A method, apparatus, device and storage medium for determining vehicle failure

Abstract

The application discloses a kind of method for determining vehicle fault, device, equipment and storage medium.The method comprises: according to the expected data of target vehicle and real-time data, determine control abnormal factor;Desired data, real-time data and control abnormal factor are input into preset neural network model, and data label is obtained;According to data label, determine whether target vehicle is abnormal, if yes, according to control abnormal factor and real-time data, determine target driving correlation index from the selected driving correlation index of target vehicle;Target data corresponding to target driving correlation index is collected, and the fault of target vehicle is determined according to target data.The technical scheme of the embodiment of the application can filter out the index related to vehicle fault from a large number of selected driving correlation indexes and collect it, which can realize the determination of vehicle fault without collecting a large amount of data, and greatly improve the safety of intelligent driving.

Description

Technical Field

[0001] This invention relates to the field of vehicle control, and more particularly to a method, apparatus, device, and storage medium for determining vehicle faults. Background Technology

[0002] In recent years, with the continuous development of intelligent vehicles, research on intelligent driving data acquisition methods has also deepened, and more and more researchers have begun to devote themselves to related research. When driving a vehicle, intelligent driving data acquisition methods can be used to collect and analyze driving data when the intelligent driving output is abnormal, thereby tracing and locating the cause of the abnormality in the intelligent driving system.

[0003] While using intelligent driving data collection methods can significantly improve the safety of intelligent driving systems, existing intelligent driving data collection methods often suffer from problems such as excessively large data volumes, low collection efficiency, and high costs that make them difficult to promote. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and storage medium for determining vehicle faults, in order to solve the problem of excessively large data volumes collected by existing intelligent driving data acquisition methods.

[0005] In a first aspect, the present invention provides a method for determining a vehicle malfunction, comprising:

[0006] Based on the expected data and real-time data of the target vehicle, control anomaly factors are determined, wherein the expected data includes the expected steering wheel angle and the expected longitudinal acceleration, the real-time data includes the steering wheel angle and the longitudinal acceleration, and the control anomaly factors include the steering wheel angle factor and the longitudinal acceleration factor.

[0007] The expected data, the real-time data, and the control anomaly factor are input into a preset neural network model to obtain data labels;

[0008] Based on the data label, determine whether the target vehicle is abnormal. If so, based on the control anomaly factor and real-time data, determine the target driving correlation index from the candidate driving correlation indexes of the target vehicle.

[0009] Collect target data corresponding to the target driving correlation index, and determine the fault of the target vehicle based on the target data.

[0010] Secondly, the present invention provides an apparatus for determining vehicle malfunctions, comprising:

[0011] The factor determination module is used to determine control anomaly factors based on the expected data and real-time data of the target vehicle. The expected data includes the expected steering wheel angle and the expected longitudinal acceleration, the real-time data includes the steering wheel angle and the longitudinal acceleration, and the control anomaly factors include the steering wheel angle factor and the longitudinal acceleration factor.

[0012] The label determination module is used to input the expected data, the real-time data, and the control anomaly factor into a preset neural network model to obtain data labels;

[0013] The correlation index determination module is used to determine whether the target vehicle is abnormal based on the data label. If so, it determines the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the control anomaly factor and real-time data.

[0014] The fault determination module is used to collect target data corresponding to the target driving correlation index and determine the fault of the target vehicle based on the target data.

[0015] Thirdly, the present invention provides an electronic device comprising:

[0016] At least one processor;

[0017] and memory that is communicatively connected to at least one processor;

[0018] The memory stores a computer program that can be executed by at least one processor, which enables the at least one processor to perform the method for determining vehicle malfunctions described in the first aspect.

[0019] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a processor to execute the method for determining a vehicle fault as described in the first aspect.

[0020] The present invention provides a scheme for determining vehicle faults. Based on expected data and real-time data of a target vehicle, control anomaly factors are determined. The expected data includes expected steering wheel angle and expected longitudinal acceleration, the real-time data includes steering wheel angle and longitudinal acceleration, and the control anomaly factors include steering wheel angle factors and longitudinal acceleration factors. The expected data, the real-time data, and the control anomaly factors are input into a preset neural network model to obtain data labels. Based on the data labels, it is determined whether the target vehicle is abnormal. If so, based on the control anomaly factors and real-time data, a target driving correlation index is determined from the candidate driving correlation indexes of the target vehicle. Target data corresponding to the target driving correlation index is collected, and the fault of the target vehicle is determined based on the target data. By adopting the above technical solution, the expected data, real-time data, and control anomaly factors that can characterize vehicle control features output by the vehicle controller are input into a preset neural network model to obtain data labels that can characterize whether the target vehicle has control anomalies. If the vehicle has anomalies, the target driving correlation index can be selected from the candidate driving correlation indexes based on the control anomaly factors and real-time data. By analyzing the target driving correlation index, the fault of the vehicle can be accurately determined. This method can select and collect indicators related to vehicle faults from a large number of candidate driving correlation indicators. It can determine vehicle faults without collecting a large amount of data, which greatly improves the safety of intelligent driving, ensures the safety of people and vehicles, and reduces the probability of traffic accidents.

