Abnormality detection method and system for automatic driving system, and computer readable medium

By calculating longitudinal acceleration anomalies, braking anomalies, and the number of "dragon strokes" in vehicle driving data, and combining Butterworth filters and weighting coefficients, the consistency and real-time issues in autonomous driving system detection are resolved, enabling comprehensive and real-time evaluation and optimization of the system.

CN118907111APending Publication Date: 2024-11-08HOZON NEW ENERGY AUTOMOBILE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411181293.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Existing methods for detecting and evaluating autonomous driving systems rely on subjective observation and simulated driving, resulting in inconsistent evaluation results and difficulty in reflecting the complexities of actual road conditions. They also lack real-time performance and cannot effectively monitor and optimize system performance.

Method used

By acquiring vehicle driving data, calculating the number of longitudinal acceleration anomalies, braking anomalies, and erratic braking events, and using a Butterworth filter for signal processing, combined with weighting coefficients to calculate user perception indicators, real-time and objective detection of the autonomous driving system can be achieved.

Benefits of technology

It enables real-time and comprehensive anomaly detection for autonomous driving systems, improves detection and evaluation effectiveness, provides multi-dimensional performance optimization guidance, and enhances driving comfort, safety, and behavioral norms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118907111A_ABST
    Figure CN118907111A_ABST
Patent Text Reader

Abstract

The invention relates to an anomaly detection method and system of an automatic driving system and a computer readable medium. The anomaly detection method comprises the steps of obtaining vehicle driving data of the automatic driving system in a vehicle driving process; calculating the number of longitudinal acceleration abnormities according to the vehicle driving data; calculating the number of abnormal braking times according to the vehicle driving data; calculating dragon drawing times according to the vehicle driving data; and obtaining a user somatosensory index of the automatic driving system according to the longitudinal acceleration abnormity times, the braking abnormity times and the dragon drawing times. The abnormal condition of the automatic driving system can be objectively and comprehensively detected in real time, and the detection and evaluation effects of the automatic driving system are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention mainly relates to the technical field of autonomous driving systems, and specifically to an abnormality detection method, system and computer-readable medium for an autonomous driving system. Background Art

[0002] With the rapid development of intelligent driving technology, various companies have launched their own autonomous driving systems, such as the Pilot system. As one of the important innovations in the automotive industry, the importance of the autonomous driving system is becoming increasingly prominent. The core function of the autonomous driving system is to fully perceive the vehicle's surrounding environment, accurately formulate driving strategies, make decisions quickly, and implement control efficiently to achieve an automated driving experience. However, to ensure that the Pilot system is both efficient and safe in actual applications, comprehensive testing, evaluation, and continuous monitoring are particularly critical.

[0003] In the past, the means of abnormal detection and evaluation of autonomous driving systems usually relied on subjective observation, complex test scenarios and simulated driving. This method of evaluating system performance through subjective observation inevitably introduces human factors, that is, the subjective bias of different drivers or observers, which may lead to significant inconsistency and deviation in the evaluation results. Although this method can simulate a variety of driving scenarios to a certain extent, it is often difficult to fully reflect the complex and changeable conditions on actual roads due to the authenticity of the simulation environment and the coverage of the data. Some evaluation methods lack real-time performance and cannot provide immediate performance feedback for the Pilot system, which not only limits the optimization space of system performance, but also hinders the effective real-time monitoring and adjustment of the system status. The existing detection and evaluation methods of autonomous driving systems have the problem of poor results. Summary of the invention

[0004] The technical problem to be solved by the present application is to provide an abnormality detection method, system and computer-readable medium for an autonomous driving system, which can detect abnormal conditions of the autonomous driving system in real time, objectively and comprehensively, and improve the detection and evaluation effects of the autonomous driving system.

[0005] The technical solution adopted by the present application to solve the above-mentioned technical problems is an abnormality detection method for an autonomous driving system, including: obtaining vehicle driving data of the autonomous driving system during vehicle driving; calculating the number of longitudinal acceleration abnormalities based on the vehicle driving data; calculating the number of braking abnormalities based on the vehicle driving data; calculating the number of dragon drawing times based on the vehicle driving data; and obtaining user physical perception indicators of the autonomous driving system based on the number of longitudinal acceleration abnormalities, the number of braking abnormalities and the number of dragon drawing times.

[0006] In one embodiment of the present application, the step of calculating the number of longitudinal acceleration anomalies based on vehicle driving data includes: filtering out longitudinal vehicle control mode data from the vehicle driving data; filtering out a longitudinal acceleration signal from the longitudinal vehicle control mode data; filtering the longitudinal acceleration signal to obtain a filtered longitudinal acceleration signal; calculating the longitudinal acceleration change rate based on the filtered longitudinal acceleration signal; if the longitudinal acceleration change rate is greater than a first preset threshold, the number of longitudinal acceleration anomalies is accumulated once; wherein the first preset threshold is greater than 0.

