Vehicle rapid acceleration risk detection method, platform and storage medium

Abstract

The present invention provides a vehicle sudden acceleration risk detection method, platform and storage medium, which relates to the field of data processing technology. By performing vehicle sudden acceleration detection based on a real-time status data sequence, a frequent sudden acceleration warning is output, and the frequent sudden acceleration warning has a risk time interval identifier; the vehicle condition detection is performed based on the risk time interval by calling and obtaining associated behavior data, and the vehicle condition detection information is output to update the warning level of the frequent sudden acceleration warning. The technical problem in the prior art of low accuracy in detecting frequent sudden acceleration behavior of vehicles, resulting in poor timeliness and accuracy of driving behavior warnings, is solved. The technical effect of quickly identifying and warning of frequent sudden acceleration behavior during driving and providing the driver with timely comprehensive warning information on driving behavior and vehicle condition is achieved, thereby preventing potential driving hazards.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a vehicle sudden acceleration risk detection method, platform, and storage medium. Background Art

[0002] With the continuous increase in traffic volume, especially in the context of accelerated urbanization, vehicle safety has become a key social concern. Frequent sudden acceleration not only increases vehicle fuel consumption and causes excessive wear and tear on vehicle components, but can also cause traffic accidents in certain situations, posing a serious threat to road safety.

[0003] Especially in complex road conditions such as highways and city streets, rapid acceleration of vehicles is often accompanied by complex driving behaviors such as sudden braking and sharp turns, which makes the detection of dangerous driving behaviors more complicated and difficult.

[0004] In summary, the existing technology has a technical problem that the detection accuracy of frequent sudden acceleration behavior of vehicles is low, resulting in poor timeliness and accuracy of driving behavior warnings. Summary of the Invention

[0005] This application provides a vehicle sudden acceleration risk detection method, platform and storage medium, which are used to solve the technical problem in the existing technology that the detection accuracy of frequent sudden acceleration behavior of vehicles is low, resulting in poor timeliness and accuracy of driving behavior warnings.

[0006] In view of the above problems, the present application provides a vehicle rapid acceleration risk detection method, platform and storage medium.

[0007] The first aspect of the present application provides a vehicle sudden acceleration risk detection method, the method comprising: presetting a state acquisition window; using the state acquisition window as a constraint, controlling a motion state unit to intermittently acquire motion state data of a target vehicle to obtain a real-time state data sequence; performing vehicle sudden acceleration detection according to the real-time state data sequence, and outputting a frequent sudden acceleration warning, wherein the frequent sudden acceleration warning has a risk time interval identifier; performing a reverse call on the vehicle behavior data of the target vehicle according to the risk time interval to obtain associated behavior data; performing a vehicle condition detection according to the associated behavior data, and outputting vehicle condition detection information; and updating the warning level of the frequent sudden acceleration warning according to the vehicle condition detection information.

[0008] The second aspect of the present application provides a vehicle sudden acceleration risk detection platform, which includes: an acquisition function configuration unit for presetting a state acquisition window; an acquisition function operation unit for controlling the motion state unit to intermittently acquire motion state data of the target vehicle with the state acquisition window as a constraint to obtain a real-time state data sequence; a driving warning output unit for performing vehicle sudden acceleration detection according to the real-time state data sequence after multi-level data preprocessing of the real-time state data sequence, and outputting a frequent sudden acceleration warning, wherein the frequent sudden acceleration warning has a risk time interval identifier; a behavior data acquisition unit for reversely calling the vehicle behavior data of the target vehicle according to the risk time interval to obtain associated behavior data; a vehicle condition detection execution unit for performing vehicle condition detection according to the associated behavior data and outputting vehicle condition detection information; and a warning level update unit for updating the warning level of the frequent sudden acceleration warning according to the vehicle condition detection information.

[0009] The third aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements the steps of the vehicle rapid acceleration risk detection method described in any one of the first aspects above.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] The method provided by the embodiment of the present application is as follows: a preset state acquisition window is set; the state acquisition window is used as a constraint to control the motion state unit to intermittently acquire the motion state data of the target vehicle, thereby obtaining a real-time state data sequence; vehicle sudden acceleration detection is performed based on the real-time state data sequence, and a frequent sudden acceleration warning is output, wherein the frequent sudden acceleration warning has a risk time interval identifier; vehicle behavior data of the target vehicle is reversely called based on the risk time interval to obtain associated behavior data; vehicle condition detection is performed based on the associated behavior data, and vehicle condition detection information is output; and the warning level of the frequent sudden acceleration warning is updated based on the vehicle condition detection information. The method achieves the technical effect of quickly identifying and warning of frequent sudden acceleration behavior during driving and providing the driver with timely comprehensive warning information on driving behavior and vehicle condition, thereby preventing potential driving hazards. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flow chart of the vehicle rapid acceleration risk detection method provided in this application;

[0013] Figure 2 A schematic diagram of the process of outputting a frequent sudden acceleration warning in the vehicle sudden acceleration risk detection method provided in this application;

[0014] Figure 3This is a schematic diagram of the structure of the vehicle rapid acceleration risk detection platform provided in this application.

