Vehicle snake running detection method and device, electronic equipment and storage medium

Abstract

The application relates to the technical field of intelligent driving and provides a vehicle snake driving detection method and device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring a vehicle speed sequence and a steering wheel rotation angle sequence; based on a sliding window, extracting key feature points from the steering wheel rotation angle sequence to obtain a rotation angle feature point set, the size of the sliding window being dynamically adjusted based on the vehicle speed sequence; and based on the vehicle speed sequence and the rotation angle feature point set, detecting whether the vehicle has a snake driving behavior. By adopting the dynamic sliding window to extract the key feature points in the steering wheel rotation angle sequence, the key feature points can be more accurately recognized, the detection accuracy is improved, by combining the vehicle speed sequence and the rotation angle feature point set, the vehicle snake driving behavior can be efficiently detected, the detection does not need to rely on image data, and therefore the detection robustness and stability are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving technology, and in particular to a method, apparatus, electronic device, and storage medium for detecting vehicle serpentine driving. Background Technology

[0002] "Hawking" refers to the swaying or back-and-forth movement of autonomous vehicles while driving on a road. This phenomenon can be caused by a variety of factors, including errors in the perception system, imperfect planning algorithms, deviations in control strategies, and vehicle hardware malfunctions. To ensure the safety and reliability of autonomous driving technology, the system needs to detect and suppress sawking in real time to avoid potential harm to passengers.

[0003] Currently, autonomous vehicles generally use cameras to detect lane lines and calculate yaw angles to identify serpentine driving. However, in poor road conditions or inclement weather, lane lines may become blurred or disappear completely, leading to a decrease in the accuracy of camera perception and increasing the risk of falsely detecting serpentine driving behavior. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, and storage medium for detecting vehicle serpentine driving, in order to solve the problems of difficulty, low accuracy, and poor robustness in detecting vehicle serpentine driving in autonomous driving.

[0005] This invention provides a method for detecting serpentine driving of vehicles, comprising:

[0006] Obtain the vehicle speed sequence and steering wheel angle sequence;

[0007] Based on a sliding window, key feature points are extracted from the steering wheel angle sequence to obtain a set of corner feature points. The size of the sliding window is dynamically adjusted based on the vehicle speed sequence.

[0008] Based on the vehicle speed sequence and the set of corner feature points, the system detects whether the vehicle exhibits serpentine driving behavior.

[0009] According to the present invention, a method for detecting vehicle serpentine driving is provided, wherein the key feature points are extracted from the steering wheel angle sequence based on a sliding window to obtain a set of corner feature points, including:

[0010] Based on a sliding window, key feature points are extracted from the steering wheel angle sequence to obtain an initial feature point set.

[0011] Data augmentation is performed on some key feature points in the initial feature point set to obtain an augmented feature point set;

[0012] The enhanced feature point set is integrated into the initial feature point set to obtain the corner feature point set.

[0013] According to the present invention, a method for detecting vehicle serpentine driving is provided, wherein the initial feature point set is obtained by extracting key feature points from the steering wheel angle sequence based on a sliding window, including:

[0014] The steering wheel angle sequence is traversed. During the traversal, based on the index and size of the current sliding window, the corresponding angle data is read from the steering wheel angle sequence to obtain the steering wheel angle set within the current sliding window.

[0015] The maximum and minimum steering angles are determined from the set of steering wheel angles, and the maximum and minimum steering angles are used as candidate feature points.

[0016] Based on a preset threshold, the candidate feature points are filtered to obtain the key feature points within the current sliding window;

[0017] The initial set of feature points is obtained based on the key feature points within each sliding window.

[0018] According to the present invention, a method for detecting serpentine driving behavior of a vehicle, wherein the detection of whether a vehicle exhibits serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points includes:

[0019] The vehicle speed sequence and the set of corner feature points are input into the detection model to obtain the detection result output by the detection model. The set of corner feature points includes multiple feature points and the time and corner data corresponding to the multiple feature points.

[0020] The detection model includes a feature extractor and a classifier. The feature extractor is used to extract features based on the vehicle speed sequence and the time and turning angle data corresponding to the multiple feature points to obtain deep features. The classifier is used to perform classification detection based on the deep features and output the detection results.

[0021] According to a vehicle serpentine driving detection method provided by the present invention, the training steps of the detection model include:

[0022] Acquire historical vehicle speed data and historical steering wheel angle data;

[0023] Preprocessing the vehicle speed history data and the steering wheel angle history data yields sample speed sequences and sample angle sequences.

[0024] Based on the sample sliding window, key feature points are extracted from the sample corner sequence to obtain the initial set of sample feature points. The size of the sample sliding window is dynamically adjusted based on the sample velocity sequence.

[0025] Data augmentation is performed on some key feature points in the initial feature point set of the sample to obtain a sample augmented feature point set, and the sample augmented feature point set is integrated into the initial feature point set of the sample to obtain a sample corner feature point set;

[0026] Based on the vehicle speed sequence, the set of sample corner feature points, and the corresponding sample detection results, the initial model is trained to obtain the detection model.

[0027] According to a vehicle serpentine driving detection method provided by the present invention, the step of acquiring the vehicle speed sequence and steering wheel angle sequence includes:

[0028] Real-time collection of vehicle speed data and steering wheel angle data;

[0029] The collected vehicle speed data and steering wheel angle data are preprocessed to obtain the vehicle speed sequence and the steering wheel angle sequence.

