the hybrid performance evaluation result.
7. A hybrid test system for a hybrid vehicle, characterized in that: The steps for implementing the method according to any one of claims 1 to 6 include: A test performance interaction unit, used to interactively obtain a target test performance set for hybrid testing of a vehicle to be tested; An associated indicator calling unit, used to pre-build a performance indicator mapping table, and perform associated indicator calling on the performance indicator mapping table with the target test performance set as a constraint to obtain a target associated indicator set; an evaluation network construction unit, configured to interactively obtain a plurality of sample operating condition sets and a plurality of sample hybrid test sets of a plurality of sample vehicles of the same model as the vehicle to be tested, and to construct a hybrid performance evaluation network according to the plurality of sample hybrid test sets, the target test performance set, and the target associated index set; A test sequence output unit, configured to generate a plurality of hybrid test conditions by performing condition aggregation on the plurality of sample condition sets, and perform condition feature correlation analysis on the plurality of hybrid test conditions to obtain a hybrid test sequence; a test data acquisition unit, configured to perform continuous hybrid tests on the vehicle to be tested based on the hybrid test sequence, and to collect data of a control process performed based on a vehicle sensor network of the vehicle to be tested, so as to generate a hybrid test data set, wherein the hybrid test data set includes a plurality of monitoring time series data of a plurality of monitoring indicators; An evaluation result output unit is used to divide the hybrid test data set according to the target associated indicator set, and synchronize the division result to the hybrid performance evaluation network for performance evaluation to generate a hybrid performance evaluation result.