[0007] In one embodiment of the present application, the step of calculating the number of braking abnormalities based on vehicle driving data includes: filtering out longitudinal vehicle control mode data from the vehicle driving data; filtering out a longitudinal acceleration signal from the longitudinal vehicle control mode data; filtering the longitudinal acceleration signal to obtain a filtered longitudinal acceleration signal; calculating the longitudinal acceleration change rate based on the filtered longitudinal acceleration signal; if the longitudinal acceleration change rate is greater than a second preset threshold and the longitudinal acceleration signal is less than a third preset threshold, the number of braking abnormalities is accumulated once; wherein the second preset threshold is greater than 0 and the third preset threshold is less than 0.

[0008] In an embodiment of the present application, the step of filtering the longitudinal acceleration signal includes: performing low-pass filtering on the longitudinal acceleration signal according to a Butterworth filter.

[0009] In one embodiment of the present application, the step of calculating the longitudinal acceleration change rate according to the filtered longitudinal acceleration signal includes: calculating the longitudinal acceleration change rate using the following formula:

[0010] Longitudinal acceleration change rate = (current frame - previous frame) / (current moment - previous moment)

[0011] The current frame represents the filtered longitudinal acceleration signal at the current moment, and the previous frame represents the filtered longitudinal acceleration signal at the previous moment.

[0012] In one embodiment of the present application, the step of calculating the number of dragon drawings based on vehicle driving data includes: filtering out first driving data in which the steering torsion bar torque is less than or equal to a fourth preset threshold from the vehicle driving data; wherein the fourth preset threshold is greater than 0; filtering out second driving data when the turn signal is off from the first driving data; filtering out a lateral acceleration signal from the second driving data; and calculating the number of dragon drawings based on the lateral acceleration signal and preset rules.

[0013] In one embodiment of the present application, the step of calculating the number of dragon drawing according to the lateral acceleration signal and preset rules includes: determining whether the lateral acceleration signal has two peaks and troughs within a preset time length; if the judgment is yes, and the absolute value of at least one peak or at least one trough is greater than the fifth preset threshold, and the difference between one of the peaks and troughs is greater than the sixth preset threshold, and the difference between the other peak and trough is greater than the seventh preset threshold, then the number of dragon drawing is accumulated once; wherein the fifth preset threshold is greater than 0, the sixth preset threshold is greater than 0, and the seventh preset threshold is greater than 0.

[0014] In one embodiment of the present application, the step of obtaining a user perception index of the automatic driving system according to the number of abnormal longitudinal acceleration, the number of abnormal braking, and the number of dragon drawing comprises: calculating the user perception index using the following formula:

[0015] User sensory index = α*number of abnormal longitudinal acceleration times + β*number of abnormal braking times + γ*number of dragon drawing times

[0016] Here, α represents the first weight coefficient, β represents the second weight coefficient, γ represents the third weight coefficient, and α+β+γ=1.

[0017] In order to solve the above-mentioned technical problems, the present application also proposes an abnormality detection system for an autonomous driving system, comprising: a memory for storing instructions executable by a processor; and a processor for executing instructions to implement the above-mentioned abnormality detection method for the autonomous driving system.

[0018] In order to solve the above technical problems, the present application also proposes a computer-readable medium storing computer program code, which, when executed by a processor, implements the above-mentioned abnormality detection method for the autonomous driving system.

[0019] The technical solution of this application can obtain the vehicle driving data of the autonomous driving system in real time. By calculating the number of abnormal longitudinal acceleration, the smoothness of the driving process can be evaluated; by calculating the number of abnormal braking, the braking system's point braking and emergency braking can be evaluated, and then the safety of the system can be evaluated; by calculating the number of dragon drawing times, the instability of the vehicle's driving trajectory can be evaluated; by combining the number of abnormal longitudinal acceleration, the number of abnormal braking and the number of dragon drawing times, the user's somatosensory index of the autonomous driving system can be obtained, which can comprehensively measure the driving comfort, safety and behavioral norms. This application can detect the abnormal conditions of the autonomous driving system in real time, objectively and comprehensively, and improve the detection and evaluation effect of the autonomous driving system. The user's somatosensory index of this application can provide multi-dimensional guidance for optimizing the performance of the autonomous driving system and improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below with reference to the accompanying drawings, wherein:

[0021] Figure 1 is an exemplary flow chart of an abnormality detection method for an autonomous driving system according to an embodiment of the present application;

[0022] Figure 2 is an exemplary result diagram of analyzing abnormal longitudinal acceleration of a vehicle in one embodiment of the present application;

[0023] Figure 3 is an exemplary result diagram of analyzing abnormal braking conditions of a vehicle in one embodiment of the present application;

[0024] Figure 4 This is an exemplary result diagram of analyzing the situation where a vehicle draws a dragon in one embodiment of the present application;

[0025] Figure 5 This is a system block diagram of an abnormality detection system of an autonomous driving system according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below with reference to the accompanying drawings.

