Method, device, server and vehicle for preventing accelerator pedal from being mistakenly stepped on

Abstract

The application relates to the technical field of vehicles, in particular to a throttle anti-misoperation method and device, a server and a vehicle. The method comprises the following steps: obtaining a driving behavior data set of a vehicle; clustering the driving behavior data set to obtain a plurality of clustering centers, and marking the driving type to which the driving behavior data in the driving behavior data set belongs according to the plurality of clustering centers; training a pre-established anti-misoperation detection model based on the marked driving behavior data set, and delivering the trained anti-misoperation detection model to the vehicle, wherein the vehicle realizes the throttle anti-misoperation function based on the misoperation detection model. Therefore, the problems that the throttle anti-misoperation is usually realized by relying on the vehicle-mounted hardware device, the vehicle-mounted hardware device has limited resources, the anti-throttle misoperation detection reliability is low, and the user experience is poor in the prior art are solved.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a method, device, server, and vehicle for preventing accidental accelerator pedal use. Background Technology

[0002] Among the various causes of traffic accidents, a significant number are due to accidental pressing of the accelerator pedal.

[0003] To prevent accidental accelerator pedal presses, one approach is to calculate the optimal solution for the accelerator pedal press model. This involves collecting accelerator pedal data and storing it in a database. Based on the collected data, a model is built, and the data is then imported into an SVM (Support Vector Machine) trainer to calculate the optimal solution for the accelerator pedal press model.

[0004] However, the related technologies rely on onboard hardware for data processing, the data samples depend on the vehicle's own data, the sample size is low, and they cannot be upgraded via OTA (Over-the-Air Technology), resulting in a poor user experience. Summary of the Invention

[0005] This application provides a method, device, server, and vehicle for preventing accidental accelerator pedal presses, in order to solve the problems in related technologies that usually rely on on-board hardware to prevent accidental accelerator pedal presses. However, the limited resources of on-board hardware make the reliability of accidental accelerator pedal press detection low and the user experience poor.

[0006] The first aspect of this application provides a method for preventing accidental accelerator pedal presses. The method is applied to a server and includes the following steps: acquiring a driving behavior dataset of a vehicle; clustering the driving behavior dataset to obtain multiple cluster centers, and labeling the driving type to which the driving behavior data in the driving behavior dataset belongs based on the multiple cluster centers; training a pre-established anti-accelerator pedal press detection model based on the labeled driving behavior dataset, and sending the trained anti-accelerator pedal press detection model to the vehicle, wherein the vehicle implements the accelerator pedal anti-accelerator pedal press function based on the anti-accelerator pedal press detection model.

[0007] Optionally, the anti-accidental step detection model includes an input layer, an output layer, and a hidden layer, wherein the hidden layer has a first activation function and the output layer has a second activation function.

[0008] Optionally, training the pre-established anti-acceleration detection model based on the labeled driving behavior dataset includes: dividing the labeled driving behavior dataset into a training set, a test set, and a cross-validation set; training the anti-acceleration detection model using the training set, wherein the input layer transmits multiple sample signals from the training set to the hidden layer, the hidden layer performs a weighted summation of the multiple sample signals to obtain the actual output value of the accelerator, and adjusts the weighting coefficients of the hidden layer based on the actual output value and the expected output value of the accelerator; supervising the training using the cross-validation set; and completing the training of the anti-acceleration detection model when the performance of the model reaches a preset performance level based on the test set.

[0009] Optionally, the step of labeling the driving behavior data in the driving behavior dataset to the driving type based on the plurality of cluster centers includes: identifying the coordinate values ​​of each cluster center in the plurality of cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; calculating the sum of squares of the coordinate values ​​of each cluster center based on the accelerator pedal depth, the vehicle direction angle, the vehicle speed, and the vehicle acceleration; designating the cluster center corresponding to the largest sum of squares as the first cluster center, the cluster center corresponding to the smallest sum of squares as the second cluster center, and the remaining cluster centers as the third cluster center; determining the driving type corresponding to each of the first to third cluster centers, and labeling the driving behavior data in the driving behavior dataset to the driving type based on the first to third cluster centers.

[0010] Optionally, the driving behavior dataset includes driving behavior data uploaded by multiple vehicles of the same type, and the driving behavior data includes one or more of the following: accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration.