[0030] According to a vehicle serpentine driving detection method provided by the present invention, the preprocessing step of the steering wheel angle data includes:

[0031] The steering wheel angle data is filtered and noise is eliminated by applying Fourier transform technology to obtain the steering wheel angle sequence.

[0032] The present invention also provides a vehicle serpentine driving detection device, comprising:

[0033] The acquisition unit is used to acquire vehicle speed sequences and steering wheel angle sequences;

[0034] An extraction unit is used to extract key feature points from the steering wheel angle sequence based on a sliding window to obtain a set of corner feature points. The size of the sliding window is dynamically adjusted based on the vehicle speed sequence.

[0035] The detection unit is used to detect whether the vehicle is exhibiting serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points.

[0036] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the vehicle serpentine driving detection method as described above.

[0037] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle serpentine driving detection method as described above.

[0038] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle serpentine driving detection method as described above.

[0039] The vehicle serpentine driving detection method, apparatus, electronic device, and storage medium provided by this invention extract key feature points from the steering wheel angle sequence by employing a sliding window whose size is dynamically adjusted based on the vehicle speed sequence. This dynamic adjustment mechanism enables the invention to more accurately identify key feature points according to driving characteristics at different speeds, thereby improving the accuracy and adaptability of detection. By combining the vehicle speed sequence and the set of corner feature points, the invention can efficiently detect whether a vehicle is exhibiting serpentine driving behavior without relying on image data for detection. This avoids interference factors such as changes in lighting, occlusion, and poor image quality that can occur with image detection, thus ensuring high robustness and stability under various driving environments. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in this invention or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating the vehicle serpentine driving detection method provided by the present invention;

[0042] Figure 2 This is the original waveform diagram of the vehicle steering wheel angle data provided by the present invention;

[0043] Figure 3 This is a waveform diagram of the filtered vehicle steering wheel angle data provided by the present invention;

[0044] Figure 4 This is a schematic diagram of the steering wheel angle feature points provided by the present invention;

[0045] Figure 5 This is a schematic diagram of the training process of the detection model provided by the present invention;

[0046] Figure 6 This is a flowchart illustrating the vehicle serpentine driving detection method based on steering wheel angle provided by the present invention.

[0047] Figure 7This is a schematic diagram of the vehicle serpentine driving detection device provided by the present invention;

[0048] Figure 8 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0050] Currently, the serpentine driving behavior of autonomous vehicles is primarily identified using cameras to detect lane lines and calculate yaw angles. However, existing serpentine driving detection technologies have significant limitations. On one hand, this technology heavily relies on the camera's clear perception of lane lines. However, in adverse road conditions or bad weather, lane lines may become blurred or even disappear completely, leading to a significant decrease in the camera's accuracy. In such cases, the system may misjudge the vehicle as serpentine, triggering unnecessary safety warnings or interventions, thus affecting driving comfort and safety. On the other hand, even when lane lines are clearly visible, limitations in the camera's sensing range and interference factors such as ambient light and shadows can still affect the accuracy and stability of serpentine driving detection.

[0051] To address this issue, this invention provides a vehicle serpentine driving detection method based on steering wheel angle. By utilizing vehicle steering wheel angle information and extracting peak and trough feature points from the steering wheel angle waveform, combined with vehicle speed information, the method can accurately determine whether the vehicle is exhibiting serpentine driving behavior. This method not only solves the problem of system self-detection of serpentine driving in autonomous driving systems, but also relies solely on low-cost hardware such as vehicle steering wheel angle and speed sensors, avoiding the high cost and insufficient robustness issues associated with image-based serpentine driving detection methods.

[0052] Figure 1 This is a flowchart illustrating the vehicle serpentine driving detection method provided by the present invention, as shown below. Figure 1 As shown, the method includes:

[0053] Step 110: Obtain the vehicle speed sequence and steering wheel angle sequence.

[0054] Specifically, a vehicle speed sequence refers to a series of speed values ​​that change over time during vehicle operation. These speed values ​​can be acquired in real time by the vehicle's speed sensors. The speed sensors measure the instantaneous speed of the vehicle at different points in time and record this data in sequence. A steering wheel angle sequence, on the other hand, refers to a series of steering wheel angle values ​​that change over time when the driver or autonomous driving system controls the steering wheel. These values ​​can be acquired by a steering wheel angle sensor. The steering wheel angle sensor monitors the steering wheel's rotation angle in real time and records this data in sequence.

[0055] Specifically, vehicles are typically equipped with sensors, including speed sensors and steering wheel angle sensors. During vehicle operation, these sensors collect real-time data on vehicle speed and steering wheel angle. This data is usually output as digital signals and can be transmitted to the data processing unit via the vehicle's data bus (such as the CAN bus). The data processing unit receives the data from the sensors and records it as a sequence. This sequence data is used for subsequent serpentine driving behavior detection. After recording the data, preprocessing can be performed, including data cleaning, filtering, and noise reduction, to improve the accuracy and reliability of the data.