[0015] Explanation of the accompanying symbols: collection function configuration unit 11, collection function operation unit 12, driving warning output unit 13, behavior data collection unit 14, vehicle condition detection execution unit 15, warning level updating unit 16. DETAILED DESCRIPTION

[0016] This application provides a vehicle rapid acceleration risk detection method, platform, and storage medium to address the existing technical issues of low accuracy in detecting frequent rapid acceleration, resulting in poor timeliness and accuracy in driving behavior warnings. The method achieves the technical effect of rapidly identifying and warning of frequent rapid acceleration during driving, providing drivers with timely, comprehensive warning information on driving behavior and vehicle conditions, thereby preventing potential driving hazards.

[0017] Below, the technical solutions of the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. It should also be noted that, for the convenience of description, only the parts related to the present invention, rather than all, are shown in the accompanying drawings.

[0018] Example 1

[0019] like Figure 1 As shown, the present application provides a vehicle rapid acceleration risk detection method, the method comprising:

[0020] A100: Preset status acquisition window.

[0021] Specifically, in this embodiment, the preset state collection window refers to a defined time period or interval for data collection during vehicle rapid acceleration testing. Specifically, the state collection window controls the timing of data collection. It specifies the time window (i.e., how often data is collected) and the amount of data collected each time from vehicle motion sensors (such as accelerometers and gyroscopes). The state collection window is typically measured in milliseconds (ms), for example, setting the data sampling frequency to 50Hz.

[0022] A200: Using the state acquisition window as a constraint, control the motion state unit to intermittently acquire motion state data of the target vehicle to obtain a real-time state data sequence.

[0023] In one embodiment, the state acquisition window is used as a constraint to control the motion state unit to intermittently acquire the motion state data of the target vehicle to obtain a real-time state data sequence. Previously, step A200 of the method provided in this application includes:

[0024] A201: Deploy the motion state unit on the target vehicle, wherein the motion state unit includes a three-axis acceleration sensor and a three-axis gyroscope.

[0025] A202: Perform deviation calibration and temperature compensation on the motion state unit to complete the localization of the data acquisition function of the motion state unit.

[0026] A203: Preset a data sampling frequency, and set the status acquisition window based on the data sampling frequency.

[0027] A204: Using the state collection window as a constraint, control the motion state unit to intermittently collect the motion state data of the target vehicle.

[0028] In one embodiment, the state acquisition window is used as a constraint to control the motion state unit to intermittently acquire the motion state data of the target vehicle to obtain a real-time state data sequence. Step A200 of the method provided in this application further includes:

[0029] A210: Pre-build the first doubly linked list and the second doubly linked list.

[0030] A220: After the target vehicle is started, the motion state unit is controlled to intermittently collect the motion state data of the target vehicle with the state collection window as a constraint to obtain first real-time state data, wherein the first real-time state data includes first acceleration data and first angular velocity data.

[0031] A230: Temporarily store the first acceleration data in the first bidirectional linked list, wherein a first cache node in the first bidirectional linked list for storing the first acceleration data includes pointers to previous and next nodes.

[0032] A240: Similarly, temporarily store the first angular velocity data in the second bidirectional linked list.

[0033] A250: Similarly, control the motion state unit to intermittently collect the motion state data of the target vehicle, and map the obtained real-time state data sequence to the first bidirectional linked list and the second bidirectional linked list.

[0034] In one embodiment, the method steps provided by the present application further include:

[0035] A231: Predefined data buffer length.

[0036] A232: Initialize a standard bidirectional linked list using the data cache length to obtain the first bidirectional linked list, wherein the first bidirectional linked list is provided with a ring buffer.

[0037] A233: When temporarily storing the first acceleration data in the first bidirectional linked list, the first bidirectional linked list synchronously and dynamically removes historical data from the ring buffer.

[0038] Specifically, before data collection, this embodiment deploys the motion state unit on the target vehicle, wherein the motion state unit includes a three-axis acceleration sensor and a three-axis gyroscope.

[0039] The three-axis accelerometer collects real-time acceleration data in the X, Y, and Z directions (ACCx, ACCy, and ACCz), while the three-axis gyroscope collects real-time angular velocity data in the X, Y, and Z directions (Gyrox, Gyroy, and Gyroz). Together, these sensors enable the motion state unit to accurately record the vehicle's real-time motion state, providing raw data for subsequent rapid acceleration detection and vehicle condition monitoring.

