An optimization method and device for improving vehicle fault diagnosis robustness

Abstract

The present invention relates to an optimization method and device for improving the robustness of vehicle fault diagnosis. The method includes: an ECU extracting fault monitoring values of the vehicle during a current driving cycle and presetting a limit range for each fault monitoring; judging the fault monitoring value against a fault threshold; determining that a fault has occurred when the monitoring value is greater than or less than the set fault threshold; and calculating the difference between the monitoring value and the fault threshold; calculating a ratio of the difference to the limit range, using the ratio as a basis for data statistics; calculating the mean and variance of the ratio, and using both as a basis for data statistics; and a diagnostic instrument using the calculated ratio, mean, and variance to correct the fault threshold, thereby optimizing the robustness of fault diagnosis. Compared with existing technologies, the present invention has the advantages of not requiring a data acquisition instrument, saving costs, time, and effort, and further enhancing diagnostic robustness.

Description

Technical Field

[0001] The present invention relates to the technical field of on-board diagnostic systems, and in particular to an optimization method and device for improving the robustness of vehicle fault diagnosis. Background Art

[0002] Diagnostic functions are essential for vehicle controllers. For DEC-ECUs, such as those in transmission controllers, compliance with OBD regulations is mandatory. Fault monitoring is a crucial component of diagnostics, accurately detecting system faults, operating in a degraded mode, and promptly alerting the driver through instrument panel warnings to ensure driving safety. In actual vehicle use, OBD systems are subject to a certain degree of false positives and missed faults due to factors such as component production consistency, the vehicle's operating environment, and operating conditions. Therefore, robustness is both a key and challenging aspect of fault monitoring. Avoiding misdiagnoses and missed diagnoses significantly benefits driving safety, customer experience, and after-sales maintenance.

[0003] Current robustness diagnostic methods typically require engineers to perform worst-case analysis to determine initial fault thresholds, but this is insufficient for engineering applications. Considering the interplay between the entire system and the operating environment after component integration and assembly, extensive road testing is required on a large number of test vehicles. After collecting relevant data, statistical processing is performed to refine the current fault diagnosis thresholds. If conditions permit, faulty components are selected and assembled on the entire vehicle, and similar data is collected and statistically processed to refine the fault diagnosis thresholds.

[0004] This operation can improve the robustness of diagnosis, but there are also many problems:

[0005] 1. Data statistics require a sufficient sample size, so data acquisition instruments need to be installed on a large number of test vehicles. This places a high demand on equipment and also requires high levels of operator skill.

[0006] 2. Test data needs to be regularly collected and processed by dedicated engineers. Test vehicles are difficult to control, and the processing of huge amounts of data is time-consuming and labor-intensive. Summary of the Invention

[0007] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide an optimization method and device for improving the robustness of vehicle fault diagnosis.

[0008] The purpose of the present invention can be achieved by the following technical solutions:

[0009] An optimization method for improving the robustness of vehicle fault diagnosis is implemented based on an ECU and includes:

[0010] The ECU extracts the fault monitoring enable status from the vehicle data. When the fault monitoring is enabled, it extracts the monitoring values of the vehicle's faults during the current driving cycle and pre-sets a limit range for each fault monitoring.

[0011] The fault monitoring value and the fault threshold are judged. When the monitoring value is greater than or less than the set fault threshold, it is determined that a fault has occurred, and the difference between the monitoring value and the fault threshold is calculated;

[0012] Calculate the ratio of the difference to the limit range and use this ratio as the basis for data statistics. The ratio is used to characterize the degree of proximity between the ECU's monitoring value and the fault threshold during operation. This value is an integer between 0 and 255. The ratio calculation includes a method for determining a fault when the monitoring value is less than the fault threshold and a method for calculating the case where no fault has occurred.

[0013] Calculate the mean and variance of the proportional value and use them as the basis for data statistics. The calculation of the proportional value and its mean and variance data are all centralized in a statistical data function module, which is pre-set in the ECU and contains rules for setting the proportional value, calculating the proportional value in actual use, and calculating the mean and variance.

[0014] The diagnostic instrument calls the calculated ratio value and its mean and variance to correct the fault threshold and optimize the robustness of fault diagnosis;

[0015] The calculation method for the ratio of the monitoring value being determined as a fault when it is less than the fault threshold to the value when no fault occurs is:

[0016]

[0017] Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmin is the minimum value of the limit range.

