Fault self-checking method and system for circulating ball type electric power steering system and readable medium

Abstract

The invention discloses a recirculating ball type electric power steering system fault self-checking method and system and a readable medium, and relates to the technical field of vehicles, and the method comprises the steps: storing a signal transmission path and each function node, determining the type of an influence factor which influences the instruction execution accuracy of the function node, and determining the instruction execution accuracy of the function node; analyzing and acquiring the relationship between the fluctuation amplitude of the data signal at each function node and the numerical value of the influence factor, storing the relationship as a correction model, and storing a first relationship model for reflecting the data association relationship among the function nodes on the same signal transmission path; acquiring real-time data and source data of various impact factors, and generating theoretical response data according to the first relation model, the real-time impact factor data and the correction model; the actual response data is collected, the deviation amplitude between the actual response data and the theoretical response data is calculated, whether the function node breaks down or has a potential fault is judged according to the deviation amplitude, early warning is output, and the fault detection accuracy and the potential fault pre-judgment capability are greatly improved through the technical scheme.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, specifically to a self-diagnosis method, system, and readable medium for a recirculating ball type electric power steering system. Background Technology

[0002] A recirculating ball steering gear mainly consists of a screw, a nut, a steering gear housing, and numerous small steel balls. When the screw, connected to the reduction mechanism, rotates, it pushes the nut along the screw's axial direction with the help of the steel balls. During this movement, the nut drives the sector gear to rotate, which in turn drives the steering rocker arm to reciprocate, ultimately achieving steering. Compared to rack and pinion steering gears, recirculating ball steering gears rely more on rolling friction, resulting in higher transmission efficiency and easier, more comfortable operation. They are typically found in off-road vehicles or medium-to-large commercial vehicles.

[0003] To make steering control easier and more precise for drivers, improving driving comfort, safety, and energy efficiency, most modern cars are equipped with Electronic Power Steering (EPS) systems. Combined with the aforementioned recirculating ball steering gear, EPS uses a torque sensor to detect the torque output from the steering wheel during operation. This torque signal is then transmitted to the Electronic Control Unit (ECU). The ECU integrates a control algorithm that outputs a PWM speed control signal to the power steering motor based on vehicle speed, engine pulses, and other signals. The power steering motor's shaft is connected to a reduction gear via an electromagnetic clutch, ultimately delivering electric power assistance to the screw of the recirculating ball steering gear, thus providing steering assistance.

[0004] To ensure the stable and safe operation of EPS (Electric Power Steering), current automotive infotainment systems and the EPS itself are equipped with fault detection systems. Specifically, these systems collect and monitor data such as sensor data, motor operating parameters, ECU status, and communication signals. This data is used to diagnose the EPS's operating status in real time, ensuring the safety and reliability of power steering. The data monitored by the EPS fault detection system can be divided into two categories: explicit data and implicit data. Explicit data refers to data that can be directly collected, such as the temperature of the power steering motor and the power supply voltage of the ECU. To determine whether this data is abnormal, it is only necessary to compare it with a set safety threshold. Implicit data refers to data obtained indirectly, such as the response time of power steering and the accuracy of control command execution. To determine whether this data is abnormal, at least two sets of data need to be analyzed. For example, to determine whether the power steering response time is abnormal, it is necessary to obtain the time point when the steering wheel rotates (command output time point) and the time point when the steering screw inside the recirculating ball steering gear rotates (command execution time point). The power steering response time is then obtained by subtracting these two time points.

[0005] When designing the EPS fault self-diagnosis system, the inventors discovered that existing EPS fault detection systems, when detecting faults related to abnormal command execution, such as excessively long power steering response time or uneven power steering, often only output a result data indicating the presence of an anomaly, without accurately pinpointing the specific cause. For example, the fault self-diagnosis system can detect abnormal steering response time, but the causes of such anomalies could include signal transmission delay, ECU processing delay, motor output delay, or even jamming of the recirculating ball steering transmission components. Similarly, the fault self-diagnosis system can detect abnormal power steering, but the causes could include a power steering motor malfunction, an ECU signal processing malfunction, or a malfunction of the torque sensor inside the recirculating ball steering system. Clearly, relying solely on the result data is insufficient to accurately pinpoint the cause of the fault, hindering troubleshooting and preventing a precise assessment of subsequent driving safety. Furthermore, during EPS fault monitoring, different types of monitoring data are often affected by various factors. For example, loose signal transmission terminals and external electromagnetic radiation can cause short-term fluctuations in signal transmission. How to determine whether the above anomalies are caused by environmental interference or system (potential) faults is also a problem that existing EPS fault detection systems have not addressed.

[0006] Clearly, when designing an EPS fault self-diagnosis system, the key technical problem that urgently needs to be solved is how to enable the fault self-diagnosis system to accurately locate the fault node for instruction execution anomalies, and at the same time make accurate and reliable judgments or predictions on various fault information. Summary of the Invention