[0027] In the following description, many specific details are set forth to facilitate a full understanding of the present application, but the present application may also be implemented in other ways different from those described herein, and therefore the present application is not limited to the specific embodiments disclosed below.

[0028] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0029] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously. At the same time, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0030] The present application proposes an anomaly detection method for an autonomous driving system, which can be applied to scenarios where the autonomous driving system is to be detected and evaluated in real time. The anomaly detection method for an autonomous driving system of the present application can be run on the vehicle side, for example, in the controller of the vehicle, and can also be run on a cloud platform. When the anomaly detection method for an autonomous driving system is run on a cloud platform, the vehicle-side data and the cloud platform data interact through a wireless network. Exemplarily, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an interconnected cloud, multiple clouds, etc., or any combination thereof. The present application does not limit the operating environment of the anomaly detection method for an autonomous driving system.

[0031] Figure 1 is an exemplary flow chart of an abnormality detection method for an autonomous driving system according to an embodiment of the present application, with reference to Figure 1 As shown, the abnormality detection method of the automatic driving system of this embodiment includes the following steps:

[0032] Step S1: Acquire vehicle driving data of the automatic driving system during vehicle driving.

[0033] Step S2: Calculate the number of abnormal longitudinal accelerations based on the vehicle driving data.

[0034] Step S3: Calculate the number of braking abnormalities according to the vehicle driving data.

[0035] Step S4: Calculate the number of dragon drawing times according to the vehicle driving data.

[0036] Step S5: Obtain the user perception index of the automatic driving system according to the number of abnormal longitudinal acceleration, the number of abnormal braking and the number of dragon drawing.

[0037] Exemplarily, the execution order of the above steps S2, S3, and S4 can be changed, or steps S2, S3, and S4 can be executed simultaneously. The present application does not limit the execution order of steps S2, S3, and S4.

[0038] The above steps S1 to S5 are described in detail below:

[0039] In step S1, the vehicle driving data of the automatic driving system is obtained during the driving process of the vehicle. Exemplarily, the vehicle may upload a CAN (Controller Area Network) data packet every 10 seconds, and the frequency of data contained in each data packet is 1Hz. The CAN data packet may include data during manual driving and data during automatic driving, so this application needs to extract the vehicle driving data of the automatic driving system in the automatic driving mode from the CAN data packet. For example, according to the CAN data packet of each smart driving car uploaded to the smart driving cloud platform by the smart driving domain controller, take the data when the ADCS8_Npilot_SysState (N-pilot System State) PILOT smart driving system state is in the activated state.

[0040] When the abnormality detection method of this application is run on the cloud platform, the CAN data packet is reliably transmitted to the intelligent driving cloud platform through the vehicle intelligent driving domain controller and then enters a dedicated data warehouse. This application can filter out the required vehicle driving data from the data warehouse in real time, and then evaluate the performance of the autonomous driving system by detecting abnormal conditions. In the data warehouse, data can be recorded, analyzed and processed in real time to monitor and evaluate the number of abnormal longitudinal acceleration (hereinafter referred to as the number of longitudinal abnormal jerk times in this application), the number of abnormal braking (hereinafter referred to as the number of point braking / emergency braking times in this application), and the number of dragon drawing times.

[0041] For example, in actual applications, the user's physical perception of the intelligent driving system Pilot can be evaluated by combining camera data, lidar data, and ultrasonic sensor data of the vehicle driving on the actual road, or simulated simulation data in the simulation scene. Diversified data can provide more information about the surrounding environment and be used to identify obstacles, road signs, and the position and speed of other vehicles.

[0042] In step S2, the number of abnormal longitudinal accelerations is calculated based on the vehicle driving data. Exemplarily, the number of abnormal longitudinal accelerations can be referred to as the number of abnormal longitudinal jerks. In an intelligent autonomous driving system, abnormal longitudinal jerk means that the vehicle has experienced unusual longitudinal acceleration changes when the intelligent driving function is activated. This abnormal jerk may be caused by abnormal acceleration or braking behavior. The occurrence of abnormal longitudinal jerk may cause discomfort to the driver and passengers, and may also have potential impacts on road safety. Therefore, it is very important for the autonomous driving system to monitor and reduce abnormal longitudinal jerk.