[0021] It should be understood that the description in this section is not intended to identify key or essential features of the invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of a method for determining vehicle faults according to Embodiment 1 of the present invention;

[0024] Figure 2 This is a flowchart of a method for determining vehicle faults according to Embodiment 2 of the present invention;

[0025] Figure 3 This is a schematic diagram of a device for determining vehicle malfunctions according to Embodiment 3 of the present invention;

[0026] Figure 4 This is a schematic diagram of the structure of an electronic device provided according to Embodiment 4 of the present invention. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. In the description of this invention, unless otherwise stated, "a plurality of" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist; for example, A and / or B can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0029] Example 1

[0030] Figure 1 The flowchart of a method for determining vehicle faults is provided in Embodiment 1 of the present invention. This embodiment is applicable to situations where vehicle faults are determined. The method can be executed by a device for determining vehicle faults. The device for determining vehicle faults can be implemented in hardware and / or software. The device for determining vehicle faults can be configured in an electronic device, which can be composed of two or more physical entities or a single physical entity.

[0031] like Figure 1 As shown, the method for determining vehicle malfunctions provided in Embodiment 1 of the present invention specifically includes the following steps:

[0032] S101. Based on the expected data and real-time data of the target vehicle, determine the control anomaly factors, wherein the expected data includes the expected steering wheel angle and the expected longitudinal acceleration, the real-time data includes the steering wheel angle and the longitudinal acceleration, and the control anomaly factors include the steering wheel angle factor and the longitudinal acceleration factor.

[0033] In this embodiment, the target vehicle can be understood as the current vehicle whose real-time data is being collected. The expected data output by the target vehicle's (hereinafter referred to as "this vehicle") controller, along with the real-time collected steering wheel angle and longitudinal acceleration, can be used to determine control anomaly factors. For example, if the expected data output by the vehicle's controller is too large or too small, the difference between the expected steering wheel angle and the real-time collected steering wheel angle is determined as the steering wheel angle factor, and the difference between the expected longitudinal acceleration and the real-time collected longitudinal acceleration is determined as the longitudinal acceleration factor. These control anomaly factors can be used to characterize the vehicle's driving characteristics.

[0034] S102. Input the expected data, the real-time data, and the control anomaly factor into a preset neural network model to obtain data labels.

[0035] In this embodiment, since the expected data, real-time data, and control anomaly factors all contain the vehicle's driving characteristics, a pre-trained neural network model can be used to process the above data to obtain data labels characterizing whether the vehicle has any anomalies. These data labels can include labels such as "normal data" and "abnormal data." The neural network model has strong robustness and fault tolerance, can process data in parallel, is computationally fast, has strong information synthesis capabilities, can process both quantitative and qualitative information simultaneously, and can effectively coordinate the relationships between multiple input information. Therefore, it is suitable for determining whether a vehicle has data anomalies.

[0036] S103. Determine whether the target vehicle is abnormal based on the data label. If so, determine the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the control anomaly factor and real-time data.

[0037] In this embodiment, as described above, if the data tag indicates that the data is normal, it can be determined that the vehicle is normal and the further analysis of the vehicle can be ended. If the data tag indicates that the data is abnormal, it can be determined that the vehicle is abnormal and step 104 can be executed. Otherwise, the process can be terminated.

[0038] S104. Based on the control anomaly factor and real-time data, determine the target driving correlation index from the candidate driving correlation indexes of the target vehicle.

[0039] In this embodiment, based on the control anomaly factor and real-time data, such as the magnitude relationship between the control anomaly factor and real-time data, it is possible to determine which indicators of the vehicle are related to the potential faults of the vehicle. These indicators are the target driving-related indicators. For example, if, within a certain time period, the absolute value of the slope of the straight line l1 fitted by the steering wheel angle factor and the actual steering wheel angle is greater than the absolute value of the slope of the straight line l2 fitted by the longitudinal acceleration factor and the actual longitudinal acceleration, and the absolute value of the intercept of l1 is greater than the absolute value of the intercept of l2, then it can be determined that the vehicle has a lateral anomaly. That is, the target driving-related indicator should be the lateral data corresponding to the lateral indicators of the vehicle. Here, the candidate driving-related indicators can be understood as indicators related to the vehicle's driving, such as the current time, the vehicle speed, and the vehicle's lateral acceleration. A lateral anomaly can be understood as an anomaly in the vehicle's lateral data. The vehicle's driving direction can be understood as longitudinal. The vehicle's lateral indicators can be understood as indicators related to the direction perpendicular to the vehicle's driving direction, such as the vehicle's lateral acceleration, vehicle yaw rate, and steering wheel speed.

[0040] S105. Collect target data corresponding to the target driving correlation index, and determine the fault of the target vehicle based on the target data.

[0041] In this embodiment, target data corresponding to the target driving-related indicators can be collected, and the faults existing in the vehicle can be determined by analyzing the target data.