[0011] A second aspect of this application provides a method for preventing accidental accelerator pedal presses. The method is applied to a vehicle and includes the following steps: acquiring current driving behavior data when the accelerator pedal is triggered; inputting the current driving behavior data into an anti-accelerator pedal press detection model, wherein the anti-accelerator pedal press detection model outputs the actual output value of the accelerator pedal, wherein the anti-accelerator pedal press detection model is trained by a server based on the vehicle's driving behavior dataset; if the actual output value is consistent with the expected output value corresponding to the accelerator pedal, then responding to the accelerator pedal's accelerator signal, otherwise generating an accelerator pedal press warning message.

[0012] A third aspect of this application provides an accelerator pedal misoperation prevention device. The device is applied to a server and includes: a first acquisition module for acquiring a vehicle driving behavior dataset; a labeling module for clustering the driving behavior dataset to obtain multiple cluster centers and labeling the driving behavior data in the driving behavior dataset to the driving type according to the multiple cluster centers; and a training module for training a pre-established anti-accelerator pedal misoperation detection model based on the labeled driving behavior dataset and sending the trained anti-accelerator pedal misoperation detection model to the vehicle, wherein the vehicle implements the accelerator pedal misoperation prevention function based on the misoperation detection model.

[0013] Optionally, the anti-accidental step detection model includes an input layer, an output layer, and a hidden layer, wherein the hidden layer has a first activation function and the output layer has a second activation function.

[0014] Optionally, the training module is further configured to: divide the labeled driving behavior dataset into a training set, a test set, and a cross-validation set; train the anti-accidental pedal detection model using the training set, wherein the input layer transmits multiple sample signals from the training set to the hidden layer, the hidden layer performs a weighted summation of the multiple sample signals to obtain the actual output value of the accelerator, and adjusts the weighting coefficients of the hidden layer based on the actual output value and the expected output value of the accelerator; supervise training using the cross-validation set, and complete the training of the anti-accidental pedal detection model when the performance of the model reaches a preset performance level based on the test set.

[0015] Optionally, the labeling module is further configured to: identify the coordinate values ​​of each cluster center among the plurality of cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; calculate the sum of squares of the coordinate values ​​of each cluster center based on the accelerator pedal depth, the vehicle direction angle, the vehicle speed, and the vehicle acceleration, designate the cluster center corresponding to the largest sum of squares as the first cluster center, the cluster center corresponding to the smallest sum of squares as the second cluster center, and the remaining cluster centers as the third cluster center; determine the driving type corresponding to each of the first to third cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the first to third cluster centers.

[0016] Optionally, the driving behavior dataset includes driving behavior data uploaded by multiple vehicles of the same type, and the driving behavior data includes one or more of the following: accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration.

[0017] A fourth aspect of this application provides an accelerator pedal misoperation prevention device, applied to a vehicle, comprising: a second acquisition module for acquiring current driving behavior data when the accelerator pedal is triggered; an output module for inputting the current driving behavior data into an anti-accelerator pedal misoperation detection model, wherein the anti-accelerator pedal misoperation detection model outputs the actual output value of the accelerator pedal, wherein the anti-accelerator pedal misoperation detection model is trained by a server based on a vehicle driving behavior dataset; and an execution module for responding to the accelerator pedal signal when the actual output value matches the expected output value corresponding to the accelerator pedal, otherwise generating an accelerator pedal misoperation warning message.

[0018] A fifth aspect of this application provides a server, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the accelerator pedal anti-accidental pressing method as described in the above embodiments.

[0019] A sixth aspect of this application provides a server, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the accelerator pedal anti-accidental pressing method as described in the above embodiments.

[0020] Therefore, this application has at least the following beneficial effects:

[0021] This application embodiment can establish an accelerator pedal misoperation detection model and train it using a clustered and labeled vehicle driving behavior dataset. The vehicle can then detect accidental accelerator pedal operation based on the trained model. Furthermore, the analysis algorithm can be optimized to increase the sample data capacity. While improving computational efficiency and reliability, training the model in the cloud reduces vehicle-side hardware and computational costs, thus meeting practical application needs.