[0018] The ratio value calculation also includes a calculation method for determining a fault when the monitoring value is greater than the fault threshold value and a fault occurrence situation. The ratio value calculation method for determining a fault when the monitoring value is greater than the fault threshold value and a fault occurrence situation is:

[0019]

[0020] Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmax is the maximum value of the limit range.

[0021] The ratio value calculation also includes a calculation method for the case where the monitoring value is greater than the fault threshold and the fault is determined to have not occurred. The ratio value calculation method for the case where the monitoring value is greater than the fault threshold and the fault is determined to have not occurred is:

[0022]

[0023] Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmax is the maximum value of the limit range.

[0024] The ratio value calculation also includes a calculation method for determining a fault when the monitoring value is less than the fault threshold value and a fault occurrence situation. The ratio value calculation method for determining a fault when the monitoring value is less than the fault threshold value and a fault occurrence situation is:

[0025]

[0026] Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmin is the minimum value of the limit range.

[0027] The present invention also provides an optimization device for improving the robustness of vehicle fault diagnosis, including an ECU and a data acquisition device. The ECU is provided with a statistical analysis tool, and the statistical analysis tool is added with a statistical data function module. The statistical data function module includes:

[0028] A proportional value calculation unit is used to calculate a proportional value representing the degree of proximity between the monitoring value and the fault threshold during operation of the ECU based on the fault monitoring value, the fault threshold and a set limit range;

[0029] The statistical data function module is specifically used for:

[0030] Extracting the fault monitoring enablement status from vehicle data, extracting the fault monitoring values of the vehicle during the current driving cycle when the fault monitoring is enabled, and pre-setting a limit range for each fault monitoring;

[0031] The fault monitoring value and the fault threshold are judged. When the monitoring value is greater than or less than the set fault threshold, it is determined that a fault has occurred, and the difference between the monitoring value and the fault threshold is calculated;

[0032] receiving a proportional value calculated by the proportional value calculation unit and using the proportional value as a basis for data statistics, wherein the proportional value is used to characterize the degree of proximity between the monitoring value of the ECU during operation and the fault threshold, and the value is an integer between 0 and 255, and the proportional value calculation includes a method for determining a fault when the monitoring value is less than the fault threshold and a method for calculating a case where no fault has occurred;

[0033] Calculate the mean and variance of the proportional value and use them as the basis for data statistics. The calculation of the proportional value and its mean and variance data are all centralized in a statistical data function module, which is pre-set in the ECU and contains rules for setting the proportional value, calculating the proportional value in actual use, and calculating the mean and variance.

[0034] The diagnostic instrument calls the calculated ratio value and its mean and variance to correct the fault threshold and optimize the robustness of fault diagnosis;

[0035] The calculation method for the ratio of the monitoring value being determined as a fault when it is less than the fault threshold to the value when no fault occurs is:

[0036]

[0037] Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmin is the minimum value of the limit range.

[0038] The statistical data function module includes:

[0039] The statistical data calculation unit is used to calculate the average value and variance of the proportion value obtained by the proportion value calculation unit, store the calculation result in NVRAM, and associate the proportion value, average value and variance data with the DID.

[0040] The optimization method and device for improving the robustness of vehicle fault diagnosis provided by the present invention have at least the following beneficial effects compared to the prior art:

[0041] 1) By adding the Ranking Logger function to the ECU, the controller can automatically collect and store data while the test vehicle is running. When collecting real vehicle test data, there is no need to equip a data acquisition instrument, saving costs, time and effort;

[0042] 2) When using the present invention for fault diagnosis, there is no need for engineers to extract useful information from complex data; by adding a data statistics function (Ranking logger function) in the ECU, which is based on the entire vehicle test data, during diagnosis, it is only necessary to call the statistical data obtained by this function to analyze the operation of the diagnostic monitoring, and then correct the fault threshold based on the analysis results. This makes it very convenient to collect test data, and R&D engineers are more efficient when performing data analysis, which can further enhance the robustness of diagnosis.