[0043] In some embodiments, the step of calculating the number of abnormal longitudinal accelerations according to the vehicle driving data includes:

[0044] Step S2a: Filter out the longitudinal vehicle control mode data from the vehicle driving data. Exemplarily, the longitudinal vehicle control mode refers to controlling the forward and backward movement of the vehicle (i.e., longitudinal movement), which mainly focuses on the adjustment of parameters such as vehicle speed, acceleration, and the distance between the front and rear vehicles or obstacles. In the longitudinal vehicle control mode, the vehicle will automatically adjust the speed according to the preset algorithm or the driver's settings. In actual applications, if ADCS2_longitudCtrlType (ADAS longitudinal control request type) is 0x2: NPilot Mode, it means that the vehicle is in PILOT longitudinal vehicle control mode.

[0045] Step S2b: Filter out the longitudinal acceleration signal from the longitudinal vehicle control mode data. For example, in practical applications, if the vehicle longitudinal acceleration sensor value ACU2_LongAccSensorValue (Longitude AccelerationSensor Value) is a valid value, it corresponds to the longitudinal acceleration signal.

[0046] Step S2c: Filter the longitudinal acceleration signal to obtain a filtered longitudinal acceleration signal.

[0047] Step S2d: Calculate the longitudinal acceleration change rate according to the filtered longitudinal acceleration signal.

[0048] Step S2e: If the longitudinal acceleration change rate is greater than a first preset threshold, the number of longitudinal acceleration abnormalities is accumulated once, that is, counted as one longitudinal abnormal jerk; wherein the first preset threshold is greater than 0. Exemplarily, the first preset threshold can be set to 3.65.

[0049] In some embodiments, in step S2c, the step of filtering the longitudinal acceleration signal includes: performing low-pass filtering on the longitudinal acceleration signal according to a Butterworth filter. For example, in practical applications, ACU2_LongAccSensorValue can be low-pass filtered to smooth the signal, for example, setting function b, a = signal.butter(2, 0.4, output = 'ba', fs = 1). The present application can remove high-frequency noise in the longitudinal acceleration signal through low-pass filtering, making subsequent calculations more accurate.

[0050] In some embodiments, in step S2d, the step of calculating the longitudinal acceleration change rate according to the filtered longitudinal acceleration signal includes: calculating the longitudinal acceleration change rate using the following formula (1):

[0051] Longitudinal acceleration change rate = (current frame - previous frame) / (current moment - previous moment) (1)

[0052] The current frame represents the filtered longitudinal acceleration signal at the current moment, and the previous frame represents the filtered longitudinal acceleration signal at the previous moment. For example, the "frame" and "moment" in the above formula (1) of the present application can be understood as the continuously sampled longitudinal acceleration signal and its corresponding timestamp.

[0053] In step S3, the number of abnormal braking is calculated based on the vehicle driving data. For example, the number of abnormal braking can be understood as the number of point braking / emergency braking. In the intelligent automatic driving system, point braking / emergency braking refers to the vehicle performing abrupt braking when the intelligent driving function is activated, that is, rapid deceleration or stopping. This operation usually occurs when there is a need to respond to a dangerous situation urgently or when emergency braking is required.

[0054] For the braking / emergency braking situation when the intelligent driving function is activated, the factors usually considered include: (1) Emergency situation: Braking / emergency braking usually occurs in emergency situations, such as when an obstacle suddenly appears in front, other vehicles brake suddenly or engage in dangerous behavior, etc. This may be to avoid collision or reduce risks. (2) Intelligent decision-making: Intelligent driving systems usually monitor road conditions through sensors and algorithms. When the system detects a potential dangerous situation, it may automatically perform braking / emergency braking operations. The system needs to be able to quickly identify and respond to potential dangers to maintain the safety of the vehicle and passengers. (3) Driving comfort: In the case of braking / emergency braking, driving comfort is usually affected because such sudden braking may cause discomfort to passengers. Therefore, the intelligent driving system also needs to consider the comfort of passengers when braking / emergency braking. (4) Monitoring and recording: The intelligent driving system needs to record and analyze braking / emergency braking events in order to analyze and evaluate its performance. This can be used to improve the system's decision-making algorithm and improve driving safety.

[0055] In some embodiments, the step of calculating the number of braking abnormalities according to the vehicle driving data includes:

[0056] Step S3a: Filter out longitudinal vehicle control mode data from the vehicle driving data. For example, in practical applications, if ADCS2_longitudCtrlType (ADAS longitudinal control request type) is 0x2: NPilot Mode, it means that the vehicle is in PILOT longitudinal vehicle control mode.