[0007] To address the problem that EPS fault detection systems in practical applications struggle to accurately pinpoint fault nodes and reliably and accurately determine fault information, including intermittent faults, this application aims to provide a self-diagnosis method for recirculating ball electric power steering systems. This method can accurately and reliably locate fault nodes and accurately predict potential, intermittent faults, thereby improving the accuracy and reliability of EPS fault self-diagnosis. To achieve the aforementioned self-diagnosis method for recirculating ball electric power steering systems, this application also aims to provide a self-diagnosis system for recirculating ball electric power steering systems. This system, through an algorithm module configured within the ECU combined with relevant hardware modules, can significantly improve the accuracy and reliability of EPS fault self-diagnosis. Finally, this application also proposes protection for a computer-readable storage medium. The specific solution is as follows: A self-diagnostic method for a recirculating ball type electric power steering system includes: Set and store the signal transmission path corresponding to each source data type and the functional nodes contained in the signal transmission path, and number and store the above source data types, signal transmission paths and functional nodes together. Identify the types of influencing factors affecting the stability of data signal transmission between functional nodes and the accuracy of instruction execution at each functional node. Analyze the correlation between the signal transmission path and the magnitude of data signal fluctuations at each functional node and the corresponding influencing factor values, and store this information as a correction model. Establish and store a first relational model to reflect the data correspondence between functional nodes on the same signal transmission path, and a second relational model to reflect the data correspondence between functional nodes on the signal transmission path and specific reference nodes on different signal transmission paths; Acquire real-time data on various impact factors and store them temporarily; Theoretical response data at each functional node on the signal transmission path is generated by acquiring and combining the source data with the first relationship model, real-time impact factor data, and the correction model. Collect the actual response data at each functional node along the signal transmission path and calculate the deviation between the actual response data and the corresponding theoretical response data; Locate and identify functional nodes whose deviation exceeds a set value, mark these functional nodes as suspected fault nodes, and store them in association with the functional node number and data sampling time. After a set time interval, output the source data for verification, collect the actual response data at suspected fault nodes and compare it with the corresponding theoretical response data: If the deviation exceeds the set value, the first suspected fault node on the signal transmission path will be marked as a fault node and an alarm will be output. If the deviation does not exceed the set value, the data information at the specific reference node associated with the first suspected fault node on the signal transmission path is obtained as reference data. Analyze the correspondence between actual response data and reference data: If the second relation model is satisfied, the suspected fault node is marked as a potential fault node and an early warning is output; If the second relation model is not satisfied, then remove the suspected fault mark from the suspected fault node; The signal transmission path includes the forward transmission path from the source data output node to the instruction execution node, and the feedback transmission path from the functional node where each monitoring data is collected to the signal processing node. The types of influencing factors include vehicle driving status parameters, road condition parameters, environmental parameters, and driving habit data.

[0008] Through the above technical solution, when abnormalities occur in the execution of commands during EPS operation, such as excessively long power assist response delays or uneven power assist, the fault detection system can accurately locate the specific fault node, facilitating subsequent fault troubleshooting and accurately assessing the impact of the current fault on driving safety. Simultaneously, the fault detection system can distinguish potential faults and interference data from intermittent abnormal data, greatly improving the accuracy of fault detection and the ability to predict potential faults.

[0009] Furthermore, identifying the types of influencing factors includes: Select potential influencing factors and source data for testing, and set up the testing process; The EPS simulation equipment was placed in the test environment chamber to obtain the response data corresponding to the source data at each functional node on the signal transmission path under the conditions of various influencing factors or their combinations. The types of influencing factors affecting the accuracy of response data were determined based on correlation analysis. Analysis yields a corrected model, including: Based on the identified types of influencing factors, the numerical values ​​of these factors in the environmental chamber are varied to obtain the fluctuation range of the output response data of each functional node under different influencing factor values. The correlation between the magnitude of data signal fluctuations at each functional node and the magnitude of the corresponding influencing factor values ​​is obtained by fitting the data using a linear regression algorithm and stored as the corrected model.

[0010] The above technical solution can obtain the response accuracy of each functional node to different source data under various influencing factors, which is conducive to the fault detection system accurately eliminating interference data signals when making fault judgment and improving the accuracy of fault detection.

[0011] Furthermore, the fault self-diagnosis method also includes: Establish and store the correspondence between the data signal fluctuation amplitude and the sampling frequency at each functional node; Based on the real-time data of the influencing factors, the fluctuation amplitude of the data signal at each functional node is obtained by combining the modified model. Select a functional node whose fluctuation amplitude exceeds the set value, and adjust the sampling frequency of the data signal at that functional node according to the above correspondence between fluctuation amplitude and sampling frequency.

[0012] The above technical solution allows for adjusting the sampling frequency of data signals at each functional node based on the magnitude of the influence of the influencing factors, thereby reducing or avoiding the impact of interference signals on fault determination results and improving the accuracy of fault detection.

[0013] Furthermore, the fault detection method also includes: Acquire and store the actual response data and sampling time points of each functional node to each source of data of the same type; The numerical values ​​of the source data and their corresponding actual response data are normalized and stored together. Cluster analysis was performed on the above actual response data to obtain the abnormal data and their corresponding sampling time points; Obtain reference data collected at the above sampling time points for specific reference nodes associated with the above functional nodes, and analyze and compare the correspondence between the above abnormal data and the reference data: If the second relationship model is satisfied, the above abnormal data will be marked as potential fault data and stored in association with the functional node number and sampling time point; If the second relational model is not satisfied, the above abnormal data will be marked as interference data.

[0014] The above technical solution allows for the periodic analysis of the collected and stored source data and its actual response data, identifying and marking potential fault data. This helps to discover potential fault nodes in EPS, make accurate predictions about faults, and improve the accuracy and timeliness of fault detection.

[0015] Furthermore, the fault self-diagnosis method also includes: Count the number of times each functional node is marked as a faulty node or potential faulty node, and calculate the fault marking rate of each functional node; Configure different data signal sampling frequencies for each functional node; The data signal sampling frequency is set in direct proportion to the fault marking rate.

[0016] Through the above technical solution, the higher the fault marking rate of the functional node, the higher the sampling frequency of the corresponding data signal, which avoids the omission of potential fault data information due to data loss or low sampling frequency, and improves the accuracy of fault detection.

[0017] Furthermore, the fault self-diagnosis method also includes: Establish the correlation between the accuracy of instruction execution at each functional node and the performance value of driving safety; Based on the above relationships, configure security impact weights for each functional node; The safety status value of each functional node is generated based on the magnitude of the deviation between the actual response data and the corresponding theoretical response data at each functional node. The security judgment value is obtained by weighted calculation based on the security status value of each functional node and its corresponding security impact weight; The weighted judgment value is compared with the set driving safety threshold, and driving safety alarms of different levels are output according to the difference between the weighted judgment value and the driving safety threshold.