[0042] The method for determining vehicle faults provided in this invention involves determining control anomaly factors based on expected data and real-time data of a target vehicle. The expected data includes expected steering wheel angle and expected longitudinal acceleration, the real-time data includes steering wheel angle and longitudinal acceleration, and the control anomaly factors include steering wheel angle factors and longitudinal acceleration factors. The expected data, the real-time data, and the control anomaly factors are input into a preset neural network model to obtain data labels. Based on the data labels, it is determined whether the target vehicle is abnormal. If so, a target driving correlation index is determined from candidate driving correlation indices of the target vehicle based on the control anomaly factors and real-time data. Target data corresponding to the target driving correlation index is collected, and the fault of the target vehicle is determined based on the target data. The technical solution of this invention involves inputting the expected data, real-time data, and control anomaly factors that characterize the vehicle's driving characteristics into a preset neural network model. This yields data labels that characterize whether the target vehicle has an anomaly. If the vehicle has an anomaly, a target driving-related indicator can be selected from the candidate driving-related indicators based on the control anomaly factors and real-time data. By analyzing this target driving-related indicator, the vehicle's fault can be accurately determined. This method can select and collect indicators related to vehicle faults from a large number of candidate driving-related indicators. It can determine vehicle faults without collecting a large amount of data, significantly improving the safety of intelligent driving, ensuring the safety of people and vehicles, and reducing the probability of traffic accidents.

[0043] Example 2

[0044] Figure 2 This is a flowchart of a method for determining vehicle faults provided in Embodiment 2 of the present invention. The technical solution of the present invention is further optimized based on the above optional technical solutions, and a specific method for determining vehicle faults is given.

[0045] Optionally, determining the control anomaly factor based on the target vehicle's expected data and real-time data includes: determining a first absolute value of the difference between the expected steering wheel angle and the steering wheel angle; determining a second absolute value of the difference between the expected longitudinal acceleration and the longitudinal acceleration; determining a steering wheel angle factor based on the first absolute value and the absolute value of the steering wheel angle; and determining a longitudinal acceleration factor based on the second absolute value and the absolute value of the longitudinal acceleration. The advantage of this configuration is that by utilizing the first and second absolute values, the steering wheel angle factor characterizing the vehicle's steering wheel characteristics and the longitudinal acceleration factor characterizing the vehicle's longitudinal acceleration characteristics can be accurately determined.

[0046] Optionally, determining the target driving correlation index from the candidate driving correlation indices of the target vehicle based on the control anomaly factor and real-time data includes: determining first information of a first fitted straight line based on the steering wheel angle factor and the steering wheel angle, wherein the first information includes a first coefficient and a first intercept; determining second information of a second fitted straight line based on the longitudinal acceleration factor and the longitudinal acceleration, wherein the second information includes a second coefficient and a second intercept; and determining the target driving correlation index from the candidate driving correlation indices of the target vehicle based on the first information and the second information. The advantage of this configuration is that the fitted curve can determine the first information characterizing the steering wheel angle control characteristics and the second information characterizing the longitudinal acceleration control characteristics. Using this first and second information, the target driving correlation index related to vehicle faults can be determined, thereby enabling targeted collection of vehicle data and reducing the memory footprint of cloud data.

[0047] like Figure 2 As shown in Embodiment 2 of the present invention, a method for determining vehicle faults specifically includes the following steps:

[0048] S201. Determine the first absolute value of the difference between the desired steering wheel angle and the steering wheel angle.

[0049] For example, if the desired steering wheel angle is δ AD If the steering wheel angle is δ1, then the first absolute value can be expressed as |δ AD -δ1|, where || represents taking the absolute value.

[0050] S202. Determine the second absolute value of the difference between the expected longitudinal acceleration and the longitudinal acceleration.

[0051] For example, if the desired longitudinal acceleration is a AD If the longitudinal acceleration is a1, then the second absolute value can be expressed as |a1|. AD -a1|.

[0052] S203. Determine the steering wheel angle factor based on the first absolute value and the absolute value of the steering wheel angle, and determine the longitudinal acceleration factor based on the second absolute value and the absolute value of the longitudinal acceleration.

[0053] For example, the steering wheel angle factor F δ It can be represented as:

[0054]

[0055] Where T is a coefficient, which can be taken as 0.001, and F is the longitudinal acceleration factor. a It can be represented as:

[0056]

[0057] Where U is a coefficient, which can be 0.01.

[0058] S204. Input the expected data, the real-time data, and the control anomaly factor into a preset neural network model to obtain data labels.

[0059] Optionally, the method for determining the preset neural network model includes: determining multiple sample data vectors and sample labels for each sample data in each sample data vector, wherein the sample data contained in each sample data vector corresponds one-to-one with the expected data, the real-time data, and the control anomaly factor; determining the vector distance between the sample data vectors and multiple neuron vectors in the preset initial model; determining the neuron weights in the preset initial model based on the vector distances and the sample labels, and determining the preset neural network model based on the neuron weights. The advantage of this setup is that by utilizing the vector distances between the neuron vectors and the sample data vectors, as well as the sample labels, the neuron weights in the preset initial model can be accurately determined, thereby completing the construction of the preset neural network model.