[0057] Step S3b: Filter out the longitudinal acceleration signal from the longitudinal vehicle control mode data. For example, in practical applications, if the vehicle longitudinal acceleration sensor value ACU2_LongAccSensorValue (Longitude AccelerationSensor Value) is a valid value, it corresponds to the longitudinal acceleration signal.

[0058] Step S3c: Filter the longitudinal acceleration signal to obtain a filtered longitudinal acceleration signal. For example, in practical applications, ACU2_LongAccSensorValue may be low-pass filtered to smooth the signal, for example, setting function b, a = signal.butter (2, 0.4, output = 'ba', fs = 1).

[0059] Step S3d: Calculate the longitudinal acceleration change rate according to the filtered longitudinal acceleration signal. For example, the longitudinal acceleration change rate can be calculated using the above formula (1).

[0060] Step S3e: If the longitudinal acceleration change rate is greater than the second preset threshold and the longitudinal acceleration signal (corresponding to ACU2_LongAccSensorValue) is less than the third preset threshold, the number of abnormal braking is accumulated once, that is, recorded as one braking; wherein the second preset threshold is greater than 0, and the third preset threshold is less than 0. Exemplarily, the second preset threshold can be set to 3.65, and the third preset threshold can be set to -1.34.

[0061] In step S4, the number of dragon drawing is calculated based on the vehicle driving data. For example, in intelligent driving, "dragon drawing" is a specific driving phenomenon, which is usually manifested as a series of small and rapid left and right turns of the vehicle's steering wheel, like drawing a curved line. This phenomenon may occur when the intelligent driving function is activated, and is usually related to specific driving conditions or control logic.

[0062] There are many reasons why the dragon-drawing phenomenon may occur, some of which may include: (1) Misidentification or environmental interference: The intelligent driving system may misidentify the road or traffic conditions, causing the vehicle to make unnecessary steering wheel adjustments. This may be caused by visual or sensor errors, unusual road conditions, or other environmental factors. (2) Control logic problems: There may be problems with the control logic of the intelligent driving system, resulting in unnecessary changes in the steering wheel. This may be caused by software errors, sensor failures, or communication problems. (3) Human-computer interaction problems: The dragon-drawing phenomenon may also be related to the interaction between the driver and the intelligent driving system, which may be caused by inappropriate or confusing communication instructions between the driver and the autonomous driving system.

[0063] After understanding the dragon-drawing phenomenon, it is necessary to identify and eliminate the occurrence of this phenomenon, which requires detailed troubleshooting and analysis of the hardware, software, and sensors of the intelligent driving system. After determining the root cause, developers can take corresponding measures to improve the system and improve its performance and reliability to ensure that the dragon-drawing phenomenon does not occur again in the activated state. Therefore, identifying the occurrence of the dragon-drawing phenomenon can effectively evaluate the safety and comfort of the intelligent driving system.

[0064] In some embodiments, the step of calculating the number of dragon drawing times according to the vehicle driving data includes:

[0065] Step S4a: Filter out first driving data in which the steering torsion bar torque is less than or equal to a fourth preset threshold value from the vehicle driving data; wherein the fourth preset threshold value is greater than 0. Exemplarily, the fourth preset threshold value may be set to 1 Nm (Newton meter). In practical applications, if the steering torsion bar torque EPS1_SteeringAngle (driver hand torque, left positive and right negative) of the steering wheel is less than or equal to 1 Nm, it can be considered that the torque manually applied by the driver is very small, which generally indicates that the driver has little or no intervention in the steering wheel.

[0066] Step S4b: Filter out the second driving data when the turn signal is off from the first driving data. For example, in practical applications, data in which BDCS1_LeftTurnLightSt is in an off state (0x0: Inactive) and BDCS1_RightTurnLightSt is in an off state (0x0: Inactive) can be filtered out.

[0067] Step S4c: Filter out the lateral acceleration signal from the second driving data. For example, in practical applications, the lateral acceleration signal can be filtered out according to ACU2_LatAccSensorValue.

[0068] Step S4d: Calculate the number of dragon drawing times according to the lateral acceleration signal and a preset rule.

[0069] In some embodiments, in step S4d, the step of calculating the number of dragon drawing according to the lateral acceleration signal and a preset rule includes:

[0070] Step S4d1: determining whether the lateral acceleration signal has two peaks and two troughs within a preset time period (e.g., 10 seconds);

[0071] Step S4d2: If the judgment is yes, and the absolute value of at least one peak or at least one trough is greater than the fifth preset threshold, and the difference between one peak and trough is greater than the sixth preset threshold, and the difference between the other peak and trough is greater than the seventh preset threshold, then the number of dragon drawing is accumulated once; wherein the fifth preset threshold is greater than 0, the sixth preset threshold is greater than 0, and the seventh preset threshold is greater than 0. Exemplarily, the fifth preset threshold can be set to 0.5, the sixth preset threshold can be set to 1, and the seventh preset threshold can be set to 0.6.