[0022] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0023] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0024] Figure 1 This is a flowchart of the accelerator pedal anti-accidental pressing method according to an embodiment of this application;

[0025] Figure 2 This is a schematic diagram of the accelerator pedal misoperation detection system architecture according to an embodiment of this application;

[0026] Figure 3 This is a schematic diagram of data acquisition in an embodiment of this application;

[0027] Figure 4 This is a schematic diagram illustrating the calculation and learning principle of the anti-accidental step detection model in an embodiment of this application;

[0028] Figure 5 This is a flowchart of another method for preventing accidental accelerator pedal press according to an embodiment of this application;

[0029] Figure 6 This is an example diagram of the accelerator pedal anti-accidental pressing device according to an embodiment of this application;

[0030] Figure 7 This is an example diagram of another accelerator pedal anti-accidental pressing device according to an embodiment of this application;

[0031] Figure 8 This is a schematic diagram of the server structure according to an embodiment of this application;

[0032] Figure 9 This is a schematic diagram of the vehicle structure according to an embodiment of this application. Detailed Implementation

[0033] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0034] To address the problems mentioned in the background section, this application provides a method for preventing accidental accelerator pedal press. The following description, with reference to the accompanying drawings, outlines an embodiment of the method, apparatus, server, and vehicle for preventing accidental accelerator pedal press.

[0035] Specifically, Figure 1 This is a flowchart illustrating a method for preventing accidental accelerator pedal press provided in an embodiment of this application.

[0036] like Figure 1 As shown, this method for preventing accidental throttle input is applied to a server and includes the following steps:

[0037] In step S101, the vehicle's driving behavior dataset is obtained.

[0038] The driving behavior dataset includes driving behavior data uploaded by multiple vehicles of the same type. The driving behavior data includes one or more of the following: accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration.

[0039] It is understandable that, such as Figure 2As shown, the embodiments of this application can collect driving data, including accelerator pedal data, brake pedal data, real-time vehicle speed data, vehicle acceleration data, vehicle direction angle and distance data of objects in front and behind, etc. The vehicle driving conditions include but are not limited to uphill, downhill, straight driving, reversing, turning and stationary turning conditions. The data is uploaded to a cloud server and stored.

[0040] It should be noted that the time and number of vehicles for data collection in this application embodiment can be set according to actual conditions, such as collecting data from no fewer than 10 vehicles, with each vehicle's data collected for no less than one month; this application embodiment can use at least one means to acquire vehicle driving behavior data, such as... Figure 3 As shown, the embodiments of this application can collect driving behavior data of the same vehicle model through relevant data acquisition devices and systems.

[0041] In step S102, the driving behavior dataset is clustered to obtain multiple cluster centers, and the driving type to which the driving behavior data in the driving behavior dataset belongs is marked according to the multiple cluster centers.

[0042] It is understood that the embodiments of this application can cluster the obtained driving behavior data to obtain multiple cluster centers, and then label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the cluster centers; wherein, the embodiments of this application can use at least one method to label the driving type, as follows:

[0043] In this embodiment of the application, the driving type to which the driving behavior data in the driving behavior dataset belongs is labeled according to multiple cluster centers, including: identifying the coordinate values ​​of each cluster center in the multiple cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; calculating the sum of squares of the coordinate values ​​of each cluster center according to the accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration, taking the cluster center corresponding to the largest sum of squares of coordinate values ​​as the first cluster center, the cluster center corresponding to the smallest sum of squares of coordinate values ​​as the second cluster center, and the remaining cluster centers as the third cluster center; determining the driving type corresponding to each of the first to third cluster centers, and labeling the driving type to which the driving behavior data in the driving behavior dataset belongs based on the first to third cluster centers.

[0044] It is understood that the embodiments of this application can cluster the driving behavior dataset, setting the number of clusters to three. After the clustering operation is completed, three cluster centers (first to third cluster centers) are obtained. Each center contains four coordinate values, corresponding to accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration, respectively. The sum of the squares of the coordinate values ​​of the four cluster centers (accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration) is calculated. The largest sum is used as the aggressive driving behavior cluster center (first cluster center), the smallest sum is used as the conservative driving behavior cluster center (third cluster center), and the remaining sum is used as the normal driving behavior cluster center (second cluster center). The coordinates of the three cluster centers are used to label the original experimental data. The driving behavior type of the original experimental data is determined by the membership degree of its normalized data with the three cluster centers. That is, the driving state type corresponding to the cluster center with the highest membership degree is the type to which the original experimental data belongs. In this embodiment, the number of labels for the aggressive driving type can be set to "A", the number of labels for the normal driving type can be set to "B", and the number of labels for the conservative driving type can be set to "C".