[0060] Specifically, you can first initialize the parameters of each neuron in the preset initial model, such as setting random numbers as the initial values ​​for the weights between the input layer and the mapping layer. Then, obtain multiple sample data vectors, sample data vector X. k It can be represented as X k =(x1,x2,…,x m ) T x j Let j = 1, 2, ..., m, k = 1, 2, ..., n, where m can be 6. The sample data in the sample data vector corresponds one-to-one with the expected data, real-time data, and control anomaly factors. For example, x1 can be the expected sample data for steering wheel angle, x2 can be the expected sample data for longitudinal acceleration, ..., x... m This can be longitudinal acceleration factor sample data, meaning the dimension of each neuron vector in the initial model is pre-defined to match the dimension of the sample data vector. Next, the sample data vector X... kThe data is sequentially input into the preset initial model. For each sample data vector, the vector distance between the sample data vector and each neuron vector in the preset initial model is calculated. Using this vector distance and the sample label, the neuron weight of each neuron in the preset initial model can be determined. For example, if the sample label is 0 and 1, where 0 indicates normal data and 1 indicates abnormal data, when the label of the output of the neuron corresponding to the shortest vector distance matches the sample label, the result of mathematical operations (such as addition, subtraction, multiplication, and division) between the previous sample data vector (e.g., if the current sample data vector is X2, then the previous sample data vector is X1) and the neuron vector corresponding to the shortest vector distance can be determined as the neuron weight corresponding to the shortest vector distance. Finally, after obtaining the neuron weight of each neuron in the preset initial model, the preset neural network model is obtained.

[0061] Furthermore, determining the neuron weights in the preset initial model based on the vector distance and the sample label includes: determining the minimum vector distance corresponding to each sample data vector for multiple vector distances associated with each sample data vector, and identifying the neuron corresponding to the minimum vector distance as the winning neuron; determining a learning factor based on the sample label and the training data label output by the winning neuron, and determining the neuron weights in the preset initial model based on the learning factor and the weights of the winning neuron. The advantage of this setup is that by filtering for the minimum vector distance, the neuron closest to the sample data vector can be identified, which is the winning neuron. Then, using the learning factor and a large number of sample data vectors, the weights of the winning neuron are iteratively calculated, thereby obtaining the neuron weights of each neuron in the preset initial model.

[0062] For example, as described above, the sample data vector X can be... k Input the data into the preset initial model sequentially. For each sample data vector, calculate the vector distance between the sample data vector and each neuron vector in the preset initial model, and then select the sample data vector X. k The neuron corresponding to the minimum vector distance is identified as the winning neuron. Vector distance d t The determination method can be expressed as follows:

[0063]

[0064] x k,j (t) represents the j-th sample data in the k-th sample data vector, ω i,j (t) represents the j-th sample data in the vector of the i-th neuron, and ∑ represents the summation operation. Let represent the square root operation, o represent the total number of neuron vectors, and t represent the number of epochs of the input sample data vector, which is generally consistent with the value of k. The learning factor η(t) can be determined as follows:

[0065]

[0066] In this system, A, B, and C are coefficients. A can be 0.2, C can be 10000, and the value of B is related to the label output by the winning neuron and the sample label. For example, if the label output by the winning neuron matches the sample label, B can be 1; otherwise, it can be -1. The neuron weight ω of the winning neuron... j* It can be expressed as follows:

[0067] ω j* (t+1)=ω j* (t)+η(t)*[x(t)-ω j* (t)]

[0068] Where, ω j* (t+1) represents the current neuron weight, ω j* x(t) represents the neuron weights determined in the previous round, and x(t) represents the sample data vector in the current round, i.e., x k , k = t.

[0069] S205. Determine whether the target vehicle is abnormal based on the data tag. If yes, proceed to step 206; otherwise, end the process.

[0070] S206. Determine the first information of the first fitted straight line based on the steering wheel angle factor and the steering wheel angle.

[0071] The first information includes a first coefficient and a first intercept.

[0072] For example, the steering wheel angle can be continuously collected at fixed time intervals, such as 0.02 seconds, for six consecutive moments to obtain six steering wheel angle factors. These factors are then combined to obtain the vector y_lateral. The following program can then be entered into MATLAB to determine the first coefficient k_lateral (i.e., the slope) and the first intercept b_lateral of the first fitted straight line:

[0073] a1 = 1:6;

[0074] y_lateral = [0.11, 0.12, 0.15, 0.17, 0.20, 0.24]; % Example

[0075] numb1=polyfit(a1,y_lateral,1);

[0076] k_lateral = numb1(1);

[0077] b_lateral = numb1(2);

[0078] Here, polyfit() is a MATLAB function used for curve fitting, and a1 represents the longitudinal acceleration of the vehicle at the six time points.

[0079] S207. Determine the second information of the second fitted straight line based on the longitudinal acceleration factor and the longitudinal acceleration.

[0080] The second information includes the second coefficient and the second intercept.