[0040] To improve the accuracy and stability of data acquisition, the motion state unit requires bias calibration and temperature compensation before data acquisition. Specifically, the initial bias of the accelerometer and gyroscope is corrected to ensure that the data read by the sensors is close to the actual value. Sensor measurements can be affected by ambient temperature fluctuations. Temperature compensation corrects the impact of temperature fluctuations on sensor data, ensuring data accuracy under different temperature conditions.

[0041] In this embodiment, to ensure real-time and continuous data collection, a data sampling frequency (e.g., 50 Hz) is pre-set, and the size of the state collection window is determined based on this sampling frequency. The sampling frequency determines the time interval between each data collection, which in turn affects the length of the state collection window.

[0042] In this embodiment, the state acquisition window is used to constrain the data acquisition period of the motion state unit. Each time the acquisition is performed, the acceleration data and angular velocity data of the target vehicle are intermittently acquired according to the size of the acquisition window to form a real-time state data sequence.

[0043] The storage management method of the real-time status data sequence is as follows:

[0044] A first doubly linked list and a second doubly linked list are pre-built to store acceleration data and angular velocity data, respectively. Both the first doubly linked list and the second doubly linked list support efficient data storage and fast insertion / deletion operations.

[0045] After the target vehicle starts, the motion state unit is controlled to intermittently collect motion state data of the target vehicle according to a set collection window to obtain first real-time state data, which includes first acceleration data and first angular velocity data.

[0046] The first acceleration data is temporarily stored in the first bidirectional linked list, wherein the first cache node in the first bidirectional linked list for storing the first acceleration data includes pointers to previous and next nodes, and the pointers and the data stored in the nodes ensure the continuity of the data and the traversal efficiency of the linked list.

[0047] To ensure that the amount of data stored in the doubly linked list is appropriate, a cache length N (predefined data cache length) is set to prevent the linked list memory from being overused. The first doubly linked list is initialized with the predefined data cache length and a ring buffer is set for it. This way, when the cache reaches the maximum length, the new data in the first doubly linked list will overwrite the oldest data.

[0048] When temporarily storing the first acceleration data in the first bidirectional linked list, the first bidirectional linked list synchronously performs dynamic removal of historical data from the ring buffer. That is, when the data stored in the linked list reaches a set length N, the oldest data will be automatically removed, thereby maintaining the memory utilization and update efficiency of the linked list.

[0049] Similarly, the real-time angular velocity data is stored in a second doubly linked list. Each new data collection is inserted at the end of the linked list in chronological order. By storing acceleration and angular velocity data in a doubly linked list, the continuity and stability of the data sequence are ensured, facilitating subsequent rapid acceleration detection and analysis.

[0050] This embodiment achieves the technical effect of optimizing memory management, ensuring the real-time nature of vehicle driving status data, and avoiding delays or storage overflows caused by data backlogs by storing acceleration and angular velocity data in a bidirectional linked list and using a ring buffer to dynamically remove historical data.

[0051] A300: After performing multi-level data preprocessing on the real-time status data sequence, perform vehicle rapid acceleration detection based on the real-time status data sequence and output a frequent rapid acceleration warning, wherein the frequent rapid acceleration warning has a risk time interval identifier.

[0052] In one embodiment, Figure 2 As shown, vehicle rapid acceleration detection is performed based on the real-time status data sequence, and a frequent rapid acceleration warning is output, wherein the frequent rapid acceleration warning has a risk time interval identifier. The method step A300 provided in this application also includes:

[0053] A310: Extracting and obtaining real-time acceleration data from the first bidirectional linked list based on a sliding time window.

[0054] A320: Calculate the average of the real-time acceleration data, obtain the acceleration average, and compare the preset acceleration threshold with the acceleration average.

[0055] A330: If the preset acceleration threshold is less than the acceleration mean, the sudden acceleration detection mechanism is triggered.

[0056] A340: After the sudden acceleration detection mechanism is triggered, sudden acceleration behaviors are accumulated to obtain real-time behavior accumulation records.

[0057] A350: When the real-time behavior accumulation record meets the preset acceleration behavior threshold, generate and send the frequent sudden acceleration warning.

[0058] In one embodiment, if the preset acceleration threshold is less than the average acceleration, the sudden acceleration detection mechanism is triggered. Previously, step A300 of the method provided in this application further includes:

[0059] A331: Taking the acquisition time of the real-time acceleration data as a constraint, extract the real-time angular velocity data from the second bidirectional linked list based on a sliding time window.

[0060] A332: Predefines multi-dimensional angular velocity thresholds and deviation frequency thresholds.

[0061] A333: Using the multi-dimensional angular velocity threshold to traverse the real-time angular velocity data to record angular velocity deviations and obtain angular velocity deviation frequencies.

[0062] A334: If the angular velocity deviation frequency does not meet the deviation frequency threshold, triggering a sudden acceleration detection mechanism.