[0056] It is understood that the embodiments of the present invention rely only on the vehicle's steering wheel angle sensor and speed sensor, both of which are typically standard equipment in vehicles, thus requiring no additional high-cost hardware investment. Compared to methods relying on image detection, the present invention can significantly reduce costs while avoiding the complexity and insufficient robustness that image detection may bring.

[0057] Step 120: Based on the sliding window, extract key feature points from the steering wheel angle sequence to obtain a set of corner feature points. The size of the sliding window is dynamically adjusted based on the vehicle speed sequence.

[0058] Specifically, for a steering wheel angle sequence obtained over a period of time, a sliding window approach can be used to extract key feature points from the sequence, thereby reducing the amount of data. Here, the sliding window is a commonly used data processing technique that slides a fixed-size or dynamically sized window across a data sequence to analyze the data within the window. In vehicle serpentine behavior detection, the sliding window is used to analyze the steering wheel angle sequence to identify potential serpentine behaviors.

[0059] In this embodiment of the invention, the size of the sliding window is not fixed, but dynamically adjusted based on the vehicle speed sequence. This is because the driving characteristics of a vehicle are different at different speeds. For example, when driving at high speed, the vehicle's handling becomes more sensitive, and a small turn of the steering wheel may cause significant changes in the vehicle's driving direction and trajectory. When driving at low speed, the vehicle's handling is relatively less sensitive, and a small turn of the steering wheel will not have a significant impact on the vehicle's driving direction and trajectory.

[0060] To dynamically adjust the size of the sliding window, the vehicle speed sequence can be iterated through, and the window size can be adjusted based on each speed value encountered. For example, a base window size can be set, and the window size can be adjusted linearly or non-linearly based on speed changes. When vehicle speed is low, the sliding window size increases accordingly to smooth out minor fluctuations in steering wheel angle data; when vehicle speed is high, the sliding window size decreases accordingly to more sensitively capture changes in steering wheel angle. It should be understood that a dynamic sliding window can more accurately identify key feature points compared to a fixed-size window.

[0061] Once the size of the sliding window is determined, it can be used to extract key feature points from the steering wheel angle sequence. Specifically, the sliding window is moved across the steering wheel angle sequence, and statistical characteristics (such as maximum and minimum values) of the data within the window are calculated at each slide. Then, key feature points are determined based on these statistical characteristics.

[0062] Understandably, in vehicle serpentine behavior detection, key feature points refer to points with specific significance identified in the steering wheel angle sequence, such as peaks and troughs. The set of corner feature points is a collection of these key feature points, encompassing all peaks and troughs identified in the steering wheel angle sequence. Here, a peak represents a point where the steering wheel angle increases sharply and then decreases within a short period, while a trough represents a point where the steering wheel angle decreases sharply and then increases within a short period. These peaks and troughs may represent unstable states during vehicle movement and are important features for detecting serpentine behavior.

[0063] Step 130: Based on the vehicle speed sequence and the set of corner feature points, detect whether the vehicle exhibits serpentine driving behavior.

[0064] Specifically, after extracting the set of corner feature points, the presence of serpentine driving behavior can be detected by analyzing the vehicle speed sequence and the set of corner feature points. First, analyzing the vehicle speed sequence reveals the vehicle's driving state. If the vehicle speed remains relatively stable, it indicates normal driving. However, frequent fluctuations in speed, especially when these fluctuations correlate with changes in steering wheel angle, may indicate serpentine driving behavior. Next, the set of corner feature points—the peaks and troughs in the steering wheel angle sequence—can be analyzed. These feature points represent sharp steering wheel rotations, potentially reflecting frequent adjustments to the vehicle's direction by the autonomous driving system. For example, the distance between adjacent feature points (peaks or troughs) can be calculated. Short distances between adjacent feature points indicate frequent steering wheel rotations within a short period, potentially suggesting serpentine driving behavior. For example, the angle changes of feature points can also be analyzed, that is, the angle difference between adjacent feature points can be calculated to understand the direction and degree of steering wheel rotation. If the angle difference between adjacent feature points is large, it indicates that the steering wheel has rotated significantly in a short period of time, which may also indicate that the vehicle is exhibiting serpentine driving behavior.

[0065] Finally, the analysis results of the vehicle speed sequence and the set of turning feature points can be combined. If the vehicle speed fluctuates frequently, and this fluctuation is correlated with the sharp turning of the steering wheel (i.e., the speed fluctuation is accompanied by frequent steering wheel turning), then it can be determined that the vehicle is exhibiting serpentine driving behavior.

[0066] It should be noted that the detection method described above is only one possible implementation. In another embodiment, the vehicle speed sequence and the set of corner feature points can be directly input into a neural network model. The model analyzes and detects the vehicle, outputting corresponding detection results. Based on the detection results, it can be determined whether the vehicle exhibits serpentine driving behavior. Here, the neural network model can be pre-trained based on a large amount of sample data.

[0067] The method provided in this invention extracts key feature points from a steering wheel angle sequence by employing a sliding window whose size is dynamically adjusted based on the vehicle speed sequence. This dynamic adjustment mechanism enables the invention to more accurately identify key feature points according to driving characteristics at different speeds, thereby improving the accuracy and adaptability of detection. By combining the vehicle speed sequence and the set of corner feature points, the method can efficiently detect whether a vehicle exhibits serpentine driving behavior without relying on image data. This avoids interference factors such as changes in lighting, occlusion, and poor image quality that can occur with image detection, thus ensuring high robustness and stability under various driving environments.