[0045] It should be noted that this application can also process the driving behavior data of vehicles uploaded to the cloud server by identifying the coordinate values ​​of each cluster center in multiple cluster centers. The processing can include data preprocessing, planning the original experimental data, and filtering outlier data points, as detailed below:

[0046] (1) Data preprocessing: In this embodiment of the application, the driving behavior data of the vehicle uploaded to the cloud server can be filtered for missing data, some abnormal data that deviates from the discrete model can be removed, the data information that needs to be calculated can be selected, and then the data can be classified.

[0047] (2) Planning the original experimental data: The embodiments of this application can further use the following formula for normalization calculation:

[0048]

[0049] Where i is the data number, j is the variable number, and x i,j Let X represent the j-th variable in the i-th unnormalized data set. j Let j represent the set of all variable data values ​​corresponding to j, min represents the minimum value of the variable in the experimental data after removing outliers, and max represents the maximum value of the relevant variable in the experimental data after removing outliers.

[0050] (3) Screening abnormal data points: According to the embodiment of this application, abnormal driving behavior points and normal driving behavior points in the normalized test data can be screened out based on the Euclidean distance between equally spaced test points and normalized test data.

[0051] Specifically, in this embodiment, the Euclidean distance between each equally spaced test point and each normalized test data can be calculated. If the Euclidean distance between an equally spaced test point and one or more normalized test data is less than or equal to a threshold L, then the equally spaced test point is a normal operating condition point; otherwise, it is an abnormal operating condition point. The threshold L can be set as the interval value between two adjacent equally spaced test points. In the following embodiments, the threshold L will be set to 0.05 as an example for illustration.

[0052] In step S103, the pre-established anti-accelerator pedal detection model is trained based on the labeled driving behavior dataset, and the trained anti-accelerator pedal detection model is sent to the vehicle. The vehicle implements the anti-accelerator pedal function based on the anti-accelerator pedal detection model.

[0053] The anti-accidental step detection model includes an input layer, an output layer, and a hidden layer. The hidden layer has a first activation function, and the output layer has a second activation function.

[0054] It is understood that, in this application embodiment, an anti-accelerator pedal misoperation detection model can be pre-established, and the trained model can be distributed to the vehicle based on the trained anti-accelerator pedal misoperation detection model. The vehicle then implements anti-accelerator pedal misoperation prevention based on the received model. In this application embodiment, the model can be distributed and the software upgraded via OTA, without requiring hardware upgrades. This application embodiment can use at least one method to train the anti-accelerator pedal misoperation detection model, as follows:

[0055] In this embodiment, training a pre-established anti-acceleration detection model based on a labeled driving behavior dataset includes: dividing the labeled driving behavior dataset into a training set, a test set, and a cross-validation set; training the anti-acceleration detection model using the training set, wherein the input layer transmits multiple sample signals from the training set to the hidden layer, the hidden layer performs a weighted summation of the multiple sample signals to obtain the actual output value of the accelerator, and adjusts the weighting coefficients of the hidden layer based on the actual output value and the expected output value of the accelerator; training supervision is performed using the cross-validation set, and when the performance of the anti-acceleration detection model reaches the preset performance level, the training of the anti-acceleration detection model is completed.

[0056] The preset performance can be set according to the actual situation, and there are no specific limitations on it.

[0057] It is understood that, in the embodiments of this application, the above-mentioned labeled driving behavior dataset can be divided into training sets, test sets are used to train the anti-accidental step detection model, cross-validation sets are used to test the performance of the anti-accidental step detection model, and cross-validation sets are used for training supervision.

[0058] Specifically, (1) Modeling of the anti-accidental step detection model:

[0059] The embodiments of this application can use training data to perform fusion road feel modeling, and cross-validation is used during training; wherein, when training the fusion road feel model, the input variables include accelerator pedal data, vehicle speed, vehicle acceleration and steering wheel angle, and the output variable is accelerator pedal depth.

[0060] Therefore, the obtained anti-accidental step detection model includes an input layer, an output layer, and a hidden layer. The hidden layer has a first activation function, and the output layer has a second activation function. In the following embodiment, the obtained anti-accidental step detection model can be specified to include one input layer, three output layers, and ten hidden layers, where each hidden layer has ten neurons, and this will be used as an example for illustration.