[0081] For example, longitudinal acceleration can be continuously collected at fixed time intervals for six consecutive moments to obtain six longitudinal acceleration factors. These factors can then be combined to obtain a vector y_long. The following program can then be entered into MATLAB to determine the second coefficient k_long (i.e., the slope) and the second intercept b_long of the second fitted line:

[0082] a2 = 1:6;

[0083] y_long = [0.12, 0.13, 0.15, 0.16, 0.19, 0.22]; % Example

[0084] numb2=polyfit(a2,y_long,1);

[0085] k_long = numb2(1);

[0086] b_long = numb2(2)

[0087] Where a2 represents the longitudinal acceleration of the vehicle at the six time points.

[0088] S208. Based on the first information and the second information, determine the target driving correlation index from the candidate driving correlation indexes of the target vehicle.

[0089] Specifically, the target driving correlation index can be determined based on the relationship between the first and second information. For example, if the absolute value of the first coefficient is greater than the absolute value of the second coefficient, and the absolute value of the first intercept is greater than the absolute value of the second intercept, then it can be determined that the vehicle may have a lateral anomaly, that is, the target driving correlation index should be the lateral data corresponding to the vehicle's lateral index.

[0090] Optionally, determining the target driving correlation index from the candidate driving correlation indices of the target vehicle based on the first information and the second information includes: if the absolute value of the first coefficient is greater than the absolute value of the second coefficient, and the absolute value of the first intercept is greater than the absolute value of the second intercept, then determining a first target driving correlation index from the candidate driving correlation indices of the target vehicle; if the absolute value of the first coefficient is less than the absolute value of the second coefficient, and the absolute value of the first intercept is less than the absolute value of the second intercept, then determining a second target driving correlation index from the candidate driving correlation indices of the target vehicle, wherein the first target driving correlation index includes the lateral index of the target vehicle, and the second target driving correlation index includes the longitudinal index of the target vehicle. The advantage of this setup is that by calculating the relationship between the coefficients of the two fitted lines and the relationship between the intercepts of the two fitted lines, it is possible to accurately determine whether the lateral data or the longitudinal data of the vehicle is abnormal.

[0091] Specifically, longitudinal indicators can be understood as indicators that are related to the direction parallel to the vehicle's direction of travel, such as the vehicle's longitudinal acceleration, accelerator pedal travel, and brake pedal travel.

[0092] For example, as described above, if the candidate driving-related indicators include: current time (common indicator), vehicle speed (common indicator), vehicle longitudinal acceleration (longitudinal indicator), vehicle lateral acceleration (lateral indicator), vehicle yaw rate (lateral indicator), accelerator pedal travel (longitudinal indicator), brake pedal travel (longitudinal indicator), master cylinder pressure (longitudinal indicator), steering wheel angle (lateral indicator), steering wheel speed (lateral indicator), steering wheel torque (lateral indicator), vehicle gear (common indicator), vehicle stationary state (longitudinal indicator), start button status (common indicator), preceding vehicle type (longitudinal indicator), preceding vehicle stationary state (longitudinal indicator), preceding vehicle distance (longitudinal indicator), preceding vehicle speed (longitudinal indicator), and preceding vehicle acceleration (longitudinal indicator). The following parameters are considered: left lane line quality (lateral indicator), right lane line quality (lateral indicator), left lane line color (lateral indicator), right lane line color (lateral indicator), left lane line type (lateral indicator), right lane line type (lateral indicator), distance from the center of the vehicle's front bumper to the left lane line (lateral indicator), distance from the center of the vehicle's front bumper to the right lane line (lateral indicator), left lane line curvature (lateral indicator), right lane line curvature (lateral indicator), expected steering wheel angle (lateral indicator), expected longitudinal acceleration (longitudinal indicator), autonomous driving control status (common indicator), dynamic distance threshold for emergency operation activation (common indicator), steering wheel angle factor (common indicator), and longitudinal acceleration factor (common indicator).

[0093] When |k_lateral|>|k_long| and |b_lateral|>|b_long|, a first target driving correlation indicator can be determined from the candidate driving correlation indicators. This first target driving correlation indicator includes lateral indicators and common indicators. When |k_lateral|<|k_long| and |b_lateral|<|b_long|, a second target driving correlation indicator can be determined from the candidate driving correlation indicators. This second target driving correlation indicator includes longitudinal indicators and common indicators. Common indicators can be understood as indicators related to vehicle movement but independent of driving direction.

[0094] S209. Collect target data corresponding to the target driving correlation index, and determine the fault of the target vehicle based on the target data.

[0095] The method for determining vehicle faults provided in this invention can accurately determine the steering wheel angle factor, which characterizes the steering wheel characteristics of the vehicle, and the longitudinal acceleration factor, which characterizes the longitudinal acceleration characteristics of the vehicle, by using a first absolute value and a second absolute value. Then, by fitting a curve, first information characterizing the steering wheel angle control characteristics and second information characterizing the longitudinal acceleration control characteristics can be determined. Using the first and second information, target driving correlation indicators related to vehicle faults can be determined, thereby enabling targeted collection of vehicle data. This reduces the memory usage of cloud data, ensures the safety of people and vehicles, and reduces the probability of traffic accidents.