[0068] Based on the above embodiments, step 110 specifically includes:

[0069] Step 111: Real-time collection of vehicle speed data and steering wheel angle data;

[0070] Step 112: Preprocess the collected vehicle speed data and steering wheel angle data to obtain the vehicle speed sequence and the steering wheel angle sequence.

[0071] Specifically, during vehicle operation, speed and steering wheel angle data can be collected in real time using speed sensors and steering wheel angle sensors installed on the vehicle. Here, vehicle speed data refers to the speed value measured and output by the speed sensor during the current driving process. Steering wheel angle data refers to the rotation angle value measured by the steering wheel angle sensor when the driver or autonomous driving system operates the steering wheel.

[0072] The collected vehicle speed and steering wheel angle data can be preprocessed to ensure data quality and improve the accuracy of subsequent analysis. Preprocessing can include steps such as data cleaning (e.g., removing outliers, filling in missing values), data smoothing, and data standardization.

[0073] Furthermore, the preprocessing steps for the steering wheel angle data include:

[0074] The steering wheel angle data is filtered and noise is eliminated by applying Fourier transform technology to obtain the steering wheel angle sequence.

[0075] Specifically, for the acquired digital steering wheel angle data, Fourier transform technology can be used to filter and eliminate noise, thereby improving the signal-to-noise ratio. Here, Fourier transform is a mathematical method that transforms a signal from the time domain to the frequency domain to identify different frequency components in the signal, thus achieving signal filtering and noise elimination.

[0076] Filtering and noise reduction of steering wheel angle data using Fourier transform techniques can be achieved through the following steps: First, perform a Fourier transform on the steering wheel angle data, converting it from the time domain to the frequency domain to obtain a frequency-domain signal containing the amplitude and phase information of different frequency components. Then, in the frequency domain, certain frequency components can be selectively retained or removed based on the signal's frequency characteristics. For example, for steering wheel angle data, high-frequency noise components can be removed, as these are typically caused by sensor noise or other external interference. Finally, perform an inverse Fourier transform on the filtered frequency-domain signal to convert it back to the time domain, yielding the filtered and noise-reduced steering wheel angle data.

[0077] Figure 2 This is the original waveform diagram of the vehicle steering wheel angle data provided by the present invention, such as... Figure 2 As shown, the steering wheel angle sensor installed on the vehicle can collect and continuously store steering wheel angle data in real time. This data is output in the form of digital signals, and the resulting raw waveform is shown in the figure. Figure 2 As shown. Figure 3 This is a waveform diagram of the filtered vehicle steering wheel angle data provided by the present invention, such as... Figure 3 As shown, by using a filter and Fourier transform technology to filter and eliminate noise in the steering wheel angle data, the following can be obtained: Figure 3 The waveform of the filtered signal is shown below. It should be noted that... Figure 2 and Figure 3 In the diagram, the horizontal axis represents time, and the vertical axis represents the steering wheel angle.

[0078] Based on any of the above embodiments, step 120 specifically includes:

[0079] Step 121: Based on the sliding window, extract key feature points from the steering wheel angle sequence to obtain an initial feature point set.

[0080] Specifically, a dynamically resizing sliding window is used to slide across the steering wheel angle sequence. Each slide calculates the statistical characteristics of the data within the window (such as maximum and minimum values), and these characteristics determine key feature points. Based on the key feature points determined within each window, an initial feature point set can be formed. Here, the initial feature point set refers to the set of preliminary key feature points extracted from the steering wheel angle sequence based on the sliding window during vehicle serpentine driving behavior detection. These feature points mainly include peaks and troughs, which represent significant locations of steering wheel angle changes.

[0081] Furthermore, step 121 specifically includes:

[0082] Step 1211: Traverse the steering wheel angle sequence. During the traversal, based on the index and size of the current sliding window, read the corresponding angle data from the steering wheel angle sequence to obtain the steering wheel angle set within the current sliding window.

[0083] Step 1212: Determine the maximum and minimum steering angles from the set of steering wheel angles, and use the maximum and minimum steering angles as candidate feature points;

[0084] Step 1213: Based on a preset threshold, the candidate feature points are filtered to obtain the key feature points within the current sliding window;

[0085] Step 1214: Based on the key feature points within each sliding window, obtain the initial set of feature points.

[0086] Specifically, firstly, a basic window size can be set, and the window size can be dynamically adjusted according to the corresponding speed values ​​in the vehicle speed sequence. Simultaneously, a sliding step size is set, which determines the frequency at which the window moves along the steering wheel angle sequence. For example, the step size can be set to 1, meaning the window moves one data point at a time, or it can be adjusted according to the analysis requirements; this embodiment of the invention does not impose specific limitations on this. An index variable also needs to be set to track the current position of the sliding window in the steering wheel angle sequence. Next, a loop structure is used to traverse the entire steering wheel angle sequence. In each loop, the index is incremented by the sliding step size. In each loop, based on the current index and the window size, the corresponding data points (i.e., angle data) are read from the steering wheel angle sequence, forming the steering wheel angle set within the current sliding window. Subsequently, statistical analysis is performed on all angle data in the steering wheel angle set within the current sliding window to determine the maximum angle (i.e., the maximum value) and the minimum angle (i.e., the minimum value).