[0061] In this embodiment, the Sigmoid function can be selected as the first function and the Softmax function as the second function; thereby, the aggressive driving class marker value "A", the normal driving class marker value "B", and the conservative driving class marker value "C" can be output respectively.

[0062] (2) Training the anti-accidental step detection model:

[0063] In this embodiment, the processed test data can be used to train an artificial neural network model. When training the model, the cross-validation ratio can be set to 20%, with the goal of maximizing prediction accuracy. If the prediction accuracy of the obtained cross-validation points is greater than 85%, the model is acceptable, i.e., the modeling is successful. Otherwise, the simulated driving test is repeated.

[0064] It should be noted that the input layer in this embodiment is a feature vector that transmits values ​​to the hidden layer, calculates the results using weights and activation functions, and further transmits the output results down the layer until the entire process ends when a certain number of neuron categories are reached. In this embodiment, the weighting coefficients of the hidden layer can be adjusted based on the actual output value and the expected output value of the throttle. The weights of each layer can be calculated and learned by training an ANN (Artificial Neural Network) algorithm, thereby improving the computational analysis capabilities.

[0065] For example, such as Figure 4 As shown, X1, X2...Xn are the input sample signals, and W1, W2...Wn are the weight coefficients. The input sample signals are applied to u through the weight coefficients to produce the output result WiXi, i.e., u = ΣWiXi = W1X1 + W2X2 + ... + WnXn. The desired output signal Y(t) is then compared with u to generate the error signal e. The weight adjustment mechanism modifies the weight coefficients of the learning system based on the error e, and the modification direction should be to reduce the error e. This process continues until the error e is zero. At this point, the actual output value u and the desired output value Y(t) are exactly the same, and the learning process ends.

[0066] In summary, the accelerator pedal misoperation prevention method proposed in this application can establish an accelerator pedal misoperation prevention detection model and train it using a clustered and labeled vehicle driving behavior dataset. The vehicle can then detect accelerator pedal misoperation based on the trained model. Furthermore, the analysis algorithm can be optimized to increase the sample data capacity, thereby improving computational efficiency and reliability. By training the model in the cloud, the vehicle-side hardware and computational costs can be reduced, meeting the needs of practical use.

[0067] Next, referring to the accompanying drawings, another method for preventing accidental throttle pedal use according to an embodiment of this application is described.

[0068] Figure 5 This is a flowchart illustrating a method for preventing accidental accelerator pedal press provided in an embodiment of this application.

[0069] like Figure 5 As shown, this method for preventing accidental accelerator pedal presses when applied to a vehicle includes the following steps:

[0070] In step S101, the current driving behavior data when the accelerator pedal is triggered is obtained.

[0071] It is understood that the embodiments of this application can obtain current driving behavior data when the accelerator pedal is actually triggered; wherein, the current driving behavior data can be collected and transmitted by relevant data acquisition devices and systems, etc., and there is no specific limitation thereto.

[0072] In step S102, the current driving behavior data is input into the anti-acceleration detection model, and the anti-acceleration detection model outputs the actual output value of the accelerator. The anti-acceleration detection model is trained by the server based on the vehicle's driving behavior dataset.

[0073] It is understood that, in this embodiment of the application, after obtaining the anti-accidental step detection model, the current actual driving behavior data can be input into the model to obtain the actual output value; wherein, in this embodiment of the application, after obtaining the data processing result, the data processing result can be upgraded to the vehicle via OTA.

[0074] In step S103, if the actual output value is consistent with the expected output value corresponding to the accelerator pedal, the accelerator pedal throttle signal is responded to; otherwise, a reminder message for accidental accelerator pedal press is generated.

[0075] It is understood that, in this embodiment of the application, after the vehicle's accelerator pedal misoperation prevention system is activated, each time an accelerator signal is input, the on-board ECU (Electronic Control Unit) software determines whether the input value matches the algorithm analysis result; when the actual output value matches the expected output value corresponding to the accelerator pedal, the accelerator signal is output, and the corresponding pedal accelerator signal is activated; when the actual output value does not match the expected output value corresponding to the accelerator pedal, a warning signal is output, that is, a reminder message for accelerator pedal misoperation is generated; wherein, the reminder message can achieve the reminder function in at least one way, and there is no specific limitation on this.