[0018] By using the above technical solution, the deviation of data signals at each functional node is correlated with driving safety. When multiple functional nodes or functional nodes with large safety impact weight values ​​fail, driving safety warnings can be output in a timely manner, prompting the driver to drive cautiously or park to troubleshoot the fault, which helps to improve driving safety.

[0019] To implement the above-mentioned self-diagnosis method for recirculating ball electric power steering systems, this application also proposes a self-diagnosis system for recirculating ball electric power steering systems, comprising: The data storage unit is configured to store the signal transmission paths corresponding to each source data type and the functional nodes contained in the signal transmission paths, a correction model to reflect the correlation between the magnitude of data signal fluctuations on each signal transmission path / functional node and the magnitude of the corresponding influencing factor values, a first relation model to reflect the data correspondence between functional nodes on the same signal transmission path, a second relation model to reflect the data correspondence between functional nodes on the signal transmission path and specific reference nodes on different signal transmission paths, the acquired various influencing factor data, and the fault markers corresponding to each functional node. The impact factor data acquisition unit is configured to acquire various impact factor data in real time. The source data monitoring unit is configured to monitor and acquire source data in real time. The theoretical response data generation unit is configured to generate theoretical response data at each functional node on the signal transmission path based on the source data, combined with the first relationship model, real-time influence factor data, and the correction model. The actual response data acquisition unit is configured to acquire actual response data at each functional node along the signal transmission path based on a set sampling frequency. The deviation calculation unit is configured to receive and calculate the deviation between the actual response data and the corresponding theoretical response data at each functional node. The marking verification unit is connected to the deviation calculation unit and configured to find and determine functional nodes whose deviation exceeds a set value. The functional nodes are marked as suspected fault nodes. Then, the source data for verification is output at set intervals. The actual response data at the suspected fault node is obtained and compared with the corresponding theoretical response data. The verification result of the deviation of the actual response data and the set value is output. The fault alarm output unit is connected to the marking verification unit to obtain the verification result. If the above deviation exceeds the set value, the first suspected fault node on the signal transmission path is marked as a fault node and an alarm is output. The potential fault confirmation unit receives the verification result. If the deviation does not exceed a set value, it acquires data information from a specific reference node associated with the first suspected fault node on the signal transmission path as reference data and analyzes the correspondence between the actual response data and the reference data. If the second relation model is satisfied, the suspected fault node is marked as a potential fault node and an early warning is output; If the second relational model is not satisfied, then the suspected fault mark of the suspected fault node is removed.

[0020] The types of influencing factors include vehicle driving status parameters, road condition parameters, environmental parameters, and driving habit data.

[0021] Furthermore, the data storage unit also stores the correspondence between the data signal fluctuation amplitude and the sampling frequency at each functional node; The system also includes: The signal fluctuation amplitude calculation unit is configured to acquire and, based on real-time data of influencing factors, combine the correction model to obtain the data signal fluctuation amplitude at each functional node. The sampling frequency configuration unit is configured to select a functional node whose fluctuation amplitude exceeds a set value, adjust the sampling frequency of the data signal at the functional node according to the above correspondence between the fluctuation amplitude and the sampling frequency, and output it to the actual response data acquisition unit.

[0022] The above technical solution allows for focused data collection on functional nodes with large fluctuations in data signals, preventing data omissions and ensuring the accuracy of fault detection.

[0023] Furthermore, the data storage unit also stores the correlation between the accuracy of instruction execution at each functional node and the driving safety performance value; The system also includes a driving safety alarm module, comprising: The safety impact weight configuration unit configures safety impact weights for each functional node based on the correlation between the above-mentioned instruction execution accuracy and driving safety characterization value. The safety status value assignment unit is configured to generate the corresponding safety status value for each functional node based on the magnitude of the deviation between the actual response data and the corresponding theoretical response data at each functional node. The driving safety judgment value generation unit is configured to obtain and calculate the safety judgment value by weighting the safety status value and its corresponding safety impact weight at each functional node. The driving safety alarm output unit is configured to compare the weighted judgment value with the set driving safety threshold, and output driving safety alarms of different levels according to the difference between the weighted judgment value and the driving safety threshold.

[0024] With the above technical solution, when a critical node on the EPS signal transmission path fails, a driving safety alarm can be output in a timely manner to ensure driving safety.

[0025] A computer-readable storage medium having a program module loaded thereon, which, when executed by a processor, is used to implement the aforementioned self-diagnosis method for a recirculating ball electric power steering system fault.

[0026] The above technical solution will help promote the application of the EPS fault self-testing method in this application.

[0027] This application includes at least one of the following beneficial effects: (1) By introducing a correction model, the data signals at each functional node are corrected under the influence of different influencing factors, eliminating the interference of environment or vehicle condition on EPS fault detection and improving the accuracy of fault detection. (2) By detecting and marking each functional node along the signal transmission path, the specific fault node can be located directly when a fault occurs, which helps to eliminate the fault later and make an accurate assessment of driving safety. (3) The fault detection system can distinguish potential faults and interference data from occasional abnormal data, which greatly improves the accuracy of fault detection and the ability to predict potential faults. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the main functional modules and their connections in an electric power steering (EPS) system. Figure 2 This is a schematic diagram of the fault self-testing method of this application; Figure 3 This is a schematic diagram of a signal transmission path; Figure 4 A schematic diagram illustrating the steps involved in determining the type of impact factor; Figure 5 This is a schematic diagram showing the functional module connections of the EPS fault self-test system.