[0087] It should be noted that, considering the influence of various factors during actual signal acquisition (such as sensor accuracy, environmental factors, electromagnetic interference, etc.), the acquired steering wheel angle data often contains a certain amount of noise. This noise may cause some local maxima or minima in the data, which do not represent the true peaks or troughs. Therefore, directly using the maximum and minimum values ​​within each window as peaks and troughs (i.e., key feature points) may introduce a large number of misjudgments. To accurately identify peaks and troughs, in each loop, the maximum and minimum values ​​within the current window can be used as candidate feature points. By further filtering these candidate feature points, effective key feature points can be obtained.

[0088] Since true peaks and troughs are typically significant feature points of signal change over a period of time, they possess a certain smoothness and continuity. In contrast, local maxima or minima caused by noise often exhibit abrupt jumps and lack this smoothness. Therefore, to avoid misidentifying noise-induced local maxima or minima as peaks or troughs, a threshold can be set to filter candidate feature points. For example, a threshold for corner changes can be set; when the difference between the maximum or minimum value and the corresponding value in an adjacent window exceeds this threshold, it is considered a key feature point. Figure 4 This is a schematic diagram of the steering wheel angle feature points provided by the present invention, such as... Figure 4 As shown in the figure, the feature points marked in red are the peaks, and the feature points marked in green are the troughs.

[0089] Understandably, the key feature points within the current sliding window refer to the significant locations of steering wheel angle changes determined after filtering within the current sliding window range; that is, the actual peaks and troughs. After traversing the steering wheel angle sequence, an initial set of feature points can be obtained based on the key feature points determined within each sliding window.

[0090] Step 122: Perform data augmentation on some key feature points in the initial feature point set to obtain an augmented feature point set.

[0091] Specifically, after obtaining the initial set of feature points, a subset of key feature points can be randomly selected from this set, and data augmentation can be performed on these key feature points to address the potential shortage of data for serpentine driving detection, thereby increasing the robustness and generalization ability of the model. Here, data augmentation is a technique that increases data diversity by adding noise, transforming data, etc. In vehicle serpentine driving behavior detection, data augmentation can be used to improve the model's generalization ability, enabling it to better cope with various complex driving situations.

[0092] Data augmentation of key feature points can be achieved by adding Gaussian noise to the steering wheel angle data of these feature points. Specifically, first, the mean and standard deviation of the Gaussian noise are set. Then, a certain proportion (e.g., 40%) of the feature points are randomly selected from the initial feature point set for augmentation to maintain the authenticity of the data distribution. For each selected feature point, random noise conforming to a Gaussian distribution can be generated and added to the steering wheel angle value of the selected feature point. After augmenting all selected feature points, an augmented feature point set is obtained. Here, the augmented feature point set is the set of feature points obtained after data augmentation. These feature points contain information from the original feature points and also incorporate noise and other variations, thereby increasing the diversity of the data.

[0093] Step 123: Integrate the enhanced feature point set into the initial feature point set to obtain the corner feature point set.

[0094] Specifically, after obtaining the enhanced feature point set, each enhanced feature point in the enhanced feature point set can be merged with each original feature point in the initial feature point set to form a new feature point set, namely the corner feature point set. This new set contains information from both the original feature points and the feature points after data augmentation, thereby improving the diversity of feature points and the generalization ability of the model.

[0095] Based on any of the above embodiments, step 130 specifically includes:

[0096] Step 131: Input the vehicle speed sequence and the set of corner feature points into the detection model to obtain the detection result output by the detection model. The set of corner feature points includes multiple feature points and the time and corner data corresponding to the multiple feature points.

[0097] The detection model includes a feature extractor and a classifier. The feature extractor is used to extract features based on the vehicle speed sequence and the time and turning angle data corresponding to the multiple feature points to obtain deep features. The classifier is used to perform classification detection based on the deep features and output the detection results.

[0098] Specifically, when detecting vehicle serpentine driving behavior based on vehicle speed sequences and corner feature point sets, the vehicle speed sequences and corner feature point sets can be directly input into the detection model to obtain the detection results output by the model. Here, the corner feature point set includes multiple feature points and the time and angle data (i.e., steering wheel angle data) corresponding to each feature point. Here, multiple feature points refer to the extracted key feature points (including peaks and troughs) and some key feature points after data augmentation.

[0099] Understandably, the detection model is a trained machine learning or deep learning model used to detect whether a vehicle exhibits serpentine driving behavior. It receives a vehicle speed sequence and a set of corner feature points as input and outputs the detection result. The detection model mainly consists of a feature extractor and a classifier. The feature extractor extracts useful feature information from the input vehicle speed sequence and corner feature point set, i.e., obtains deep features that better reflect the vehicle's driving state. The classifier receives the deep features output by the feature extractor and performs classification detection based on these features, aiming to determine whether the input data belongs to the serpentine driving behavior category. It should be understood that deep features refer to high-level, abstract features extracted from the original input data by the feature extractor.