[0076] According to the accelerator pedal mis-pressing method proposed in the embodiments of this application, the current driving behavior data can be input into the mis-pressing detection model, and the actual output value of the mis-pressing detection model can be compared with the expected output value corresponding to the accelerator pedal to determine whether the accelerator pedal mis-pressing has occurred. Since the data analysis of the embodiments of this application is performed in the cloud, the computing efficiency and output value reliability can be improved, thus meeting the safety requirements of actual accelerator pedal mis-pressing detection.

[0077] The accelerator pedal anti-accidental pressing device proposed according to the embodiments of this application is described below with reference to the accompanying drawings.

[0078] Figure 6 This is a block diagram of the accelerator pedal anti-accidental pressing device according to an embodiment of this application.

[0079] like Figure 6 As shown, the accelerator pedal anti-accidental pressing device 10 is applied to a server and includes: a first acquisition module 110.

[0080] The first acquisition module 110 is used to acquire the driving behavior dataset of the vehicle; the labeling module 120 is used to cluster the driving behavior dataset to obtain multiple cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the multiple cluster centers; the training module 130 is used to train the pre-established anti-accelerator pedal detection model based on the labeled driving behavior dataset, and send the trained anti-accelerator pedal detection model to the vehicle, wherein the vehicle implements the anti-accelerator pedal function based on the anti-accelerator pedal detection model.

[0081] In this embodiment, the anti-accidental step detection model includes an input layer, an output layer, and a hidden layer. The hidden layer has a first activation function, and the output layer has a second activation function.

[0082] In this embodiment, the training module 130 is further configured to: divide the labeled driving behavior dataset into a training set, a test set, and a cross-validation set; train the anti-accidental pedal detection model using the training set, wherein the input layer transmits multiple sample signals from the training set to the hidden layer, the hidden layer performs a weighted summation of the multiple sample signals to obtain the actual output value of the accelerator, and adjusts the weighting coefficients of the hidden layer based on the actual output value and the expected output value of the accelerator; supervise the training using the cross-validation set, and complete the training of the anti-accidental pedal detection model when the performance of the anti-accidental pedal detection model reaches the preset performance based on the test set.

[0083] In this embodiment, the labeling module 120 is further configured to: identify the coordinate values ​​of each cluster center among multiple cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; calculate the sum of squares of the coordinate values ​​of each cluster center based on the accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration, designate the cluster center corresponding to the largest sum of squares as the first cluster center, designate the cluster center corresponding to the smallest sum of squares as the second cluster center, and designate the remaining cluster centers as the third cluster center; determine the driving type corresponding to each of the first to third cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the first to third cluster centers.

[0084] In this embodiment of the application, the driving behavior dataset includes driving behavior data uploaded by multiple vehicles of the same type. The driving behavior data includes one or more of the following: accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration.

[0085] It should be noted that the foregoing explanation of the embodiment of the method for preventing accidental throttle pedal use also applies to the throttle pedal prevention device of this embodiment, and will not be repeated here.

[0086] According to the accelerator pedal misoperation prevention device proposed in the embodiments of this application, an accelerator pedal misoperation prevention detection model can be established and trained using a clustered and labeled vehicle driving behavior dataset. The vehicle can detect accelerator pedal misoperation based on the trained anti-misoperation detection model. Furthermore, the analysis algorithm can be optimized to increase the sample data capacity. While improving computational efficiency and reliability, the model can be trained in the cloud to reduce vehicle-side hardware and computational costs, thus meeting the needs of actual use.

[0087] The accelerator pedal anti-accidental pressing device proposed according to the embodiments of this application is described below with reference to the accompanying drawings.

[0088] Figure 7 This is a block diagram of the accelerator pedal anti-accidental pressing device according to an embodiment of this application.

[0089] like Figure 7As shown, the accelerator pedal anti-accidental pressing device 10 includes: a second acquisition module 210, an output module 220, and an execution module 230.

[0090] The second acquisition module 210 is used to acquire the current driving behavior data when the accelerator pedal is triggered; the output module 220 is used to input the current driving behavior data into the anti-accelerator pedal detection model, and the anti-accelerator pedal detection model outputs the actual output value of the accelerator pedal. The anti-accelerator pedal detection model is trained by the server based on the vehicle's driving behavior dataset; the execution module 230 is used to respond to the accelerator pedal signal when the actual output value is consistent with the expected output value corresponding to the accelerator pedal, otherwise it generates a reminder message for accidental accelerator pedal press.