[0063] In this embodiment, the multi-level data preprocessing of the real-time status data sequence is essentially based on a low-pass filtering algorithm and a Kalman filtering algorithm to gradually filter out data noise and uncertainty at different levels, thereby improving the accuracy and stability of the data.

[0064] Specifically, the real-time collected acceleration data (ACCx, ACCy, ACCz) and angular velocity data (Gyrox, Gyroy, Gyroz) are used as input data. Conventional low-pass filtering and Kalman filtering are performed to optimize the accuracy of the real-time data to obtain optimized acceleration data and angular velocity data. Then, the optimized data is stored in a bidirectional linked list in real time. The stored data in the bidirectional linked list is used to subsequently determine whether the vehicle has frequently accelerated suddenly.

[0065] In this embodiment, the method for determining whether a vehicle has frequent sudden acceleration behavior based on the data recorded in the bidirectional linked list is as follows:

[0066] Real-time acceleration data (data series) is extracted from the first bidirectional linked list using a sliding time window. The acceleration data series reflects the acceleration changes of the vehicle within a specific time period. This sliding time window allows the latest acceleration data to be continuously obtained and processed from the linked list to monitor the vehicle's acceleration status in real time.

[0067] The extracted real-time acceleration data is averaged to smooth out short-term noise and sudden fluctuations, resulting in a more stable and accurate acceleration signal. The calculated average acceleration value is then compared with a preset acceleration threshold. If the average acceleration value exceeds the threshold, it indicates that the vehicle may be experiencing rapid acceleration. Step A330 triggers the rapid acceleration detection mechanism to further analyze the vehicle's behavior.

[0068] To enhance the accuracy and reliability of sudden acceleration detection, the present invention also incorporates analysis of angular velocity data. Specifically, based on the timestamps of the data sequences of the collected real-time acceleration data, real-time angular velocity data (such as Gyrox, Gyroy, and Gyroz) are extracted from a second bidirectional linked list. These angular velocity data are primarily used to analyze dynamic changes in the vehicle, especially on bumpy roads. The angular velocity data can help the system identify vibrations or shocks caused by uneven road surfaces.

[0069] Preset multi-dimensional angular velocity thresholds and deviation frequency thresholds are used to identify abnormal fluctuations in angular velocity data. The multi-dimensional angular velocity thresholds are specifically a set of angular velocity thresholds for different directions (or axes). These thresholds help determine whether the vehicle is experiencing vibrations caused by sharp turns, bumps, or uneven roads, thereby avoiding misjudgment of sudden acceleration.

[0070] By setting these thresholds, it's possible to detect angular velocity deviations caused by uneven road surfaces. If angular velocity fluctuations in multiple directions exceed a uniform threshold, it can be determined that the vehicle is currently traveling on an uneven road, rather than experiencing frequent sudden acceleration.

[0071] Based on this, after obtaining the real-time angular velocity data, the system traverses the real-time angular velocity data and records the deviation frequency. When the angular velocity deviation reaches the set frequency, it is determined that the current acceleration change deviating from the preset threshold is not caused by frequent sudden acceleration, but by the uneven road surface, rather than sudden acceleration behavior. If the angular velocity deviation frequency does not meet the set deviation frequency threshold, the system will trigger the sudden acceleration detection mechanism, ensuring that the warning is only triggered when the characteristics of sudden acceleration and uneven road section interact.

[0072] After the sudden acceleration detection mechanism is triggered, each time the acceleration is detected to exceed the preset threshold, the behavior will be recorded as a sudden acceleration event in the real-time behavior accumulation record. The record will be continuously updated to reflect the frequency of sudden acceleration behaviors that occur in the vehicle within a certain period of time.

[0073] When the real-time behavior accumulation record meets the preset acceleration behavior threshold, the frequent rapid acceleration warning is generated and sent.

[0074] Through high-frequency sampling and linked list cache management analysis, this embodiment can process and analyze acceleration data in real time, achieve a rapid response to frequent and sudden acceleration of the vehicle, and achieve the technical effect of making judgments and providing feedback in a very short time when frequent and sudden acceleration behaviors suddenly occur during driving, providing timely warning information to the driver, and helping to prevent potential dangers.

[0075] A400: Reversely call the vehicle behavior data of the target vehicle according to the risk time interval to obtain associated behavior data.

[0076] Specifically, in this embodiment, during the sudden acceleration detection process, once a potential sudden acceleration risk is identified, the vehicle's behavior data (such as vehicle acceleration, steering, braking, etc.) will be reversely queried based on the risk time interval (i.e., the time period when the sudden acceleration event occurs), providing the necessary information for subsequent vehicle condition detection.

[0077] A500: Perform vehicle condition detection based on the associated behavior data and output vehicle condition detection information.