[0043] 3) The statistically processed data can be read through the diagnostic instrument, without spending a lot of energy on data statistics;

[0044] 4) Based on the performance of statistical data, relevant parameters can be adjusted to achieve the purpose of accurately correcting the fault threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 Schematic diagram of a flow chart of an optimization method for improving vehicle fault diagnosis robustness in an embodiment;

[0046] Figure 2Graph showing the relationship between TV, RV, FT, and limit ranges when a fault is determined when the monitoring value is less than the fault threshold and when no fault occurs in the embodiment;

[0047] Figure 3 Graph showing the relationship between TV, RV, FT, and limit ranges when a fault is determined when the monitoring value is greater than the fault threshold in the embodiment and when a fault occurs;

[0048] Figure 4 Graph showing the relationship between TV, RV, FT, and the limit range when a fault is determined when the monitoring value is greater than the fault threshold and when no fault occurs in the embodiment;

[0049] Figure 5 Graph showing the relationship between TV, RV, FT, and the limit range when a fault is determined when the monitoring value is less than the fault threshold in the embodiment and when a fault occurs;

[0050] Figure 6 2 is a functional principle diagram of the Ranking logger functional module in the embodiment. DETAILED DESCRIPTION

[0051] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments described are only a portion of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0052] Example

[0053] The present invention relates to an optimization method for improving the robustness of vehicle fault diagnosis. The method adds a ranking logger function to the ECU. When a test vehicle is running, the controller can automatically perform data statistics and store the data. There is no need for a dedicated data acquisition instrument, nor is there a need for engineers to extract useful information from the complex data. Diagnostic engineers only need to retrieve the statistical data in the ranking logger to analyze the operation of diagnostic monitoring, and then modify the fault threshold based on the analysis results to enhance diagnostic robustness.

[0054] The method of the present invention is implemented based on ECU, such as Figure 1 As shown, the specific implementation process includes the following steps:

[0055] Step 1: During the fault monitoring process, the ECU's fault monitoring strategy calculates the monitoring value and first extracts the monitoring value of the fault during the vehicle's current driving cycle.

[0056] Step 2: Determine the fault monitoring value and the fault threshold (FT). When the monitoring value is greater than or less than the set fault threshold (FT), it is considered that a fault has occurred; proceed to the next step.

[0057] Step 3: To unify different fault monitoring, a limit range (Rmax: Range Max, Rmin: Range Min) is pre-set for each fault monitoring. The difference between the monitoring value (TV: Test Value) and the fault threshold is calculated to obtain a ratio value (RV: Ranking Value) relative to the limit range for data statistics.

[0058] Ranking value is an integer between 0 and 255, which is used to indicate the degree to which the monitoring value of the controller is close to the fault threshold during operation.

[0059] Ranking Value algorithms vary for different fault monitoring types, including the following:

[0060] 1) When the monitoring value is less than the fault threshold, it is determined to be a fault - no fault has occurred. In this case, the RV calculation formula is as follows:

[0061]

[0062] The relationship between TV, RV, FT and limit range in this case is as follows Figure 2 shown.

[0063] 2) When the monitoring value is greater than the fault threshold, it is determined that a fault-fault has occurred. In this case, the RV is calculated as follows:

[0064]

[0065] The relationship between TV, RV, FT and limit range in this case is as follows Figure 3 shown.

[0066] 3) When the monitoring value is greater than the fault threshold, it is determined to be a fault - no fault has occurred. In this case, the RV is calculated as follows:

[0067]

[0068] The relationship between TV, RV, FT and limit range in this case is as follows Figure 4 shown.

[0069] 4) When the monitoring value is less than the fault threshold, it is determined that a fault has occurred. In this case, the RV is calculated as follows:

[0070]

[0071] The relationship between TV, RV, FT and limit range in this case is as follows Figure 5 shown.

[0072] Step 4: In traditional data statistical processing, the monitoring value signal is normally distributed and the fault threshold is defined according to the Six Sigma principle. Similarly, the present invention also needs to calculate the mean and variance of the Ranking Value for subsequent analysis.

[0073] The formula for calculating the average value is as follows:

[0074]

[0075] The formula for calculating variance is as follows:

[0076]

[0077] Where n is the number of driving cycles, It is the maximum value of RV in this driving cycle.

[0078] Each fault monitoring corresponds to a Ranking Value. After data statistical processing, its average value and variance are obtained and stored in NVRAM.