[0072] Exemplarily, in the above steps S4d1 and S4d2, if the lateral acceleration ACU2_LatAccSensorValue has two peaks and troughs within 10 seconds, wherein the absolute value of at least one peak or trough is greater than 0.5, and the difference between one of the peaks and troughs is greater than 1, and the difference between the other peak and trough is greater than 0.6, it is calculated as one dragon drawing occurs.

[0073] In step S5, the user perception index of the automatic driving system is obtained according to the number of abnormal longitudinal acceleration, the number of abnormal braking and the number of dragon drawing. For example, the driving experience of the intelligent driving system Pilot, such as comfort, safety and user satisfaction, can be comprehensively evaluated according to the user perception index combined with the subjective feedback of the driver and passengers.

[0074] In some embodiments, the step of obtaining a user perception index of the automatic driving system according to the number of abnormal longitudinal acceleration, the number of abnormal braking, and the number of dragon drawing comprises: calculating the user perception index using the following formula (2):

[0075] User sensory index = α*number of longitudinal acceleration abnormalities + β*number of braking abnormalities + γ*number of dragon drawing times (2)

[0076] Here, α represents the first weight coefficient, β represents the second weight coefficient, γ represents the third weight coefficient, and α+β+γ=1.

[0077] This application analyzes the abnormal longitudinal acceleration, abnormal braking, and dragon drawing of the vehicle. Figure 2 is an exemplary result diagram of analyzing abnormal longitudinal acceleration of a vehicle in an embodiment of the present application, with reference to Figure 2 As shown, interface 200 shows the results of the analysis of longitudinal abnormal jerks per 100 kilometers. For example, area 210 shows that 3.2 longitudinal abnormal jerks occurred on October 7, 2023. Figure 3 is an exemplary result diagram of analyzing abnormal braking of a vehicle in an embodiment of the present application, with reference to Figure 3 As shown, interface 300 shows the results of the temporary braking / emergency braking analysis per 100 kilometers. For example, area 310 shows that 3.6 temporary braking / emergency braking occurred on October 7, 2023. Figure 4 This is an exemplary result diagram of analyzing the situation of a vehicle drawing a dragon in one embodiment of the present application, refer to Figure 4 As shown, interface 400 shows the results of the trend analysis of dragon drawing per 100 kilometers. For example, area 410 shows that 0.2 dragon drawings occurred on October 14, 2023.

[0078] The technical solution of this application can obtain the vehicle driving data of the autonomous driving system in real time. By calculating the number of abnormal longitudinal acceleration, the smoothness of the driving process can be evaluated; by calculating the number of abnormal braking, the braking system's point braking and emergency braking can be evaluated, and then the safety of the system can be evaluated; by calculating the number of dragon drawing times, the instability of the vehicle's driving trajectory can be evaluated; by combining the number of abnormal longitudinal acceleration, the number of abnormal braking and the number of dragon drawing times, the user's somatosensory index of the autonomous driving system can be obtained, which can comprehensively measure the driving comfort, safety and behavioral norms. This application can detect the abnormal conditions of the autonomous driving system in real time, objectively and comprehensively, and improve the detection and evaluation effect of the autonomous driving system. The user's somatosensory index of this application can provide multi-dimensional guidance for optimizing the performance of the autonomous driving system and improving the user experience.

[0079] In practical applications, the technical solution of this application is used as the Pilot somatosensory index evaluation system of the intelligent driving system. By focusing on the frequency analysis of longitudinal abnormal jerk, point braking / sudden braking, dragon drawing and other behaviors in every 100 kilometers of driving in the activated state, the performance of the intelligent driving system can be comprehensively evaluated, especially in terms of driving comfort, safety and behavior patterns. This application adopts objective and real-time CAN data indicators to reduce subjective interference, ensure the accuracy and richness of the evaluation, and support real-time monitoring and performance optimization. This application not only improves the safety and performance detection effect of the intelligent driving system, but also realizes the rapid response and analysis of driving behavior through a highly timely data upload and recording system. This real-time performance ensures that abnormal situations can be captured in a timely manner, providing the system with timely feedback and adjustment basis. At the same time, the accumulated data continuously promotes the system's adaptability to different driving situations and enhances its reliability and safety.