[0078] In one embodiment, the vehicle condition detection is performed based on the associated behavior data, and the vehicle condition detection information is output. Step A500 of the method provided in this application further includes:

[0079] A510: Interactively obtain multiple sample association data combinations of multiple sample vehicle failures.

[0080] A520: Decompose the associated behavior data based on the plurality of sample associated data combinations as constraints to obtain a plurality of associated behavior arrays.

[0081] A530: Perform vehicle condition fault directional analysis based on the multiple associated behavior arrays to obtain multiple real-time fault analysis results.

[0082] A540: Using the multiple real-time fault analysis results, traverse the fault risk assignment table to obtain multiple real-time fault risk coefficients.

[0083] A550: Add the multiple real-time fault risk coefficients to obtain the vehicle condition detection information, wherein the vehicle condition detection information is identified by the multiple real-time fault analysis results.

[0084] In one embodiment, a vehicle fault directional analysis is performed based on the multiple associated behavior arrays to obtain multiple real-time fault analysis results. Step A530 of the method provided in this application further includes:

[0085] A531: Using the plurality of sample associated data combinations as constraints, collect fault vehicle condition data to obtain a plurality of sample fault behavior arrays.

[0086] A532: Utilize the plurality of sample vehicle faults and the plurality of sample fault behavior arrays for training to obtain a plurality of vehicle condition fault identification branches.

[0087] A533: Connect the multiple vehicle condition fault identification branches in parallel to complete the construction of the vehicle condition fault analysis model.

[0088] A534: Based on the mapping relationship between the multiple associated behavior arrays and multiple sample vehicle faults, the multiple associated behavior arrays are loaded into the multiple vehicle condition fault identification branches of the vehicle condition fault analysis model to simultaneously perform vehicle condition fault directional analysis to obtain the multiple real-time fault analysis results.

[0089] This embodiment uses the associated behavior data extracted from A400 to detect the vehicle condition to determine whether the vehicle has potential faults or abnormal operating conditions under frequent rapid acceleration.

[0090] Specifically, this embodiment constructs a vehicle condition fault analysis model for vehicle condition detection.

[0091] We interactively extract samples from fault data of different vehicles and combine them into multiple sample-related data sets. Each sample-related data set represents the vehicle condition data indicators required to determine whether a certain fault has occurred.

[0092] Using these multiple sample association data combinations as constraints, the system collects fault vehicle condition data, generating multiple sample fault behavior arrays that detail the vehicle's dynamic behavior under various fault scenarios. Fault scenarios may include engine failure, brake failure, or suspension system issues. All collected fault behavior data serves as input for the analysis model.

[0093] Based on multiple sample fault behavior arrays, the system uses machine learning or pattern recognition algorithms to build multiple vehicle condition fault identification branches. These branches represent independent identification logic for different fault types (such as engine failure and brake problems). During training, the algorithm extracts features from a large number of fault behavior arrays, learns the behavioral patterns of different fault types, and builds a dedicated identification model for each fault type. Each vehicle condition fault identification branch determines whether the vehicle is in a specific fault state based on input vehicle behavior data (such as acceleration and angular velocity).

[0094] The multiple vehicle condition and fault identification branches are connected in parallel to complete the construction of the vehicle condition and fault analysis model. This parallel structure allows all identification branches to run simultaneously, analyzing the behavioral data of the same vehicle and identifying multiple potential fault types. Each branch independently outputs a fault type determination result, which is then combined to generate a comprehensive vehicle condition analysis report.

[0095] The specific implementation method of performing vehicle condition fault analysis based on the vehicle condition fault analysis model is as follows:

[0096] Using these multiple sample association data combinations as constraints, the vehicle fault data is broken down and classified to obtain multiple association behavior arrays. Each sample association data combination represents the vehicle's behavior pattern under different fault scenarios and includes multiple vehicle condition data indicators (such as acceleration, angular velocity, steering angle, etc.).

[0097] Based on the mapping relationships between multiple associated behavior arrays and various sample vehicle faults, the multiple associated behavior arrays are loaded into the multiple vehicle condition fault identification branches of the vehicle condition fault analysis model for simultaneous analysis. Each associated behavior array represents vehicle behavior data for a specific fault type, and the mapping relationships between this data and different fault types have been established through training. Based on these mapping relationships, the associated behavior arrays are input into the corresponding vehicle condition fault identification branches for parallel fault-directed analysis. Each identification branch independently analyzes the input data, identifies the possible fault type, and outputs real-time fault analysis results, thereby obtaining the multiple real-time fault analysis results.

[0098] The fault risk assignment score table presets the risk assessment criteria for each fault type. The multiple real-time fault analysis results are used to traverse the fault risk assignment score table to obtain multiple real-time fault risk coefficients. The multiple real-time fault risk coefficients are added to obtain the vehicle condition detection information, wherein the vehicle condition detection information is identified by the multiple real-time fault analysis results.