[0091] It should be noted that the foregoing explanation of the embodiment of the method for preventing accidental throttle pedal use also applies to the throttle pedal prevention device of this embodiment, and will not be repeated here.

[0092] According to the accelerator pedal mis-pressing device proposed in the embodiments of this application, the current driving behavior data can be input into the mis-pressing detection model. The actual output value of the mis-pressing detection model is compared with the expected output value corresponding to the accelerator pedal to determine whether the accelerator pedal has been mis-pressed. Since the data analysis of the embodiments of this application is performed in the cloud, the computing efficiency and output value reliability can be improved, thus meeting the safety requirements of actual accelerator pedal mis-pressing detection.

[0093] Figure 8 This is a schematic diagram of the server structure provided in an embodiment of this application. The server may include: a memory 801, a processor 802, and a computer program stored in the memory 801 and executable on the processor 802; when the processor 802 executes the program, it implements the accelerator pedal anti-accidental pressing method provided in the above embodiment.

[0094] Figure 9 This is a schematic diagram of the structure of a vehicle provided in an embodiment of this application. The vehicle may include: a memory 901, a processor 902, and a computer program stored in the memory 901 and executable on the processor 902; when the processor 902 executes the program, it implements the accelerator pedal anti-accidental pressing method provided in the above embodiment.

[0095] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0096] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0097] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0098] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (FPGAs), field-programmable gate arrays (FPGAs), etc.

[0099] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0100] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for preventing accidental accelerator pedal press, characterized in that, The method is applied to a server, and the method includes the following steps: Obtain a dataset of vehicle driving behavior; The driving behavior dataset is clustered to obtain multiple cluster centers, and the driving type to which the driving behavior data in the driving behavior dataset belongs is marked according to the multiple cluster centers; The pre-established anti-accelerator pedal detection model is trained based on the labeled driving behavior dataset, and the trained anti-accelerator pedal detection model is sent to the vehicle, wherein the vehicle implements the anti-accelerator pedal function based on the anti-accelerator pedal detection model. The anti-accidental step detection model includes an input layer, an output layer, and a hidden layer. The hidden layer has a first activation function, and the output layer has a second activation function. The step of labeling the driving behavior data in the driving behavior dataset according to the multiple cluster centers includes: Identify the coordinate values ​​of each cluster center among the plurality of cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; The sum of squared coordinate values ​​of each cluster center is calculated based on the accelerator pedal depth, the vehicle direction angle, the vehicle speed, and the vehicle acceleration. The cluster center corresponding to the largest sum of squared coordinate values ​​is designated as the first cluster center, the cluster center corresponding to the smallest sum of squared coordinate values ​​is designated as the second cluster center, and the remaining cluster centers are designated as the third cluster center. Determine the driving type corresponding to each of the first to third cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the first to third cluster centers.

2. The method according to claim 1, characterized in that, The pre-established anti-accidental step detection model is trained based on the labeled driving behavior dataset, including: The labeled driving behavior dataset is divided into a training set, a test set, and a cross-validation set; The anti-accidental pedal detection model is trained using the training set, wherein the input layer transmits multiple sample signals from the training set to the hidden layer, the hidden layer performs a weighted summation of the multiple sample signals to obtain the actual output value of the accelerator, and the weighting coefficients of the hidden layer are adjusted based on the actual output value and the expected output value of the accelerator. The training of the anti-accidental step detection model is completed when the performance of the model reaches the preset performance level using the cross-validation set.

3. The method according to any one of claims 1-2, characterized in that, The driving behavior dataset includes driving behavior data uploaded by multiple vehicles of the same type, and the driving behavior data includes one or more of the following: accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration.