[0080] The core advantages of the evaluation system of this application include: (1) Objectivity and accuracy: The evaluation is based on CAN data, eliminating subjective interference, accurately recording and analyzing various driving behavior indicators, and providing solid data support for system performance. (2) Real-time: Through real-time monitoring and CAN data judgment, any abnormalities during driving can be captured immediately, which helps to provide early warning and improve driving safety. (3) Comprehensiveness: Covering indicators such as longitudinal abnormal jerk, point braking / sudden braking, and dragon drawing, it provides a comprehensive system performance evaluation to help identify and solve the root causes of performance problems. (4) Data-driven decision-making: Data-based evaluation prompts the system to make precise adjustments, optimize performance according to actual driving conditions, and improve the driving experience. (5) User experience improvement: By focusing on driving comfort and safety, it effectively reduces driving discomfort and improves the overall comfort of drivers and passengers.

[0081] The present application also includes an abnormality detection system for an autonomous driving system, comprising a memory and a processor, wherein the memory is used to store instructions executable by the processor; and the processor is used to execute the instructions to implement the abnormality detection method for the autonomous driving system described above.

[0082] Figure 5 is a system block diagram of an abnormality detection system of an automatic driving system according to an embodiment of the present application. Figure 5 As shown, the abnormality detection system 500 of the autonomous driving system may include a vehicle system 510 and a server computer 520, and the vehicle system 510 exchanges data with the server computer 520 in a wireless manner through a communication network 530. The vehicle system 510 includes a vehicle communication port 511, a vehicle processor 512, a read-only memory (ROM) 513, and a random access memory (RAM) 514. In some embodiments, the vehicle processor 512 may be composed of one or more processors. The vehicle communication port 511 can realize data communication between the vehicle system 510 and the outside. The vehicle system 510 may also include a program storage unit and a data storage unit in different forms, such as a read-only memory (ROM) 513 and a random access memory (RAM) 514, which can store various data files used for processing and / or communication by the vehicle system 510, and possible program instructions executed by the vehicle processor 512. The vehicle processor 512 executes these instructions to implement the main part of the method. The results processed by the vehicle processor 512 are transmitted to the user device through the vehicle communication port 511 and displayed on the user interface.

[0083] The above-mentioned abnormality detection method of the autonomous driving system can be implemented as a processor program, stored in the program storage unit of the vehicle system 510, and can be loaded into the vehicle processor 512 for execution to implement the abnormality detection method of the autonomous driving system of the present application.

[0084] The present application also includes a computer-readable medium storing a computer program code, which, when executed by a processor, implements the above-mentioned abnormality detection method for the autonomous driving system.

[0085] When the abnormality detection method of the autonomous driving system is implemented as a computer program, it can also be stored in a computer-readable storage medium as a product. For example, a computer-readable storage medium may include, but is not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips), optical disks (e.g., compact disks (CDs), digital versatile disks (DVDs)), smart cards, and flash memory devices (e.g., electrically erasable programmable read-only memories (EPROMs), cards, sticks, key drives). In addition, the various storage media described herein can represent one or more devices and / or other machine-readable media for storing information. The term "machine-readable medium" may include, but is not limited to, wireless channels and various other media (and / or storage media) that can store, contain and / or carry code and / or instructions and / or data.

[0086] It should be understood that the embodiments described above are only illustrative. The embodiments described herein may be implemented in hardware, software, firmware, middleware, microcode, or any combination thereof. For hardware implementation, the processor may be implemented in one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, and / or other electronic units designed to perform the functions described herein, or a combination thereof.

[0087] Some aspects of the present application may be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". The processor may be one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DAPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, or combinations thereof. In addition, various aspects of the present application may be expressed as computer products located in one or more computer-readable media, which include computer-readable program codes. For example, computer-readable media may include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, tapes ...), optical disks (e.g., compact disks CDs, digital versatile disks DVDs ...), smart cards, and flash memory devices (e.g., cards, sticks, key drives ...).

[0088] A computer-readable medium may include a propagated data signal containing computer program code, such as in baseband or as part of a carrier wave. The propagated signal may have a variety of manifestations, including electromagnetic, optical, etc., or a suitable combination. A computer-readable medium may be any computer-readable medium other than a computer-readable storage medium, which may be connected to an instruction execution system, device or apparatus to communicate, propagate or transmit a program for use. The program code on the computer-readable medium may be propagated via any suitable medium, including radio, cable, fiber optic cable, radio frequency signal, or similar medium, or any combination of the above mediums.

[0089] The basic concepts have been described above. Obviously, for those skilled in the art, the above application disclosure is only an example and does not constitute a limitation of the present application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements and amendments to the present application. Such modifications, improvements and amendments are suggested in the present application, so such modifications, improvements and amendments still belong to the spirit and scope of the exemplary embodiments of the present application.

[0090] At the same time, the present application uses specific words to describe the embodiments of the present application. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present application can be appropriately combined.