[0099] After detecting frequent sudden acceleration, this embodiment can identify potential vehicle faults or abnormal conditions behind sudden acceleration through further analysis of associated behavior data. This achieves the technical effect of providing real-time vehicle health status reminders on the basis of sudden acceleration risk reminders, thereby reducing potential risks and ensuring the safety and stable operation of the vehicle.

[0100] A600: Update the warning level of the frequent rapid acceleration warning according to the vehicle condition detection information.

[0101] Specifically, in this embodiment, after frequent sudden acceleration behavior is detected, the severity of the warning is adjusted in real time according to the health status of the vehicle and the fault detection results, ensuring that the driver can make appropriate safety decisions based on the actual risk situation of the vehicle.

[0102] For example, after frequent sudden acceleration is detected, the vehicle condition detection result is scored as 20, and the original warning level is updated according to the correspondence between the scoring and the level increase table.

[0103] This embodiment achieves the technical effect of being able to make judgments and provide feedback in a very short time when frequent rapid acceleration occurs suddenly during driving, providing the driver with timely comprehensive warning information on driving behavior and vehicle conditions, thereby preventing potential driving hazards.

[0104] Example 2

[0105] Based on the same inventive concept as the vehicle rapid acceleration risk detection method in the aforementioned embodiment, Figure 3 As shown, the present application provides a vehicle rapid acceleration risk detection platform, wherein the platform includes:

[0106] The collection function configuration unit 11 is used to preset a status collection window.

[0107] The acquisition function operation unit 12 is used to control the motion state unit to perform intermittent acquisition of the motion state data of the target vehicle based on the state acquisition window as a constraint, so as to obtain a real-time state data sequence.

[0108] The driving warning output unit 13 is used to perform vehicle rapid acceleration detection according to the real-time status data sequence after multi-level data preprocessing, and output a frequent rapid acceleration warning, wherein the frequent rapid acceleration warning has a risk time interval identifier.

[0109] The behavior data collection unit 14 is configured to reversely call the vehicle behavior data of the target vehicle according to the risk time interval to obtain associated behavior data.

[0110] The vehicle condition detection execution unit 15 is used to perform vehicle condition detection according to the associated behavior data and output vehicle condition detection information.

[0111] The warning level updating unit 16 is configured to update the warning level of the frequent rapid acceleration warning according to the vehicle condition detection information.

[0112] In one embodiment, the acquisition function execution unit 12 further includes:

[0113] The motion state unit is deployed on the target vehicle, wherein the motion state unit includes a three-axis accelerometer and a three-axis gyroscope. The motion state unit is subjected to bias calibration and temperature compensation to localize the data acquisition function of the motion state unit. A data sampling frequency is preset, and a state acquisition window is set based on the data sampling frequency. Using the state acquisition window as a constraint, the motion state unit is controlled to intermittently acquire the motion state data of the target vehicle.

[0114] In one embodiment, the acquisition function execution unit 12 further includes:

[0115] Pre-construct a first bidirectional linked list and a second bidirectional linked list. After the target vehicle is started, the motion state unit is controlled to intermittently collect the motion state data of the target vehicle, with the state collection window as a constraint, to obtain first real-time state data, wherein the first real-time state data includes first acceleration data and first angular velocity data. The first acceleration data is temporarily stored in the first bidirectional linked list, wherein the first cache node in the first bidirectional linked list for storing the first acceleration data contains pointers to the previous and next nodes. Similarly, the first angular velocity data is temporarily stored in the second bidirectional linked list. Similarly, the motion state unit is controlled to intermittently collect the motion state data of the target vehicle, and the obtained real-time state data sequence is mapped and stored in the first bidirectional linked list and the second bidirectional linked list.

[0116] In one embodiment, the acquisition function execution unit 12 further includes:

[0117] A predefined data cache length is used. A standard doubly linked list is initialized using the data cache length to obtain the first doubly linked list, wherein the first doubly linked list is provided with a ring buffer. When temporarily storing the first acceleration data in the first doubly linked list, the first doubly linked list dynamically removes historical data from the ring buffer.

[0118] In one embodiment, the driving warning output unit 13 further includes:

[0119] Real-time acceleration data is extracted from the first bidirectional linked list based on a sliding time window. The real-time acceleration data is averaged to obtain the average acceleration value, and then the average acceleration value is compared with a preset acceleration threshold. If the preset acceleration threshold is less than the average acceleration value, a sudden acceleration detection mechanism is triggered. After the sudden acceleration detection mechanism is triggered, sudden acceleration behavior is accumulated to obtain a real-time accumulated record. When the real-time accumulated record meets the preset acceleration behavior threshold, a frequent sudden acceleration warning is generated and sent.