[0100] Specifically, feature extractors can use deep learning techniques, such as Transformer, LSTM (Long Short-Term Memory), and GRU (Gated Recurrent Unit), to automatically extract useful features from the input data. Classifiers can use MLP (Multilayer Perceptron) and Softmax to receive deep features and perform classification to output detection results. It should be understood that in vehicle serpentine behavior detection, the detection result can be a binary classification result, i.e., whether the vehicle exhibits serpentine behavior. If the detection result is "yes," it indicates that the vehicle exhibits serpentine behavior; if the detection result is "no," it indicates that the vehicle does not exhibit serpentine behavior.

[0101] Based on any of the above embodiments Figure 5 This is a schematic diagram of the training process of the detection model provided by the present invention, as shown below. Figure 5 As shown, the training steps of the detection model include:

[0102] Step 510: Obtain historical vehicle speed data and historical steering wheel angle data.

[0103] Specifically, vehicle speed history data refers to recorded data on a vehicle's speed over a past period. Steering wheel angle history data refers to data on steering wheel angle changes caused by the driver or autonomous driving system's steering wheel operations over a past period. Vehicle speed history data and steering wheel angle history data can be obtained from vehicle sensors or recording systems and used as sample data for model training. This data is typically stored in time-series format, containing information on the vehicle's speed and steering wheel angle at different points in time.

[0104] Step 520: Preprocess the vehicle speed history data and the steering wheel angle history data to obtain sample speed sequences and sample steering angle sequences.

[0105] Specifically, the acquired historical vehicle speed data and historical steering wheel angle data are preprocessed, including cleaning, denoising, and normalization, to obtain cleaner and easier-to-analyze data. The preprocessed data is then organized into sample speed sequences and sample steering wheel angle sequences for subsequent feature extraction and model training.

[0106] Step 530: Based on the sample sliding window, extract key feature points from the sample corner sequence to obtain the initial feature point set of the sample. The size of the sample sliding window is dynamically adjusted based on the sample velocity sequence.

[0107] Specifically, key feature points can be extracted from the sample corner sequence using a sample sliding window. The size of the sample sliding window is dynamically adjusted based on the sample velocity sequence to better capture changes in the steering wheel angle. The extracted key feature points are organized into an initial set of sample feature points. Here, the sample sliding window is a window that slides across the sample corner sequence to extract key feature points (including peaks and troughs) from the corner data.

[0108] Step 540: Perform data augmentation on some key feature points in the initial feature point set of the sample to obtain a sample augmented feature point set, and integrate the sample augmented feature point set into the initial feature point set of the sample to obtain a sample corner feature point set.

[0109] Specifically, a subset of key feature points can be randomly selected from the initial set of feature points for the samples, and data augmentation can be performed on these key feature points, such as by adding Gaussian noise, to increase the diversity of the data and the generalization ability of the model. The augmented feature points are then integrated into the initial set of feature points for the samples, forming a set of corner feature points for the samples.

[0110] Step 550: Based on the vehicle speed sequence, the sample corner feature point set, and the corresponding sample detection results, train the initial model to obtain the detection model.

[0111] Specifically, the initial model is trained using preprocessed vehicle speed sequences, sample corner feature point sets, and corresponding sample detection results (i.e., whether the vehicle exhibits serpentine driving behavior) as training data. During training, the model continuously adjusts its parameters to minimize the difference between the predicted results and the actual detection results. Ultimately, the trained model becomes the detection model, which can be used to detect the vehicle's driving status and the driver's driving behavior in real time.

[0112] Based on any of the above embodiments Figure 6 This is a flowchart illustrating the vehicle serpentine driving detection method based on steering wheel angle provided by the present invention, as shown below. Figure 6 As shown in the diagram, the dashed lines represent the model training phase, and the solid lines represent the model inference phase. The implementation flow of this method is as follows:

[0113] Step S1: Collect and store vehicle speed sequence and steering wheel angle sequence.

[0114] Step S2: Using Fourier transform technology, the steering wheel angle sequence is filtered and noise is eliminated to improve the signal-to-noise ratio.

[0115] Step S3: Based on the vehicle speed sequence, design a dynamic sliding window to identify key feature points in the steering wheel angle sequence, including peaks and troughs, in order to reduce the amount of data and extract deeper features later.

[0116] Step S4: Using the data augmentation module, Gaussian noise is added to some of the key feature points extracted in step S3 to address the issue of serpentine driving and potentially insufficient data volume, thereby increasing the robustness and generalization of the model.

[0117] Step S5: Establish a deep neural network model, including a feature extractor and a classifier. The input to the feature extractor includes information such as the speed values ​​in the vehicle speed sequence, the time corresponding to the peak and trough positions (i.e., key feature points) in the steering wheel angle sequence, and the steering wheel angle, and outputs deep features; the deep features are input to the classifier, and finally output whether the vehicle exhibits serpentine driving behavior.

[0118] Step S6 involves dividing the preprocessed dataset into a training set and a test set. The model is trained using the training set, and its performance is evaluated using the test set. During training, strategies such as cross-validation and early stopping can be employed to prevent overfitting.

[0119] Step S7: Use the test set to evaluate the trained model. Based on the evaluation results, the model's structure, parameters, etc., can be adjusted and optimized to improve the model's performance.