[0029] Figure reference numerals: 1. Data storage unit; 2. Influence factor data acquisition unit; 3. Source data monitoring unit; 4. Theoretical response data generation unit; 5. Actual response data acquisition unit; 6. Deviation calculation unit; 7. Marking verification unit; 8. Fault alarm output unit; 9. Potential fault confirmation unit; 10. Signal fluctuation amplitude calculation unit; 11. Sampling frequency configuration unit; 12. Driving safety alarm module; 13. Safety influence weight configuration unit; 14. Safety status value assignment unit; 15. Driving safety judgment value generation unit; 16. Driving safety alarm output unit; 17. Steering wheel; 18. Torque sensor; 19. Electronic control unit; 20. Power steering motor reduction mechanism; 21. Recirculating ball steering gear. Detailed Implementation

[0030] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0031] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0032] Before providing a detailed explanation of the self-diagnosis method for the recirculating ball electric power steering system described in this application, it should be noted that the order of the method steps disclosed in this application is only a preferred embodiment. In specific applications, some steps can be selectively implemented as needed. For example, if a correction model already exists, the correction model generation step does not need to be repeated. Similarly, the implementation order of each step can be adjusted as needed.

[0033] like Figure 1 The diagram shows the main functional modules and their connections in the prior art electric power steering (EPS) system, including the connection diagram of torque sensor 18, electronic control unit 19 (ECU), power steering motor, etc. To facilitate the explanation of the fault self-diagnosis method described in this application, the relevant examples will be based on the functional architecture of the above-mentioned electric power steering system.

[0034] A self-diagnostic method for a recirculating ball type electric power steering system, such as Figure 2 As shown, the main steps include the following: S100: Set and store the signal transmission path corresponding to each source data type and the functional nodes contained in the signal transmission path, and number and store the source data types, signal transmission paths and functional nodes in association.

[0035] In this embodiment, the aforementioned source data refers to data signals output in response to driver actions, feedback data signals output by each functional node after completing a certain execution action, and data signals output from a set functional node to another functional node for fault diagnosis or verification. For example, the rotation angle output by the driver turning the steering wheel 17 is collected by the torque sensor 18 and converted into a torque sensor 18 signal, which is then output to the electronic control unit 19; the feedback data signal fed back to the ECU after the power assist motor completes its power assist response; and the temperature detection data signal fed back to the ECU from the power assist motor. Correspondingly, the signal transmission path includes the forward transmission path from the source data output node to the command execution node, and the feedback transmission path from the functional node where each monitoring data is collected to the signal processing node, etc.

[0036] A functional node refers to a functional component along the signal transmission path that processes the source data or responds to actions related to the output of the source data. For example... Figure 3 The diagram illustrates a signal transmission path: the driver outputs a rotation angle by turning the steering wheel 17. This rotation angle is converted into a torque sensor 18 signal at the functional node where the torque sensor 18 is located. The torque sensor 18 signal is then output to the functional node where the ECU is located for data signal processing to generate a PWM speed control signal, which is then output to the power steering motor. The power steering motor functional node responds to the PWM speed control signal and outputs rotational momentum via the shaft to the functional node where the reduction mechanism is located. Finally, the momentum is output to the steering screw via the worm gear from the functional node where the reduction mechanism is located. In this signal transmission path, the initial rotation angle, after being processed and transmitted by various functional nodes, ultimately acts as steering torque on the steering screw of the recirculating ball steering gear 21.

[0037] S200: Determine the types of influencing factors that affect the stability of data signal transmission between functional nodes and the accuracy of instruction execution at each functional node; analyze and obtain the correlation between the signal transmission path and the magnitude of data signal fluctuations at each functional node and the magnitude of the corresponding influencing factors; and store the results as a correction model.

[0038] Detailed, such as Figure 4 As shown, step S200, the method steps for determining the type of influencing factor specifically include: S210, Select potential influencing factors and source data for testing, and set the testing procedure. The aforementioned influencing factors include vehicle driving state parameters, road condition parameters, environmental parameters, and driving habit data. Specifically, vehicle driving state parameters include vehicle speed, engine pulse or drive motor speed, etc.; road condition parameters include road surface smoothness, road surface type (paved or unpaved), road surface friction coefficient, etc.; environmental parameters include ambient temperature, rain / snow intensity, and air humidity, etc.; and driving habit data includes the driver's acceleration when turning the steering wheel 17 and the acceleration when the steering wheel 17 returns to center, etc.

[0039] S211, The EPS simulation equipment is placed in the test environment chamber to obtain the response data corresponding to the source data at each functional node on the signal transmission path under the conditions of various influencing factors or their combinations. The aforementioned test environment chamber is used to provide a test environment with set environmental parameters and road condition parameters. The response data refers to the data signal output or the action performed by a functional node after receiving the data signal. For example, the torque sensor 18 outputs a torque sensor 18 signal in response to the rotation angle of the steering wheel 17 shaft.

[0040] S212, using correlation analysis, identifies the types of influencing factors affecting the accuracy of response data. As detailed in step S212, under conditions where no influencing factors are present, with fixed test source data input from the command input node, the response data generated at each functional node along the signal transmission path is predictable; that is, the theoretical response data is deterministic. By collecting actual response data from each functional node under different influencing factors, the impact of each factor on the accuracy of the response data at each functional node can be analyzed. When the sample data volume is large, correlation analysis can accurately and comprehensively obtain the types of influencing factors affecting the accuracy of response data.

[0041] In step S200, the method for analyzing and obtaining the corrected model includes: S220, based on the determined type of influencing factor, changes the value of the influencing factor in the environmental chamber, obtains the fluctuation range of the output response data of each functional node under different influencing factor values, and stores the values ​​of the influencing factor and the response data in association. S221, the correlation between the magnitude of data signal fluctuations on each functional node and the magnitude of the corresponding influencing factor values ​​is obtained by fitting the linear regression algorithm, and stored as the modified model.