[0120] In one embodiment, the driving warning output unit 13 further includes:

[0121] Real-time angular velocity data is extracted from the second doubly linked list based on a sliding time window, using the real-time acceleration data collection time as a constraint. A multidimensional angular velocity threshold and a deviation frequency threshold are predefined. The real-time angular velocity data is traversed using the multidimensional angular velocity threshold to record angular velocity deviations and obtain an angular velocity deviation frequency. If the angular velocity deviation frequency does not meet the deviation frequency threshold, a sudden acceleration detection mechanism is triggered.

[0122] In one embodiment, the vehicle condition detection execution unit 15 further includes:

[0123] Interactively obtain multiple sample association data combinations for multiple sample vehicle faults. Decompose the association behavior data using the multiple sample association data combinations as constraints to obtain multiple association behavior arrays. Perform vehicle condition and fault directional analysis based on the multiple association behavior arrays to obtain multiple real-time fault analysis results. Utilize the multiple real-time fault analysis results to traverse a fault risk assignment table to obtain multiple real-time fault risk coefficients. Add the multiple real-time fault risk coefficients to obtain vehicle condition detection information, wherein the vehicle condition detection information is identified by the multiple real-time fault analysis results.

[0124] In one embodiment, the vehicle condition detection execution unit 15 further includes:

[0125] Using the multiple sample association data combinations as constraints, fault vehicle condition data is collected to obtain multiple sample fault behavior arrays. Multiple vehicle condition fault identification branches are trained using the multiple sample vehicle faults and the multiple sample fault behavior arrays. The multiple vehicle condition fault identification branches are connected in parallel to complete the construction of a vehicle condition fault analysis model. Based on the mapping relationship between the multiple association behavior arrays and the multiple sample vehicle faults, the multiple association behavior arrays are loaded into the multiple vehicle condition fault identification branches of the vehicle condition fault analysis model to simultaneously perform vehicle condition fault directional analysis and obtain the multiple real-time fault analysis results.

[0126] Example 3

[0127] Based on the same inventive concept as the vehicle rapid acceleration risk detection method in the aforementioned embodiment 1, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, it implements the steps of the vehicle rapid acceleration risk detection method described in any one of the aforementioned embodiments 1.

[0128] Any of the methods or steps described above may be stored as computer instructions or programs in various types of computer memories, and the computer instructions or programs may be recognized by various types of computer processors to implement any of the methods or steps described above.

[0129] Based on the above specific embodiments of the present invention, any improvements and modifications made to the present invention by those skilled in the art without departing from the principles of the present invention shall fall within the scope of patent protection of the present invention.

Claims

1. A vehicle rapid acceleration risk detection method, characterized in that: The method comprises: Preset status acquisition window; Using the state acquisition window as a constraint, controlling the motion state unit to intermittently acquire motion state data of the target vehicle to obtain a real-time state data sequence; After multi-level data preprocessing is performed on the real-time status data sequence, vehicle rapid acceleration detection is performed according to the real-time status data sequence, and a frequent rapid acceleration warning is output, wherein the frequent rapid acceleration warning has a risk time interval identifier; Reversely call the vehicle behavior data of the target vehicle according to the risk time interval to obtain associated behavior data; Perform vehicle condition detection based on the associated behavior data and output vehicle condition detection information; The warning level of the frequent rapid acceleration warning is updated according to the vehicle condition detection information.

2. The vehicle rapid acceleration risk detection method according to claim 1, wherein: The method includes: controlling the motion state unit to intermittently collect motion state data of the target vehicle using the state acquisition window as a constraint to obtain a real-time state data sequence. Deploying the motion state unit on the target vehicle, wherein the motion state unit includes a three-axis acceleration sensor and a three-axis gyroscope; performing deviation calibration and temperature compensation on the motion state unit to complete the localization of the data acquisition function of the motion state unit; Presetting a data sampling frequency, and setting the state acquisition window based on the data sampling frequency; The motion state unit is controlled to intermittently collect the motion state data of the target vehicle using the state collection window as a constraint.

3. The vehicle rapid acceleration risk detection method according to claim 2, wherein: Using the state acquisition window as a constraint, controlling the motion state unit to intermittently acquire motion state data of the target vehicle to obtain a real-time state data sequence, the method comprising: Pre-constructing a first doubly linked list and a second doubly linked list; After the target vehicle starts, controlling the motion state unit to intermittently collect motion state data of the target vehicle with the state collection window as a constraint to obtain first real-time state data, wherein the first real-time state data includes first acceleration data and first angular velocity data; Temporarily storing the first acceleration data in the first bidirectional linked list, wherein a first cache node in the first bidirectional linked list for storing the first acceleration data includes pointers to previous and next nodes; Similarly, the first angular velocity data is temporarily stored in the second bidirectional linked list; Similarly, the motion state unit is controlled to intermittently collect the motion state data of the target vehicle, and the obtained real-time state data sequence is mapped and stored in the first bidirectional linked list and the second bidirectional linked list.