[0120] Step S8 involves deploying the trained model into a real-world application scenario to achieve real-time detection of vehicle serpentine driving behavior. For example, the model can be integrated into the vehicle's intelligent driving system, collecting steering wheel angle and speed data in real time and inputting it into the model for detection. Based on the model's output, corresponding warning or control strategies can be triggered to improve driving safety and comfort.

[0121] The method provided in this invention utilizes steering wheel angle information to extract the peaks and troughs of the steering wheel angle waveform, and combines this with vehicle speed information to input into a deep neural network to determine whether serpentine driving occurs. This method not only solves the problem of the system self-detecting whether the vehicle is serpentine in autonomous driving, but also relies solely on low-cost hardware such as the vehicle's built-in steering wheel angle sensor and speed sensor, avoiding the high cost and insufficient robustness of image-based serpentine driving detection methods.

[0122] Based on any of the above embodiments Figure 7 This is a schematic diagram of the vehicle serpentine driving detection device provided by the present invention, as shown below. Figure 7 As shown, the device includes:

[0123] Acquisition unit 710 is used to acquire vehicle speed sequence and steering wheel angle sequence;

[0124] Extraction unit 720 is used to extract key feature points from the steering wheel angle sequence based on a sliding window to obtain a set of corner feature points. The size of the sliding window is dynamically adjusted based on the vehicle speed sequence.

[0125] The detection unit 730 is used to detect whether the vehicle is exhibiting serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points.

[0126] The apparatus provided in this invention extracts key feature points from a steering wheel angle sequence by employing a sliding window whose size is dynamically adjusted based on the vehicle speed sequence. This dynamic adjustment mechanism enables the invention to more accurately identify key feature points according to driving characteristics at different speeds, thereby improving the accuracy and adaptability of detection. By combining the vehicle speed sequence and the set of corner feature points, it can efficiently detect whether a vehicle exhibits serpentine driving behavior without relying on image data for detection. This avoids interference factors such as changes in lighting, occlusion, and poor image quality caused by image detection, thus ensuring that the method maintains high robustness and stability under various driving environments.

[0127] Based on any of the above embodiments, the extraction unit 720 includes:

[0128] The feature point extraction subunit is used to extract key feature points from the steering wheel angle sequence based on a sliding window to obtain an initial feature point set.

[0129] The data augmentation subunit is used to perform data augmentation on some key feature points in the initial feature point set to obtain an augmented feature point set.

[0130] The feature point merging subunit is used to integrate the enhanced feature point set into the initial feature point set to obtain the corner feature point set.

[0131] Based on any of the above embodiments, the feature point extraction subunit is specifically used for:

[0132] The steering wheel angle sequence is traversed. During the traversal, based on the index and size of the current sliding window, the corresponding angle data is read from the steering wheel angle sequence to obtain the steering wheel angle set within the current sliding window.

[0133] The maximum and minimum steering angles are determined from the set of steering wheel angles, and the maximum and minimum steering angles are used as candidate feature points.

[0134] Based on a preset threshold, the candidate feature points are filtered to obtain the key feature points within the current sliding window;

[0135] The initial set of feature points is obtained based on the key feature points within each sliding window.

[0136] Based on any of the above embodiments, the detection unit 730 is specifically used for:

[0137] The vehicle speed sequence and the set of corner feature points are input into the detection model to obtain the detection result output by the detection model. The set of corner feature points includes multiple feature points and the time and corner data corresponding to the multiple feature points.

[0138] The detection model includes a feature extractor and a classifier. The feature extractor is used to extract features based on the vehicle speed sequence and the time and turning angle data corresponding to the multiple feature points to obtain deep features. The classifier is used to perform classification detection based on the deep features and output the detection results.

[0139] Based on any of the above embodiments, the device further includes a model training unit, which is used for:

[0140] Acquire historical vehicle speed data and historical steering wheel angle data;

[0141] Preprocessing the vehicle speed history data and the steering wheel angle history data yields sample speed sequences and sample angle sequences.

[0142] Based on the sample sliding window, key feature points are extracted from the sample corner sequence to obtain the initial set of sample feature points. The size of the sample sliding window is dynamically adjusted based on the sample velocity sequence.

[0143] Data augmentation is performed on some key feature points in the initial feature point set of the sample to obtain a sample augmented feature point set, and the sample augmented feature point set is integrated into the initial feature point set of the sample to obtain a sample corner feature point set;

[0144] Based on the vehicle speed sequence, the set of sample corner feature points, and the corresponding sample detection results, the initial model is trained to obtain the detection model.

[0145] Based on any of the above embodiments, the acquisition unit 710 includes:

[0146] The data acquisition subunit is used to collect vehicle speed data and steering wheel angle data in real time;

[0147] The preprocessing subunit is used to preprocess the collected vehicle speed data and steering wheel angle data to obtain the vehicle speed sequence and the steering wheel angle sequence.

[0148] Based on any of the above embodiments, the preprocessing subunit is specifically used for:

[0149] The steering wheel angle data is filtered and noise is eliminated by applying Fourier transform technology to obtain the steering wheel angle sequence.

[0150] Figure 8 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8 As shown, the electronic device may include a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute a vehicle serpentine driving detection method. This method includes: acquiring a vehicle speed sequence and a steering wheel angle sequence; extracting key feature points from the steering wheel angle sequence based on a sliding window to obtain a set of corner feature points, wherein the size of the sliding window is dynamically adjusted based on the vehicle speed sequence; and detecting whether the vehicle exhibits serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points.