[0096] Optionally, based on the above embodiments, the method may further include: determining a dynamic distance threshold based on real-time vehicle speed and real-time longitudinal acceleration, and determining a dynamic lateral acceleration threshold based on the real-time vehicle speed of the target vehicle, wherein the real-time vehicle speed includes the real-time speed of the target vehicle and the real-time speed of the preceding vehicle, and the real-time longitudinal acceleration includes the real-time longitudinal acceleration of the target vehicle and the real-time longitudinal acceleration of the preceding vehicle; if the actual following distance of the target vehicle is greater than the dynamic distance threshold, or the actual lateral acceleration is greater than the dynamic lateral acceleration threshold, then the candidate driving correlation index is collected, and the fault of the target vehicle is determined based on the candidate driving correlation index. The advantage of this setup is that by calculating the dynamic lateral acceleration threshold and the dynamic distance threshold, traffic accidents caused by excessively close following distances or rollovers can be avoided.

[0097] For example, dynamic distance threshold s e The determination method can be as follows:

[0098]

[0099] Where t1 and t2 are the brake dead zone time and deceleration rise time, respectively; v1 is the real-time speed of this vehicle; v2 is the real-time speed of the vehicle in front; and a 1max a1 is the maximum deceleration of this vehicle during emergency braking, a2 is the current longitudinal acceleration of the vehicle in front, d0 is the buffer distance, and d... c For calibration distance. Dynamic lateral acceleration threshold a 1ya The determination method can be as follows:

[0100]

[0101] Where D, E, F, and G are coefficients, D can be 2.5, E can be 60, F can be 60, and G can be 2.2. `max{}` represents the maximum value. If the actual following distance of the vehicle is greater than the dynamic distance threshold (indicating the vehicle is following too closely), or if the actual lateral acceleration of the vehicle is greater than the dynamic lateral acceleration threshold (indicating a risk of rollover), then candidate driving-related indicators can be immediately collected, and the potential risks or existing malfunctions of the target vehicle can be determined based on these indicators. These candidate driving-related indicators include lateral indicators, longitudinal indicators, and common indicators.

[0102] Optionally, vehicle occupants, such as the driver, can initiate the collection of candidate driving-related indicators through a preset method, such as pressing a start button, to determine the target vehicle's fault based on these indicators. Maintenance personnel can remotely initiate the collection of these indicators via a remote backend. The advantage of this setup is that both vehicle occupants and remote maintenance personnel can initiate data collection immediately based on the vehicle's actual condition, ensuring timely handling of vehicle faults.

[0103] Example 3

[0104] Figure 3 This is a schematic diagram of a device for determining vehicle malfunctions according to Embodiment 3 of the present invention. Figure 3 As shown, the device includes: a factor determination module 301, a label determination module 302, a correlation index determination module 303, and a fault determination module 304, wherein:

[0105] The factor determination module is used to determine control anomaly factors based on the expected data and real-time data of the target vehicle. The expected data includes the expected steering wheel angle and the expected longitudinal acceleration, the real-time data includes the steering wheel angle and the longitudinal acceleration, and the control anomaly factors include the steering wheel angle factor and the longitudinal acceleration factor.

[0106] The label determination module is used to input the expected data, the real-time data, and the control anomaly factor into a preset neural network model to obtain data labels;

[0107] The correlation index determination module is used to determine whether the target vehicle is abnormal based on the data label. If so, it determines the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the control anomaly factor and real-time data.

[0108] The fault determination module is used to collect target data corresponding to the target driving correlation index and determine the fault of the target vehicle based on the target data.

[0109] The device for determining vehicle faults provided in this invention inputs the expected data, real-time data, and control anomaly factors that characterize vehicle control features from the vehicle controller into a preset neural network model to obtain data labels that characterize whether the target vehicle has control anomalies. If the vehicle has an anomaly, a target driving correlation index can be selected from the candidate driving correlation indexes based on the control anomaly factors and real-time data. By analyzing the target driving correlation index, the fault of the vehicle can be accurately determined. This method can select and collect indicators related to vehicle faults from a large number of candidate driving correlation indicators. It can determine vehicle faults without collecting a large amount of data, which greatly improves the safety of intelligent driving, ensures the safety of people and vehicles, and reduces the probability of traffic accidents.

[0110] Optionally, the factor determination module includes:

[0111] The first absolute value determination unit is used to determine the first absolute value of the difference between the expected steering wheel angle and the steering wheel angle.

[0112] The second absolute value determination unit is used to determine the second absolute value of the difference between the longitudinal acceleration expectation and the longitudinal acceleration;

[0113] An acceleration factor determination unit is used to determine a steering wheel angle factor based on the first absolute value and the absolute value of the steering wheel angle, and to determine a longitudinal acceleration factor based on the second absolute value and the absolute value of the longitudinal acceleration.

[0114] Optionally, the method for determining the preset neural network model includes:

[0115] Multiple sample data vectors are determined, along with a sample label for each sample data in each sample data vector. Each sample data in each sample data vector corresponds one-to-one with the expected data, the real-time data, and the control anomaly factor. The vector distance between the sample data vectors and multiple neuron vectors in a preset initial model is determined. Based on the vector distance and the sample labels, the neuron weights in the preset initial model are determined, and a preset neural network model is determined based on the neuron weights.