[0091] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of the present application are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

Claims

1. A method for detecting anomalies in an autonomous driving system, characterized in that: include: Acquiring vehicle driving data of the automatic driving system during vehicle driving; Calculating the number of abnormal longitudinal accelerations according to the vehicle driving data; Calculating the number of abnormal braking times according to the vehicle driving data; Calculating the number of dragon drawing times according to the vehicle driving data; A user perception index of the automatic driving system is obtained according to the number of abnormal longitudinal accelerations, the number of abnormal brakings and the number of dragon drawing.

2. The abnormality detection method of the automatic driving system according to claim 1, characterized in that: The step of calculating the number of abnormal longitudinal accelerations according to the vehicle driving data comprises: Filtering longitudinal vehicle control mode data from the vehicle driving data; Filtering a longitudinal acceleration signal from the longitudinal vehicle control mode data; Performing filtering processing on the longitudinal acceleration signal to obtain a filtered longitudinal acceleration signal; Calculating a longitudinal acceleration rate of change according to the filtered longitudinal acceleration signal; If the longitudinal acceleration change rate is greater than a first preset threshold, the number of longitudinal acceleration anomalies is accumulated once; wherein the first preset threshold is greater than 0.

3. The abnormality detection method of the automatic driving system according to claim 1, characterized in that: The step of calculating the number of braking abnormalities according to the vehicle driving data comprises: Filtering longitudinal vehicle control mode data from the vehicle driving data; Filtering a longitudinal acceleration signal from the longitudinal vehicle control mode data; Performing filtering processing on the longitudinal acceleration signal to obtain a filtered longitudinal acceleration signal; Calculating a longitudinal acceleration rate of change according to the filtered longitudinal acceleration signal; If the longitudinal acceleration change rate is greater than a second preset threshold and the longitudinal acceleration signal is less than a third preset threshold, the number of braking abnormalities is accumulated once; wherein the second preset threshold is greater than 0 and the third preset threshold is less than 0.

4. The abnormality detection method of the automatic driving system according to claim 2 or 3, characterized in that: The step of filtering the longitudinal acceleration signal includes: performing low-pass filtering on the longitudinal acceleration signal according to a Butterworth filter.

5. The abnormality detection method of the automatic driving system according to claim 2 or 3, characterized in that: The step of calculating the longitudinal acceleration change rate according to the filtered longitudinal acceleration signal comprises: calculating the longitudinal acceleration change rate using the following formula: Longitudinal acceleration change rate = (current frame - previous frame) / (current moment - previous moment) The current frame represents the filtered longitudinal acceleration signal at the current moment, and the previous frame represents the filtered longitudinal acceleration signal at the previous moment.

6. The abnormality detection method of the automatic driving system according to claim 1, characterized in that: The step of calculating the number of dragon drawing times according to the vehicle driving data comprises: Filtering out first driving data in which the steering torsion bar torque is less than or equal to a fourth preset threshold from the vehicle driving data; wherein the fourth preset threshold is greater than 0; Filtering out second driving data when the turn signal lamp is turned off from the first driving data; Filtering a lateral acceleration signal from the second driving data; The number of dragon drawing times is calculated according to the lateral acceleration signal and a preset rule.

7. The abnormality detection method of the automatic driving system according to claim 6, characterized in that: The step of calculating the number of dragon drawing according to the lateral acceleration signal and a preset rule comprises: Determining whether the lateral acceleration signal has two peaks and two troughs within a preset time period; If the judgment is yes, and the absolute value of at least one peak or at least one trough is greater than the fifth preset threshold, and the difference between one peak and trough is greater than the sixth preset threshold, and the difference between the other peak and trough is greater than the seventh preset threshold, then the number of dragon drawing times is accumulated once; wherein, the fifth preset threshold is greater than 0, the sixth preset threshold is greater than 0, and the seventh preset threshold is greater than 0.

8. The abnormality detection method of the automatic driving system according to claim 1, characterized in that: The step of obtaining the user perception index of the automatic driving system according to the number of abnormal longitudinal acceleration, the number of abnormal braking and the number of dragon drawing comprises: calculating the user perception index by the following formula: user perception index = α*number of abnormal longitudinal acceleration+β*number of abnormal braking+γ*number of dragon drawing Wherein, α represents the first weight coefficient, β represents the second weight coefficient, γ represents the third weight coefficient, and α+β+γ=1.

9. An abnormality detection system for an autonomous driving system, characterized in that: include: a memory for storing instructions executable by a processor; A processor, configured to execute the instructions to implement the abnormality detection method for the autonomous driving system as described in any one of claims 1 to 8.

10. A computer readable medium storing computer program code, characterized in that: When executed by a processor, the computer program code implements the abnormality detection method for the autonomous driving system as described in any one of claims 1 to 8.