[0042] The above technical solution can obtain the accuracy of the response of each functional node to different source data under the influence of various influencing factors, which is conducive to the fault detection system to accurately eliminate interference data signals when making fault judgment and improve the accuracy of fault detection.

[0043] S300, establish and store a first relational model to reflect the data correspondence between functional nodes on the same signal transmission path, and a second relational model to reflect the data correspondence between functional nodes on the signal transmission path and specific reference nodes on different signal transmission paths.

[0044] Combination Figure 1 As shown, since the data and signal processing algorithms at each functional node along the signal transmission path are clear and fixed, the correspondence between the data signals of each functional node is clear. For example, the relationship between the steering wheel 17 steering shaft rotation angle and the signal magnitude of the torque sensor 18 is clear and fixed. Similarly, when the torque sensor 18 signal is input to the ECU, a PWM speed control signal is output through a set algorithm, and the correspondence between the input and output data is clear. Therefore, the first relationship model can directly adopt the system's automatic control theoretical model. In practical applications, for some functional nodes on different signal transmission paths, there will also be a correspondence between their data signals. For example, there will be a correlation between the torque sensor 18 signal, the power assist motor current data, and the real-time vehicle speed data. However, the above data correlation is usually implicit, meaning there is no clear theoretical model between the correlations. Therefore, the corresponding second relationship model usually needs to be obtained through regression algorithms via data fitting or through neural network training. In this embodiment, the specific reference node is preferably a functional node that has a systematic correlation with the corresponding functional node rather than a direct correlation, meaning that interference signals appearing in the signal transmission path for a short period of time will not affect the data of the specific reference node.

[0045] The S400 acquires and temporarily stores real-time data on various influencing factors through various sensors or intelligent interconnected devices installed on the vehicle body.

[0046] S500: Obtain and, based on the source data, combine the first relationship model, real-time influence factor data, and correction model to generate theoretical response data at each functional node on the signal transmission path.

[0047] The S600 collects the actual response data at each functional node along the signal transmission path and calculates the deviation between the actual response data and the corresponding theoretical response data. In practical applications, the collection of the above-mentioned actual response data is usually accomplished by setting sensors, such as collecting the temperature of the booster motor through a temperature sensor, or collecting the current of the booster motor through a current transformer.

[0048] S700: Locate and identify functional nodes whose deviation exceeds a set value, mark these functional nodes as suspected fault nodes, and store them in association with the functional node number and data sampling time. Through the correction of the correction model, the suspected fault nodes obtained in step S700 have eliminated the influence of environmental interference and other influencing factors on the judgment result, making the fault judgment more accurate.

[0049] S800 outputs source data for verification after a set time interval, collects actual response data at suspected fault nodes and compares it with the corresponding theoretical response data: S810: If the deviation exceeds the set value, the first suspected fault node on the signal transmission path will be marked as a fault node and an alarm will be output. S820: If the deviation does not exceed the set value, the data information at the specific reference node associated with the first suspected fault node on the signal transmission path is obtained as reference data.

[0050] The verification source data output in step S810 above is used to re-determine whether the functional node is faulty, so as to avoid misjudgment. When the verification source data is executed, the electronic control unit 19 controls the corresponding functional node to perform isolation actions to avoid the verification source data affecting driving safety. For example, during verification, the electronic control unit 19 controls the electromagnetic clutch to disconnect the connection between the power assist motor and the reduction mechanism.

[0051] S821, Analyze the correspondence between actual response data and reference data: S8211, if the second relational model is satisfied, the suspected fault node is marked as a potential fault node and an early warning is output. The basis for the above judgment is that interference data signals usually have the characteristics of being sporadic and transient. For example, if abnormal data occurs when the data signal is transmitted in a strong electromagnetic radiation environment, and if the data of a specific reference node also changes with the above data signal, it indicates that the abnormal data is not interference data caused by environmental electromagnetic radiation.

[0052] S8212, if the second relational model is not satisfied, then remove the suspected fault mark of the suspected fault node.

[0053] The main purpose of step S821 above is to eliminate interfering data information and identify potential faults.

[0054] During actual driving, the probability of EPS malfunctions at different functional nodes along the signal transmission path varies depending on the influencing factors. For example, the probability of the power steering motor malfunctioning is higher in hot weather than under normal conditions, and the probability of the torque sensor 18 output signal malfunctioning is higher on rough roads than on flat roads. To avoid missing abnormal data and ensure the accuracy of fault detection results, in one embodiment, the fault self-diagnosis method described in this application further includes: A100 establishes and stores the correspondence between the data signal fluctuation amplitude and the sampling frequency at each functional node. For functional nodes whose output data signals are easily affected by factors, the corresponding sampling frequency is increased from the base sampling frequency. For functional nodes that are not affected by factors, their data signal sampling frequency is maintained at a set frequency.

[0055] A200, based on the real-time data of the acquired influencing factors and combined with the modified model, obtains the data signal fluctuation amplitude at each functional node; A300: Select a function node whose fluctuation amplitude exceeds the set value, and adjust the sampling frequency of the data signal at that function node according to the above correspondence between fluctuation amplitude and sampling frequency.

[0056] In another embodiment, the fault self-testing method described in this application further includes: B100: Count the number of times each functional node or potential fault node is marked, and calculate the fault marking rate of each functional node. B200 configures different data signal sampling frequencies for each functional node, and the data signal sampling frequency is set in proportion to the fault marking rate.

[0057] Based on the above technical solution, the higher the fault marking rate of the functional node, the higher the sampling frequency of the corresponding data signal, which avoids the omission of potential fault data information due to data loss or low sampling frequency, and improves the accuracy of fault detection.