[0116] Furthermore, determining the neuron weights in the preset initial model based on the vector distance and the sample label includes: determining the minimum vector distance corresponding to each sample data vector for multiple vector distances associated with each sample data vector, and determining the neuron corresponding to the minimum vector distance as the winning neuron; determining a learning factor based on the sample label and the training data label output by the winning neuron, and determining the neuron weights in the preset initial model based on the learning factor and the weights of the winning neuron.

[0117] Optionally, the correlation indicator determination module includes:

[0118] The first information determining unit is used to determine first information of the first fitted straight line based on the steering wheel angle factor and the steering wheel angle, wherein the first information includes a first coefficient and a first intercept;

[0119] The second information determining unit is used to determine second information of the second fitted line based on the longitudinal acceleration factor and the longitudinal acceleration, wherein the second information includes a second coefficient and a second intercept;

[0120] The correlation index determination unit is used to determine the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the first information and the second information.

[0121] Optionally, determining the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the first information and the second information includes: if the absolute value of the first coefficient is greater than the absolute value of the second coefficient, and the absolute value of the first intercept is greater than the absolute value of the second intercept, then determining a first target driving correlation index from the candidate driving correlation indexes of the target vehicle; if the absolute value of the first coefficient is less than the absolute value of the second coefficient, and the absolute value of the first intercept is less than the absolute value of the second intercept, then determining a second target driving correlation index from the candidate driving correlation indexes of the target vehicle, wherein the first target driving correlation index includes the lateral index of the target vehicle, and the second target driving correlation index includes the longitudinal index of the target vehicle.

[0122] Optionally, the device may also include:

[0123] An acceleration threshold determination module is used to determine a dynamic distance threshold based on real-time vehicle speed and real-time longitudinal acceleration, and to determine a dynamic lateral acceleration threshold based on the real-time vehicle speed of the target vehicle, wherein the real-time vehicle speed includes the real-time vehicle speed of the target vehicle and the real-time vehicle speed of the preceding vehicle, and the real-time longitudinal acceleration includes the real-time longitudinal acceleration of the target vehicle and the real-time longitudinal acceleration of the preceding vehicle.

[0124] The fault analysis module is used to collect the candidate driving correlation indicators and determine the fault of the target vehicle based on the candidate driving correlation indicators if the actual following distance of the target vehicle is greater than the dynamic distance threshold, or the actual lateral acceleration is greater than the dynamic lateral acceleration threshold.

[0125] The device for determining vehicle faults provided in the embodiments of the present invention can execute the method for determining vehicle faults provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0126] Example 4

[0127] Figure 4 A schematic diagram of an electronic device 40 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0128] like Figure 4 As shown, the electronic device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42 or a random access memory (RAM) 43, communicatively connected to the at least one processor 41. The memory stores computer programs executable by the at least one processor. The processor 41 can perform various appropriate actions and processes based on the computer program stored in the ROM 42 or loaded into the RAM 43 from storage unit 48. The RAM 43 may also store various programs and data required for the operation of the electronic device 40. The processor 41, ROM 42, and RAM 43 are interconnected via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0129] Multiple components in electronic device 40 are connected to I / O interface 45, including: input unit 46, such as keyboard, mouse, etc.; output unit 47, such as various types of monitors, speakers, etc.; storage unit 48, such as disk, optical disk, etc.; and communication unit 49, such as network card, modem, wireless transceiver, etc. Communication unit 49 allows electronic device 40 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0130] Processor 41 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 41 performs the various methods and processes described above, such as methods for determining vehicle faults.

[0131] In some embodiments, the method for determining a vehicle malfunction may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 40 via ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the method for determining a vehicle malfunction described above may be performed. Alternatively, in other embodiments, processor 41 may be configured to perform the method for determining a vehicle malfunction by any other suitable means (e.g., by means of firmware).

[0132] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0133] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0134] The computer equipment provided above can be used to execute the method for determining vehicle faults provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0135] Example 5

[0136] In the context of this invention, the computer-readable storage medium may be a tangible medium, and the computer-executable instructions, when executed by a computer processor, are used to perform a method for determining a vehicle malfunction, the method comprising:

[0137] Based on the expected data and real-time data of the target vehicle, control anomaly factors are determined, wherein the expected data includes the expected steering wheel angle and the expected longitudinal acceleration, the real-time data includes the steering wheel angle and the longitudinal acceleration, and the control anomaly factors include the steering wheel angle factor and the longitudinal acceleration factor.

[0138] The expected data, the real-time data, and the control anomaly factor are input into a preset neural network model to obtain data labels;

[0139] Based on the data label, determine whether the target vehicle is abnormal. If so, based on the control anomaly factor and real-time data, determine the target driving correlation index from the candidate driving correlation indexes of the target vehicle.

[0140] Collect target data corresponding to the target driving correlation index, and determine the fault of the target vehicle based on the target data.

[0141] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by, or in conjunction with, an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0142] The computer equipment provided above can be used to execute the method for determining vehicle faults provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0143] It is worth noting that in the embodiments of the above-described device for determining vehicle malfunctions, the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the scope of protection of the present invention.