[0079] Step 5: Use the calculated Ranking Value, mean, and variance data to optimize robustness diagnosis. Diagnostic robustness involves the probabilistic assessment of two risks: α risk, false fault detection; and β risk, missed fault detection.

[0080] In this embodiment, the calculation of Ranking Value, average value and variance data are all concentrated in a statistical data function module (hereinafter described as Ranking logger function module), which is pre-set in the ECU. The module is provided with the setting rules of Ranking value and the calculation rules of Ranking value in actual use, as well as the calculation rules for average value and variance.

[0081] Typically, the same monitoring statistics correspond to a DID (Data Identifier). After the software stores the statistics, it associates the data with the DID, making it easier for engineers to read the data through diagnostic instruments. Engineers then use the statistics to calibrate fault thresholds. Diagnostic instruments can directly read the DID through the UDS service (Unified Diagnostic Services) to obtain the corresponding Ranking Value, average, and variance data. The calculated average and variance data can be used to obtain probability distribution and other analyses. Based on the analysis results, fault thresholds can then be modified to enhance diagnostic robustness.

[0082] Ranking logger statistics based on real-vehicle testing are crucial for threshold correction, improving diagnostic robustness by mitigating alpha risk. Six Sigma principles allow for the calculation of statistically significant Ranking Values. If these values exceed 128, the thresholds are set too narrow and require adjustment. Additionally, the Rmin and Rmax ranges can be adjusted to improve Ranking Value resolution, achieving more precise fault threshold correction.

[0083] This embodiment also provides an optimization device for improving the robustness of vehicle fault diagnosis, including an ECU and a data acquisition device. The ECU is provided with a statistical analysis tool, and the statistical analysis tool is added with a statistical data function module (hereinafter described as a Ranking logger function module).

[0084] Among them, the statistical analysis tool is a software tool. After importing the data file recorded by the data acquisition device, the software automatically extracts fault-related signals, synchronizes the time, resamples, and statistically analyzes the signals, and displays the analysis results in graphics and reports.

[0085] For healthy vehicles, the statistical result of the test data is the false alarm rate. For vehicles with faulty parts installed, the statistical result of the test data is the missed alarm rate. The OBD robustness level of the vehicle is evaluated based on the false alarm rate and missed alarm rate.

[0086] Among them, the Ranking logger function module is based on statistical analysis tools. In this embodiment, the Ranking value function module is pre-set in the ECU. The module is equipped with the setting rules of the Ranking value and the calculation rules of the Ranking value in actual use. Figure 6 As shown, the functional module includes:

[0087] A ranking value calculation unit is used to calculate a ranking value that represents the degree of proximity between the ECU's monitoring value and the fault threshold during operation based on the fault monitoring value, the fault threshold, and a set limit range;

[0088] The statistical data calculation unit is used to calculate the average value and variance of the Ranking Value value obtained by the Ranking Value calculation unit, and store the calculation results in NVRAM; associate each calculation result data with the DID, and provide a calling basis for the diagnostic instrument by corresponding each statistical data of the same monitoring to the same DID.

[0089] By adding a Ranking logger function to the ECU, the controller of the test vehicle can automatically perform data statistics and store data when the test vehicle is running. When collecting actual vehicle test data, there is no need to equip a data acquisition instrument, which saves costs, time and effort. When using the present invention for fault diagnosis, there is no need for engineers to extract useful information from complex data. By adding a data statistics function to the ECU, during diagnosis, it is only necessary to call up the statistical data in the Ranking logger to analyze the operation of the diagnostic monitoring, and then correct the fault threshold based on the analysis result. Collecting test data is very convenient, and R&D engineers are more efficient when performing data analysis, which can further enhance the robustness of diagnosis.