[0058] To further identify abnormal data and discover potential system faults, the fault detection method described in this application further includes: C100 acquires and stores the actual response data and sampling time points of each functional node to each source of data of the same type; C200 normalizes the numerical values ​​of each source data and their corresponding actual response data, and stores the source data and their corresponding actual response data in association. C300 performs cluster analysis on the above actual response data to obtain abnormal data and their corresponding sampling time points; C400: Obtain reference data collected by a specific reference node associated with the above functional node at the above sampling time point, and analyze and compare the correspondence between the above abnormal data and the reference data: C410, if the second relationship model is satisfied, the above abnormal data is marked as potential fault data and stored in association with the functional node number and sampling time point; C420, if the second relational model is not satisfied, the above abnormal data will be marked as interference data.

[0059] The above technical solution can periodically analyze the collected and stored source data and its actual response data, find and mark potential fault data, which helps to discover potential fault nodes in EPS, make accurate predictions of faults, and improve the accuracy and timeliness of fault detection.

[0060] To ensure driving safety, the fault self-diagnosis method described in this application also includes: D100 establishes the correlation between the accuracy of command execution at each functional node and the driving safety performance value. For example, the safety status value at the power steering motor is set to 10 points. When the deviation between the actual torque output by the power steering motor (which can be detected by the torque sensor 18 inside the recirculating ball steering gear 21) and the theoretical output torque (which can be calculated by the control parameters output by the ECU) is less than 1%, the output safety status value is 10 points. When the above deviation is 1%-2%, the output safety status value is 9 points, and so on, to establish the correspondence between the accuracy of command execution and the driving safety performance value.

[0061] D200: Based on the above relationships, configure safety impact weights for each functional node. In practical applications, the impact of a failure on driving safety varies depending on the functional node. For example, a failure of an indicator light in the EPS feedback alarm circuit will not directly affect driving safety. Therefore, configure safety impact weights for each functional node according to the relationships in step D100.

[0062] D300 generates the safety status value of each functional node based on the deviation between the actual response data and the corresponding theoretical response data at each functional node, according to the aforementioned correlation.

[0063] D400 calculates the security judgment value by weighting the security status values ​​at each functional node and their corresponding security impact weights.

[0064] The D500 compares the weighted judgment value with the set driving safety threshold, and outputs driving safety alarms of different levels according to the difference between the weighted judgment value and the driving safety threshold.

[0065] In a specific implementation, the output method of the above-mentioned safety status value is defined as follows: when the deviation exceeds the set value, the safety status value is output as 0; when the deviation does not exceed the set value, the safety status value is output as 1. For each functional node located on the same signal transmission path, under non-special configuration, its corresponding safety sampling weight value gradually decreases along the signal transmission direction on the signal transmission path.

[0066] The above technical solution links the data signal deviation amplitude at each functional node with driving safety. When multiple functional nodes or functional nodes with a large safety impact weight value fail, driving safety warnings can be output in a timely manner, prompting the driver to drive cautiously or park to troubleshoot the fault, which helps to improve driving safety.

[0067] To implement the above-mentioned self-diagnosis method for recirculating ball electric power steering systems, this application also proposes a self-diagnosis system for recirculating ball electric power steering systems, such as... Figure 5 As shown (for simplification, the connection path for storing data signals into data storage unit 1 is omitted in the figure), it mainly includes: data storage unit 1, influence factor data acquisition unit 2, source data monitoring unit 3, theoretical response data generation unit 4, actual response data acquisition unit 5, deviation calculation unit 6, mark verification unit 7, fault alarm output unit 8, potential fault confirmation unit 9, signal fluctuation amplitude calculation unit 10, and sampling frequency configuration unit 11.

[0068] Data storage unit 1 includes a data storage chip configured to store the signal transmission path corresponding to each source data type and the functional nodes contained in the signal transmission path; a correction model reflecting the correlation between the magnitude of data signal fluctuations on each signal transmission path / functional node and the magnitude of the corresponding influencing factor; a first relation model reflecting the data correspondence between functional nodes on the same signal transmission path; a second relation model reflecting the data correspondence between functional nodes on the signal transmission path and specific reference nodes on different signal transmission paths; various types of influencing factor data and fault markers corresponding to each functional node; the correspondence between the magnitude of data signal fluctuations at each functional node and the sampling frequency; and the correlation between the accuracy of instruction execution at each functional node and the driving safety characterization value.

[0069] The impact factor data acquisition unit 2 is configured to acquire various impact factor data in real time. These impact factor types include vehicle driving state parameters, road condition parameters, environmental parameters, and driving habit data. Specifically, depending on the type of impact factor, it includes sensors such as temperature and humidity sensors, rain and snow sensors, vehicle speed detection sensors, road surface unevenness sensors, and angular acceleration sensors connected to the steering shaft of the steering wheel 17.

[0070] The source data monitoring unit 3 is configured to monitor and acquire source data in real time. The theoretical response data generation unit 4 is configured to generate theoretical response data at each functional node along the signal transmission path based on the source data, combined with the first relationship model, real-time influence factor data, and the correction model. The actual response data acquisition unit 5 includes various data sensors and their corresponding data sampling subunits, configured to acquire actual response data at each functional node along the signal transmission path based on a set sampling frequency.

[0071] Deviation calculation unit 6 is configured to receive and calculate the deviation amplitude between the actual response data and the corresponding theoretical response data at each functional node. Specifically, the calculation method is as follows: subtract the theoretical response data value from the actual response data value, and then compare the result with the theoretical response data to obtain the deviation amplitude. Marking verification unit 7, data-connected to deviation calculation unit 6, is configured to locate and determine functional nodes whose deviation amplitude exceeds a set value, mark these functional nodes as suspected fault nodes, output verification source data at set intervals, obtain the actual response data at the suspected fault nodes, compare it with the corresponding theoretical response data, and output the verification result of the deviation amplitude of the actual response data against the set value.

[0072] The fault alarm output unit 8 is connected to the marking verification unit 7 to obtain the verification result. If the above deviation exceeds the set value, the first suspected fault node on the signal transmission path is marked as a fault node and an alarm is output. The alarm information is output to the display interface of the vehicle system.