[0144] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A method for determining vehicle malfunctions, characterized in that, include: Based on the expected data and real-time data of the target vehicle, control anomaly factors are determined, wherein the expected data includes the expected steering wheel angle and the expected longitudinal acceleration, the real-time data includes the steering wheel angle and the longitudinal acceleration, and the control anomaly factors include the steering wheel angle factor and the longitudinal acceleration factor. The expected data, the real-time data, and the control anomaly factor are input into a preset neural network model to obtain data labels; Based on the data label, determine whether the target vehicle is abnormal. If so, based on the control anomaly factor and real-time data, determine the target driving correlation index from the candidate driving correlation indexes of the target vehicle. Collect target data corresponding to the target driving correlation index, and determine the fault of the target vehicle based on the target data.

2. The method according to claim 1, characterized in that, The step of determining control anomaly factors based on the target vehicle's expected data and real-time data includes: Determine the first absolute value of the difference between the desired steering wheel angle and the steering wheel angle; Determine a second absolute value of the difference between the expected longitudinal acceleration and the longitudinal acceleration; The steering wheel angle factor is determined based on the first absolute value and the absolute value of the steering wheel angle, and the longitudinal acceleration factor is determined based on the second absolute value and the absolute value of the longitudinal acceleration.

3. The method according to claim 1, characterized in that, The method for determining the preset neural network model includes: Multiple sample data vectors are determined, and a sample label is defined for each sample data in each sample data vector, wherein the sample data contained in each sample data vector corresponds one-to-one with the expected data, the real-time data, and the control anomaly factor. Determine the vector distance between the sample data vector and multiple neuron vectors in the preset initial model; Based on the vector distance and the sample label, the neuron weights in the preset initial model are determined, and the preset neural network model is determined based on the neuron weights.

4. The method according to claim 3, characterized in that, The step of determining the neuron weights in the preset initial model based on the vector distance and the sample label includes: For each sample data vector associated with multiple vector distances, determine the minimum vector distance corresponding to each sample data vector, and determine the neuron corresponding to the minimum vector distance as the winning neuron; Based on the sample labels and the training data labels output by the winning neuron, a learning factor is determined, and based on the learning factor and the weights of the winning neuron, the neuron weights in the preset initial model are determined.

5. The method according to claim 1, characterized in that, The step of determining the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the control anomaly factor and real-time data includes: Based on the steering wheel angle factor and the steering wheel angle, first information of the first fitted straight line is determined, wherein the first information includes a first coefficient and a first intercept; Based on the longitudinal acceleration factor and the longitudinal acceleration, second information of the second fitted line is determined, wherein the second information includes a second coefficient and a second intercept; Based on the first information and the second information, the target driving correlation index is determined from the candidate driving correlation indexes of the target vehicle.

6. The method according to claim 5, characterized in that, The step of determining the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the first information and the second information includes: If the absolute value of the first coefficient is greater than the absolute value of the second coefficient, and the absolute value of the first intercept is greater than the absolute value of the second intercept, then a first target driving correlation index is determined from the candidate driving correlation indexes of the target vehicle. If the absolute value of the first coefficient is less than the absolute value of the second coefficient, and the absolute value of the first intercept is less than the absolute value of the second intercept, then a second target driving correlation index is determined from the candidate driving correlation indexes of the target vehicle, wherein the first target driving correlation index includes the lateral index of the target vehicle, and the second target driving correlation index includes the longitudinal index of the target vehicle.

7. The method according to claim 1, characterized in that, Also includes: A dynamic distance threshold is determined based on the real-time vehicle speed and real-time longitudinal acceleration, and a dynamic lateral acceleration threshold is determined based on the real-time vehicle speed of the target vehicle, wherein the real-time vehicle speed includes the real-time vehicle speed of the target vehicle and the real-time vehicle speed of the preceding vehicle, and the real-time longitudinal acceleration includes the real-time longitudinal acceleration of the target vehicle and the real-time longitudinal acceleration of the preceding vehicle. If the actual following distance of the target vehicle is greater than the dynamic distance threshold, or the actual lateral acceleration is greater than the dynamic lateral acceleration threshold, then the candidate driving correlation index is collected, and the fault of the target vehicle is determined based on the candidate driving correlation index.

8. A device for determining vehicle malfunctions, characterized in that, include: The factor determination module is used to determine control anomaly factors based on the expected data and real-time data of the target vehicle. The expected data includes the expected steering wheel angle and the expected longitudinal acceleration, the real-time data includes the steering wheel angle and the longitudinal acceleration, and the control anomaly factors include the steering wheel angle factor and the longitudinal acceleration factor. The label determination module is used to input the expected data, the real-time data, and the control anomaly factor into a preset neural network model to obtain data labels; The correlation index determination module is used to determine whether the target vehicle is abnormal based on the data label. If so, it determines the target driving correlation index from the candidate driving correlation indexes of the target vehicle based on the control anomaly factor and real-time data. The fault determination module is used to collect target data corresponding to the target driving correlation index and determine the fault of the target vehicle based on the target data.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method for determining a vehicle malfunction as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for determining a vehicle fault as described in any one of claims 1-7.

Images