[0090] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and such modifications or substitutions are intended to be within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. An optimization method for improving the robustness of vehicle fault diagnosis, characterized in that: The method is implemented based on ECU and includes: The ECU extracts the fault monitoring enable status from the vehicle data. When the fault monitoring is enabled, it extracts the monitoring values of the vehicle's faults during the current driving cycle and pre-sets a limit range for each fault monitoring. The fault monitoring value and the fault threshold are judged. When the monitoring value is greater than or less than the set fault threshold, it is determined that a fault has occurred, and the difference between the monitoring value and the fault threshold is calculated; Calculate the ratio of the difference to the limit range and use this ratio as the basis for data statistics. The ratio is used to characterize the degree of proximity between the ECU's monitoring value and the fault threshold during operation. This value is an integer between 0 and 255. The ratio calculation includes a method for determining a fault when the monitoring value is less than the fault threshold and a method for calculating the case where no fault has occurred. Calculate the mean and variance of the proportional value and use them as the basis for data statistics. The calculation of the proportional value and its mean and variance data are all centralized in a statistical data function module, which is pre-set in the ECU and contains rules for setting the proportional value, calculating the proportional value in actual use, and calculating the mean and variance. The diagnostic instrument calls the calculated ratio value and its mean and variance to correct the fault threshold and optimize the robustness of fault diagnosis; The calculation method for the ratio of the monitoring value being determined as a fault when it is less than the fault threshold to the value when no fault occurs is: Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmin is the minimum value of the limit range.

2. The optimization method for improving vehicle fault diagnosis robustness according to claim 1, characterized in that: The ratio value calculation also includes a calculation method for determining a fault when the monitoring value is greater than the fault threshold value and a fault occurrence situation. The ratio value calculation method for determining a fault when the monitoring value is greater than the fault threshold value and a fault occurrence situation is: Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmax is the maximum value of the limit range.

3. The optimization method for improving vehicle fault diagnosis robustness according to claim 1, characterized in that: The ratio value calculation also includes a calculation method for the case where the monitoring value is greater than the fault threshold and the fault is determined to have not occurred. The ratio value calculation method for the case where the monitoring value is greater than the fault threshold and the fault is determined to have not occurred is: Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmax is the maximum value of the limit range.

4. The optimization method for improving vehicle fault diagnosis robustness according to claim 1, characterized in that: The ratio value calculation also includes a calculation method for determining a fault when the monitoring value is less than the fault threshold value and a fault occurrence situation. The ratio value calculation method for determining a fault when the monitoring value is less than the fault threshold value and a fault occurrence situation is: Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmin is the minimum value of the limit range.

5. An optimization device for improving the robustness of vehicle fault diagnosis, characterized in that: The system includes an ECU and a data acquisition device. The ECU is provided with a statistical analysis tool. The statistical analysis tool is provided with a statistical data function module. The statistical data function module includes: A proportional value calculation unit is used to calculate a proportional value representing the degree of proximity between the monitoring value and the fault threshold during operation of the ECU based on the fault monitoring value, the fault threshold and a set limit range; The statistical data function module is specifically used for: Extracting the fault monitoring enablement status from vehicle data, extracting the fault monitoring values of the vehicle during the current driving cycle when the fault monitoring is enabled, and pre-setting a limit range for each fault monitoring; The fault monitoring value and the fault threshold are judged. When the monitoring value is greater than or less than the set fault threshold, it is determined that a fault has occurred, and the difference between the monitoring value and the fault threshold is calculated; receiving a proportional value calculated by the proportional value calculation unit and using the proportional value as a basis for data statistics, wherein the proportional value is used to characterize the degree of proximity between the monitoring value of the ECU during operation and the fault threshold, and the value is an integer between 0 and 255, and the proportional value calculation includes a method for determining a fault when the monitoring value is less than the fault threshold and a method for calculating a case where no fault has occurred; Calculate the mean and variance of the proportional value and use them as the basis for data statistics. The calculation of the proportional value and its mean and variance data are all centralized in a statistical data function module, which is pre-set in the ECU and contains rules for setting the proportional value, calculating the proportional value in actual use, and calculating the mean and variance. The diagnostic instrument calls the calculated ratio value and its mean and variance to correct the fault threshold and optimize the robustness of fault diagnosis; The calculation method for the ratio of the monitoring value being determined as a fault when it is less than the fault threshold to the value when no fault occurs is: Where RV is the proportional value, TV is the fault monitoring value, FT is the fault threshold, and Rmin is the minimum value of the limit range.

6. The optimization device for improving vehicle fault diagnosis robustness according to claim 5, characterized in that: The statistical data function module includes: The statistical data calculation unit is used to calculate the average value and variance of the proportion value obtained by the proportion value calculation unit, store the calculation result in NVRAM, and associate the proportion value, average value and variance data with the DID.

Images