[0151] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to related technologies, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the vehicle serpentine driving detection method provided by the above methods. The method includes: acquiring a vehicle speed sequence and a steering wheel angle sequence; extracting key feature points from the steering wheel angle sequence based on a sliding window to obtain a set of corner feature points, wherein the size of the sliding window is dynamically adjusted based on the vehicle speed sequence; and detecting whether the vehicle exhibits serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points.

[0153] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a vehicle serpentine driving detection method provided by the above methods. The method includes: acquiring a vehicle speed sequence and a steering wheel angle sequence; extracting key feature points from the steering wheel angle sequence based on a sliding window to obtain a set of corner feature points, wherein the size of the sliding window is dynamically adjusted based on the vehicle speed sequence; and detecting whether the vehicle exhibits serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points.

[0154] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0155] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of software products. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting serpentine driving of a vehicle, characterized in that, include: Obtain the vehicle speed sequence and steering wheel angle sequence; Based on a sliding window, key feature points are extracted from the steering wheel angle sequence to obtain a set of corner feature points. The size of the sliding window is dynamically adjusted based on the vehicle speed sequence. Based on the vehicle speed sequence and the set of corner feature points, detect whether the vehicle exhibits serpentine driving behavior; The process of extracting key feature points from the steering wheel angle sequence using a sliding window yields a set of corner feature points, including: Based on a sliding window, key feature points are extracted from the steering wheel angle sequence to obtain an initial feature point set. Data augmentation is performed on some key feature points in the initial feature point set to obtain an augmented feature point set; The enhanced feature point set is integrated into the initial feature point set to obtain the corner feature point set.

2. The vehicle serpentine driving detection method according to claim 1, characterized in that, The process of extracting key feature points from the steering wheel angle sequence using a sliding window yields an initial set of feature points, including: The steering wheel angle sequence is traversed. During the traversal, based on the index and size of the current sliding window, the corresponding angle data is read from the steering wheel angle sequence to obtain the steering wheel angle set within the current sliding window. The maximum and minimum steering angles are determined from the set of steering wheel angles, and the maximum and minimum steering angles are used as candidate feature points. Based on a preset threshold, the candidate feature points are filtered to obtain the key feature points within the current sliding window; The initial set of feature points is obtained based on the key feature points within each sliding window.

3. The vehicle serpentine driving detection method according to claim 1, characterized in that, The step of detecting whether a vehicle exhibits serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points includes: The vehicle speed sequence and the set of corner feature points are input into the detection model to obtain the detection result output by the detection model. The set of corner feature points includes multiple feature points and the time and corner data corresponding to the multiple feature points. The detection model includes a feature extractor and a classifier. The feature extractor is used to extract features based on the vehicle speed sequence and the time and turning angle data corresponding to the multiple feature points to obtain deep features. The classifier is used to perform classification detection based on the deep features and output the detection results.

4. The vehicle serpentine driving detection method according to claim 3, characterized in that, The training steps of the detection model include: Acquire historical vehicle speed data and historical steering wheel angle data; Preprocessing the vehicle speed history data and the steering wheel angle history data yields sample speed sequences and sample angle sequences. Based on the sample sliding window, key feature points are extracted from the sample corner sequence to obtain the initial set of sample feature points. The size of the sample sliding window is dynamically adjusted based on the sample velocity sequence. Data augmentation is performed on some key feature points in the initial feature point set of the sample to obtain a sample augmented feature point set, and the sample augmented feature point set is integrated into the initial feature point set of the sample to obtain a sample corner feature point set; Based on the vehicle speed sequence, the set of sample corner feature points, and the corresponding sample detection results, the initial model is trained to obtain the detection model.

5. The vehicle serpentine driving detection method according to any one of claims 1 to 4, characterized in that, The acquisition of the vehicle speed sequence and steering wheel angle sequence includes: Real-time collection of vehicle speed data and steering wheel angle data; The collected vehicle speed data and steering wheel angle data are preprocessed to obtain the vehicle speed sequence and the steering wheel angle sequence.

6. The vehicle serpentine driving detection method according to claim 5, characterized in that, The preprocessing steps for the steering wheel angle data include: The steering wheel angle data is filtered and noise is eliminated by applying Fourier transform technology to obtain the steering wheel angle sequence.

7. A vehicle serpentine driving detection device, characterized in that, include: The acquisition unit is used to acquire vehicle speed sequences and steering wheel angle sequences; An extraction unit is used to extract key feature points from the steering wheel angle sequence based on a sliding window to obtain a set of corner feature points. The size of the sliding window is dynamically adjusted based on the vehicle speed sequence. The detection unit is used to detect whether the vehicle is exhibiting serpentine driving behavior based on the vehicle speed sequence and the set of corner feature points; The extraction unit is specifically used for: Based on a sliding window, key feature points are extracted from the steering wheel angle sequence to obtain an initial feature point set. Data augmentation is performed on some key feature points in the initial feature point set to obtain an augmented feature point set; The enhanced feature point set is integrated into the initial feature point set to obtain the corner feature point set.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the vehicle serpentine driving detection method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the vehicle serpentine driving detection method as described in any one of claims 1 to 6.

Images