[0073] The potential fault confirmation unit 9 receives the verification result. If the above deviation does not exceed the set value, it obtains the data information of a specific reference node associated with the first suspected fault node on the signal transmission path as reference data and analyzes the correspondence between the actual response data and the reference data. If the second relationship model is satisfied, the suspected fault node is marked as a potential fault node and an early warning is output. If the second relationship model is not satisfied, the suspected fault node is removed from the suspected fault mark.

[0074] like Figure 5 As shown, the signal fluctuation amplitude calculation unit 10 is configured to acquire and, based on real-time data of influencing factors, combine with the correction model to obtain the data signal fluctuation amplitude at each functional node. The sampling frequency configuration unit 11 is configured to select functional nodes whose fluctuation amplitude exceeds a set value, adjust the sampling frequency of the data signal at that functional node according to the correspondence between the fluctuation amplitude and the sampling frequency, and output it to the actual response data acquisition unit 5. The above technical solution can focus on data acquisition from functional nodes with large data signal fluctuation amplitudes, avoid data omissions, and ensure the accuracy of fault detection.

[0075] In order to determine whether a fault in the EPS affects driving safety, the system in this embodiment of the application further includes a driving safety alarm module 12, which specifically includes: a safety impact weight configuration unit 13, a safety status value assignment unit 14, a driving safety judgment value generation unit 15, and a driving safety alarm output unit 16.

[0076] The safety impact weight configuration unit 13 configures safety impact weights for each functional node based on the correlation between the accuracy of instruction execution and the driving safety characterization value. The safety status value assignment unit 14 is configured to generate a safety status value for each functional node based on the deviation between the actual response data and the corresponding theoretical response data. The driving safety judgment value generation unit 15 is configured to acquire and calculate a safety judgment value by weighting the safety status value and its corresponding safety impact weight for each functional node. The driving safety alarm output unit 16 is configured to compare the weighted judgment value with a set driving safety threshold and output different levels of driving safety alarms based on the difference between the weighted judgment value and the driving safety threshold.

[0077] Based on the above technical solution, when a critical node on the EPS signal transmission path fails, a driving safety alarm can be output in a timely manner to ensure driving safety.

[0078] Finally, in order to facilitate the widespread application of the EPS fault self-diagnosis method in this application, this application also discloses a computer-readable storage medium on which a program module is loaded. When the program module is executed by a processor, it is used to implement the aforementioned recirculating ball electric power steering system fault self-diagnosis method.

[0079] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A self-diagnosis method for a recirculating ball type electric power steering system, characterized in that, include: Set and store the signal transmission path corresponding to each source data type and the functional nodes contained in the signal transmission path, and number and store the above source data types, signal transmission paths and functional nodes together. Identify the types of influencing factors affecting the accuracy of instruction execution at each functional node, analyze the correlation between the signal transmission path and the magnitude of data signal fluctuations at each functional node and the corresponding influencing factor values, and store this information as a correction model. Establish and store a first relational model to reflect the data correspondence between functional nodes on the same signal transmission path, and a second relational model to reflect the data correspondence between functional nodes on the signal transmission path and specific reference nodes on different signal transmission paths; Acquire real-time data on various impact factors and store them temporarily; Based on the source data, combined with the first relationship model, real-time impact factor data and the correction model, theoretical response data at each functional node on the signal transmission path is generated. Collect the actual response data at each functional node along the signal transmission path and calculate the deviation between the actual response data and the corresponding theoretical response data; Locate and identify functional nodes whose deviation exceeds a set value, mark these functional nodes as suspected fault nodes, and store them in association with the functional node number and data sampling time. After a set time interval, output the source data for verification, collect the actual response data at suspected fault nodes and compare it with the corresponding theoretical response data: If the deviation exceeds the set value, the first suspected fault node on the signal transmission path will be marked as a fault node and an alarm will be output. If the deviation does not exceed the set value, the data information at the specific reference node associated with the first suspected fault node on the signal transmission path is obtained as reference data. Analyze the correspondence between actual response data and reference data: If the second relation model is satisfied, the suspected fault node is marked as a potential fault node and an early warning is output; If the second relation model is not satisfied, then remove the suspected fault mark from the suspected fault node; The signal transmission path includes the forward transmission path from the source data output node to the instruction execution node, and the feedback transmission path from the functional node where each monitoring data is collected to the signal processing node. The types of influencing factors include vehicle driving status parameters, road condition parameters, environmental parameters, and driving habit data.

2. The self-diagnosis method for a recirculating ball type electric power steering system according to claim 1, characterized in that, Determining the type of impact factor includes: Select potential influencing factors and source data for testing, and set up the testing process; The EPS simulation equipment was placed in the test environment chamber to obtain the response data corresponding to the source data at each functional node on the signal transmission path under the conditions of various influencing factors or their combinations. The types of influencing factors affecting the accuracy of response data were determined based on correlation analysis. Analysis yields a corrected model, including: Based on the identified types of influencing factors, the numerical values ​​of these factors in the environmental chamber are varied to obtain the fluctuation range of the output response data of each functional node under different influencing factor values. The correlation between the magnitude of data signal fluctuations at each functional node and the magnitude of the corresponding influencing factor values ​​is obtained by fitting the data using a linear regression algorithm and stored as the corrected model.

3. The self-diagnosis method for a recirculating ball electric power steering system according to claim 1, characterized in that, The fault self-test method also includes: Establish and store the correspondence between the data signal fluctuation amplitude and the sampling frequency at each functional node; Based on the real-time data of the influencing factors, the fluctuation amplitude of the data signal at each functional node is obtained by combining the modified model. Select a functional node whose fluctuation amplitude exceeds the set value, and adjust the sampling frequency of the data signal at that functional node according to the above correspondence between fluctuation amplitude and sampling frequency.