4. The vehicle rapid acceleration risk detection method according to claim 3, wherein: The method comprises: Predefined data buffer length; Initializing a standard bidirectional linked list using the data cache length to obtain the first bidirectional linked list, wherein the first bidirectional linked list is provided with a ring buffer; When temporarily storing the first acceleration data in the first bidirectional linked list, the first bidirectional linked list synchronously performs dynamic removal of historical data in the ring buffer.

5. The vehicle rapid acceleration risk detection method according to claim 4, characterized in that: Performing vehicle rapid acceleration detection based on the real-time status data sequence and outputting a frequent rapid acceleration warning, wherein the frequent rapid acceleration warning has a risk time interval identifier, the method comprising: Extracting real-time acceleration data from the first bidirectional linked list based on a sliding time window; Calculating the mean of the real-time acceleration data to obtain the acceleration mean, and then comparing the preset acceleration threshold with the acceleration mean; If the preset acceleration threshold is less than the acceleration mean, the sudden acceleration detection mechanism is triggered; After the sudden acceleration detection mechanism is triggered, sudden acceleration behaviors are accumulated to obtain real-time behavior accumulation records; When the real-time behavior accumulation record meets the preset acceleration behavior threshold, the frequent rapid acceleration warning is generated and sent.

6. The vehicle rapid acceleration risk detection method according to claim 5, wherein: If the preset acceleration threshold is less than the acceleration mean, the sudden acceleration detection mechanism is triggered. Previously, the method includes: Taking the acquisition time of the real-time acceleration data as a constraint, extracting the real-time angular velocity data from the second bidirectional linked list based on a sliding time window; Predefine multi-dimensional angular velocity thresholds and deviation frequency thresholds; Using the multi-dimensional angular velocity threshold to traverse the real-time angular velocity data to record angular velocity deviations and obtain angular velocity deviation frequencies; If the angular velocity deviation frequency does not meet the deviation frequency threshold, the sudden acceleration detection mechanism is triggered.

7. The vehicle rapid acceleration risk detection method according to claim 1, wherein: Performing vehicle condition detection based on the associated behavior data and outputting vehicle condition detection information, the method includes: Interactively obtain multiple sample association data combinations of multiple sample vehicle faults; Decomposing the associated behavior data based on the combination of the plurality of sample associated data as constraints to obtain a plurality of associated behavior arrays; Performing vehicle condition fault directional analysis based on the multiple associated behavior arrays to obtain multiple real-time fault analysis results; Using the multiple real-time fault analysis results to traverse the fault risk assignment table to obtain multiple real-time fault risk coefficients; The multiple real-time fault risk coefficients are added to obtain the vehicle condition detection information, wherein the vehicle condition detection information is identified by the multiple real-time fault analysis results.

8. The vehicle rapid acceleration risk detection method according to claim 7, wherein: Performing vehicle condition fault directional analysis based on the multiple associated behavior arrays to obtain multiple real-time fault analysis results, the method comprising: Using the plurality of sample associated data combinations as constraints, collecting fault vehicle condition data to obtain a plurality of sample fault behavior arrays; Using the plurality of sample vehicle faults and the plurality of sample fault behavior arrays for training to obtain a plurality of vehicle condition fault identification branches; Connecting the plurality of vehicle condition fault identification branches in parallel to complete the construction of a vehicle condition fault analysis model; According to the mapping relationship between the multiple associated behavior arrays and multiple sample vehicle faults, the multiple associated behavior arrays are loaded into the multiple vehicle condition fault identification branches of the vehicle condition fault analysis model to simultaneously perform vehicle condition fault directional analysis to obtain the multiple real-time fault analysis results.

9. Vehicle rapid acceleration risk detection platform, characterized by: For implementing the method for detecting vehicle sudden acceleration risk according to any one of claims 1 to 8, the vehicle sudden acceleration risk detection platform comprises: A collection function configuration unit, used to preset a state collection window; An acquisition function operation unit, configured to control the motion state unit to intermittently acquire motion state data of the target vehicle using the state acquisition window as a constraint, and obtain a real-time state data sequence; a driving warning output unit, configured to perform vehicle rapid acceleration detection based on the real-time status data sequence after multi-level data preprocessing, and output a frequent rapid acceleration warning, wherein the frequent rapid acceleration warning has a risk time interval identifier; a behavior data collection unit, configured to reversely call the vehicle behavior data of the target vehicle according to the risk time interval to obtain associated behavior data; a vehicle condition detection execution unit, configured to perform vehicle condition detection based on the associated behavior data and output vehicle condition detection information; The warning level updating unit is used to update the warning level of the frequent rapid acceleration warning according to the vehicle condition detection information.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed, implements the steps of the vehicle rapid acceleration risk detection method according to any one of claims 1 to 8.

Images