4. A method for preventing accidental accelerator pedal press, characterized in that, The method is applied to a vehicle, and the method includes the following steps: Acquire current driving behavior data when the accelerator pedal is triggered; The current driving behavior data is input into the anti-accelerator pedal detection model, and the anti-accelerator pedal detection model outputs the actual output value of the accelerator pedal. The anti-accelerator pedal detection model is trained by the server based on the vehicle's driving behavior dataset. If the actual output value matches the expected output value corresponding to the accelerator pedal, then the accelerator pedal signal is responded to; otherwise, a reminder message for accidental accelerator pedal press is generated. Obtain a vehicle driving behavior dataset, cluster the driving behavior dataset to obtain multiple cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the multiple cluster centers; The anti-accidental step detection model includes an input layer, an output layer, and a hidden layer. The hidden layer has a first activation function, and the output layer has a second activation function. The step of labeling the driving behavior data in the driving behavior dataset according to the multiple cluster centers includes: Identify the coordinate values ​​of each cluster center among the plurality of cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; The sum of squared coordinate values ​​of each cluster center is calculated based on the accelerator pedal depth, the vehicle direction angle, the vehicle speed, and the vehicle acceleration. The cluster center corresponding to the largest sum of squared coordinate values ​​is designated as the first cluster center, the cluster center corresponding to the smallest sum of squared coordinate values ​​is designated as the second cluster center, and the remaining cluster centers are designated as the third cluster center. Determine the driving type corresponding to each of the first to third cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the first to third cluster centers.

5. A device for preventing accidental accelerator pedal press, characterized in that, The device is used in a server, wherein the device includes: The first acquisition module is used to acquire the vehicle's driving behavior dataset; The labeling module is used to cluster the driving behavior dataset to obtain multiple cluster centers, and to label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the multiple cluster centers; The training module is used to train a pre-established anti-accelerator pedal detection model based on a labeled driving behavior dataset, and to distribute the trained anti-accelerator pedal detection model to the vehicle, wherein the vehicle implements the anti-accelerator pedal function based on the anti-accelerator pedal detection model. The anti-accidental step detection model includes an input layer, an output layer, and a hidden layer. The hidden layer has a first activation function, and the output layer has a second activation function. The step of labeling the driving behavior data in the driving behavior dataset according to the multiple cluster centers includes: Identify the coordinate values ​​of each cluster center among the plurality of cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; The sum of squared coordinate values ​​of each cluster center is calculated based on the accelerator pedal depth, the vehicle direction angle, the vehicle speed, and the vehicle acceleration. The cluster center corresponding to the largest sum of squared coordinate values ​​is designated as the first cluster center, the cluster center corresponding to the smallest sum of squared coordinate values ​​is designated as the second cluster center, and the remaining cluster centers are designated as the third cluster center. Determine the driving type corresponding to each of the first to third cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the first to third cluster centers.

6. A device for preventing accidental accelerator pedal press, characterized in that, The device is applied to a vehicle, wherein the device includes: The second acquisition module is used to acquire current driving behavior data when the accelerator pedal is triggered. The output module is used to input the current driving behavior data into the anti-acceleration detection model, and the anti-acceleration detection model outputs the actual output value of the accelerator pedal. The anti-acceleration detection model is trained by the server based on the vehicle's driving behavior dataset. The execution module is used to respond to the accelerator pedal signal when the actual output value is consistent with the expected output value corresponding to the accelerator pedal; otherwise, it generates a reminder message for accidental accelerator pedal press. Obtain a vehicle driving behavior dataset, cluster the driving behavior dataset to obtain multiple cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the multiple cluster centers; The anti-accidental step detection model includes an input layer, an output layer, and a hidden layer. The hidden layer has a first activation function, and the output layer has a second activation function. The step of labeling the driving behavior data in the driving behavior dataset according to the multiple cluster centers includes: Identify the coordinate values ​​of each cluster center among the plurality of cluster centers, wherein the coordinate values ​​include accelerator pedal depth, vehicle direction angle, vehicle speed, and vehicle acceleration; The sum of squared coordinate values ​​of each cluster center is calculated based on the accelerator pedal depth, the vehicle direction angle, the vehicle speed, and the vehicle acceleration. The cluster center corresponding to the largest sum of squared coordinate values ​​is designated as the first cluster center, the cluster center corresponding to the smallest sum of squared coordinate values ​​is designated as the second cluster center, and the remaining cluster centers are designated as the third cluster center. Determine the driving type corresponding to each of the first to third cluster centers, and label the driving type to which the driving behavior data in the driving behavior dataset belongs based on the first to third cluster centers.

7. A server, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the accelerator pedal anti-accidental pressing method as described in any one of claims 1-3.

8. A vehicle, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the accelerator pedal anti-accidental pressing method as described in claim 4.

Images