4. The self-diagnosis method for a recirculating ball electric power steering system according to claim 1, characterized in that, The fault detection method further includes: Acquire and store the actual response data and sampling time points of each functional node to each source of data of the same type; The numerical values ​​of the source data and their corresponding actual response data are normalized and stored together. Cluster analysis was performed on the above actual response data to obtain the abnormal data and their corresponding sampling time points; Obtain reference data collected at the above sampling time points for specific reference nodes associated with the above functional nodes, and analyze and compare the correspondence between the above abnormal data and the reference data: If the second relationship model is satisfied, the above abnormal data will be marked as potential fault data and stored in association with the functional node number and sampling time point; If the second relational model is not satisfied, the above abnormal data will be marked as interference data.

5. The self-diagnosis method for a recirculating ball electric power steering system according to claim 1, characterized in that, The fault self-test method also includes: Count the number of times each functional node is marked as a faulty node or potential faulty node, and calculate the fault marking rate of each functional node; Configure different data signal sampling frequencies for each functional node; The data signal sampling frequency is set in direct proportion to the fault marking rate.

6. The self-diagnosis method for a recirculating ball electric power steering system according to claim 1, characterized in that, The fault self-test method also includes: Establish the correlation between the accuracy of instruction execution at each functional node and the performance value of driving safety; Based on the above relationships, configure security impact weights for each functional node; The safety status value of each functional node is generated based on the magnitude of the deviation between the actual response data and the corresponding theoretical response data at each functional node. The security judgment value is obtained by weighted calculation based on the security status value of each functional node and its corresponding security impact weight; The weighted judgment value is compared with the set driving safety threshold, and driving safety alarms of different levels are output according to the difference between the weighted judgment value and the driving safety threshold.

7. A self-diagnostic system for a recirculating ball type electric power steering system, characterized in that, include: The data storage unit (1) is configured to store the signal transmission path corresponding to each source data type and the functional nodes contained in the signal transmission path, a correction model to reflect the correlation between the magnitude of the data signal fluctuation on each signal transmission path / functional node and the magnitude of the corresponding influence factor, a first relation model to reflect the data correspondence between functional nodes on the same signal transmission path, a second relation model to reflect the data correspondence between functional nodes on the signal transmission path and specific reference nodes on different signal transmission paths, the acquired various influence factor data and the fault markers corresponding to each functional node; Impact factor data acquisition unit (2) is configured to acquire various impact factor data in real time; The source data monitoring unit (3) is configured to monitor and acquire source data in real time; The theoretical response data generation unit (4) is configured to generate theoretical response data at each functional node on the signal transmission path based on the source data, combined with the first relationship model, real-time impact factor data and the correction model; The actual response data acquisition unit (5) is configured to acquire the actual response data at each functional node on the signal transmission path based on a set sampling frequency; The deviation calculation unit (6) is configured to receive and calculate the deviation between the actual response data and the corresponding theoretical response data at each functional node; The marking verification unit (7) is connected to the deviation calculation unit (6) and configured to find and determine the functional nodes whose deviation exceeds the set value. The functional nodes are marked as suspected fault nodes. The source data for verification is output at a set interval. The actual response data at the suspected fault node is obtained and compared with the corresponding theoretical response data. The verification result of the deviation of the actual response data and the set value is output. The fault alarm output unit (8) is connected to the marking verification unit (7) to obtain the verification result. If the above deviation exceeds the set value, the first suspected fault node on the signal transmission path is marked as a fault node and an alarm is output. The potential fault confirmation unit (9) receives the verification result. If the above deviation does not exceed the set value, it acquires the data information at a specific reference node associated with the first suspected fault node on the signal transmission path as reference data and analyzes the correspondence between the actual response data and the reference data. If the second relation model is satisfied, the suspected fault node is marked as a potential fault node and an early warning is output; If the second relation model is not satisfied, then remove the suspected fault mark from the suspected fault node; The types of influencing factors include vehicle driving status parameters, road condition parameters, environmental parameters, and driving habit data.

8. The self-diagnostic system for a recirculating ball type electric power steering system according to claim 7, characterized in that, The data storage unit (1) also stores the correspondence between the data signal fluctuation amplitude and the sampling frequency at each functional node; The system also includes: The signal fluctuation amplitude calculation unit (10) is configured to acquire and obtain the data signal fluctuation amplitude at each functional node based on the real-time data of the influencing factors and the correction model. The sampling frequency configuration unit (11) is configured to select a functional node whose fluctuation amplitude exceeds a set value, adjust the sampling frequency of the data signal at the functional node according to the above correspondence between the fluctuation amplitude and the sampling frequency, and output it to the actual response data acquisition unit (5).

9. The self-diagnostic system for a recirculating ball type electric power steering system according to claim 7, characterized in that, The data storage unit (1) also stores the correlation between the accuracy of instruction execution and the driving safety characterization value at each functional node; The system also includes a driving safety alarm module (12), comprising: The safety impact weight configuration unit (13) configures safety impact weights for each functional node based on the correlation between the above-mentioned instruction execution accuracy and driving safety characterization value; The safety status value assignment unit (14) is configured to generate the corresponding safety status value at each functional node based on the magnitude of the deviation between the actual response data at each functional node and the corresponding theoretical response data. The driving safety judgment value generation unit (15) is configured to obtain and calculate the safety judgment value by weighting the safety status value and the corresponding safety impact weight at each functional node. The driving safety alarm output unit (16) is configured to compare the weighted judgment value with the set driving safety threshold, and output different levels of driving safety alarms according to the difference between the weighted judgment value and the above driving safety threshold.

10. A computer-readable storage medium, characterized in that, It is loaded with a program module, which, when executed by the processor, is used to implement the self-diagnosis method for the recirculating ball electric power steering system as described in any one of claims 1-6.

Images