A multi-resolution driving cycle recognition and classification method

Abstract

The application relates to the technical field of driving behavior analysis, and particularly provides a multi-resolution driving working condition identification and classification method. A plurality of time length sliding windows are set to perform multi-scale division on a multi-channel driving time sequence data stream, and a multi-resolution analysis system is constructed; longitudinal driving working conditions are identified and classified according to a yaw angular velocity energy index; a dynamic time warping algorithm is used to match data segments with transverse working condition templates, so that the identification and classification of the transverse driving working conditions are realized; and key driving events are calibrated, thereby providing fine data support for subsequent driving style analysis. The application combines the multi-scale sliding window and time sequence similarity measurement technology, effectively improves the driving working condition identification precision, and significantly reduces the misjudgment rate in complex scenes.

Description

Technical Field

[0001] This invention belongs to the field of driving behavior analysis technology, and in particular relates to a multi-resolution driving condition recognition and classification method. Background Technology

[0002] With the rapid development of intelligent connected vehicles and advanced driver assistance systems, onboard sensors can collect large amounts of multi-channel time-series data in real time, including steering wheel angle, accelerator and brake pedal travel, and vehicle dynamics parameters, providing a rich data foundation for driving condition identification and driving behavior analysis. Accurate identification of driving conditions is an important prerequisite for driving behavior analysis and driving style evaluation.

[0003] However, existing driving condition recognition technologies mainly have the following problems:

[0004] First, single time windows have limitations. Existing technologies mostly use fixed-length time windows for data analysis, making it difficult to adapt to the differences in time scales across various driving conditions. For example, transient situations such as emergency braking may only last 1-2 seconds, while complex situations such as lane changing and overtaking may require 5-10 seconds, and steady-state situations such as constant-speed cruising require even longer periods for accurate assessment. Using a single time window cannot effectively capture driving behavior characteristics at these different time scales simultaneously.

[0005] Secondly, the methods for identifying driving conditions are limited. Existing technologies typically use a uniform identification method to handle all types of driving conditions, neglecting the fundamental differences in data characteristics between longitudinal conditions (such as acceleration, deceleration, and constant speed) and lateral conditions (such as turning, lane changing, and weaving). Longitudinal conditions are mainly reflected in changes in indicators such as speed and acceleration, while lateral conditions are more characterized by temporal patterns of parameters such as steering wheel angle and lateral acceleration, requiring different identification strategies.

[0006] Furthermore, there is a lack of critical event identification mechanisms. Existing technologies often only provide a classification of the driving condition, lacking precise location and identification of critical driving events within that condition. This results in the inability to provide refined event-level data support for subsequent tasks such as assessing the intensity of driving behavior and analyzing driving styles.

[0007] In summary, due to the aforementioned problems, the existing technology has a low accuracy rate in identifying driving conditions in complex driving scenarios, especially when multiple driving conditions occur alternately or in combination, which can easily lead to misjudgments or omissions, thereby affecting the reliability of subsequent driving behavior analysis.

[0008] Therefore, there is an urgent need to develop a multi-resolution driving condition identification method that can adapt to different time scales, adopt differentiated identification strategies for different types of driving conditions, and accurately identify key driving events, so as to improve the accuracy and adaptability of driving condition identification and provide a high-quality data foundation for subsequent driving behavior analysis. Summary of the Invention

[0009] In view of this, the present invention aims to provide a multi-resolution driving condition recognition and classification method, constructs a multi-resolution driving condition analysis system, integrates multi-scale sliding windows and driving behavior categories, and coordinates time series similarity measurement to achieve high-precision classification of driving conditions, providing high-quality data support for subsequent driving style analysis.

[0010] To achieve the above objectives, the technical solution created by this invention is implemented as follows:

[0011] This invention provides a multi-resolution driving condition recognition and classification method, comprising:

[0012] S1: Multi-channel driving time-series data streams are collected by onboard sensors;

[0013] S2: Set up multiple sliding windows of different time lengths and the working condition category to be identified for each sliding window. Use the sliding windows to divide the multi-channel driving time-series data stream to obtain a data segment set containing driving time-series data segments of different time lengths.

[0014] S3: Set driving characteristic indicators, and select driving time series data segments corresponding to longitudinal driving conditions from the data segment set based on the driving characteristic indicators to realize the identification and classification of longitudinal driving conditions;

[0015] S4: Set a lateral driving condition template, and use the time series similarity measurement method to match the driving time series data segments in the data segment set with the lateral driving condition template, filter out the driving time series data segments corresponding to the lateral driving condition, and realize the identification and classification of lateral driving conditions.

[0016] S5: Calibrate key driving events in driving time sequence data segments under longitudinal and lateral driving conditions.

[0017] Preferably, the multi-channel driving time-series data stream includes: steering wheel angle, steering wheel angle direction indicator, vehicle speed, longitudinal acceleration, lateral acceleration, yaw rate, and timestamp.

[0018] Preferably, after the multi-channel driving time-series data stream is acquired by the vehicle-mounted sensors, the multi-channel driving time-series data stream is preprocessed. The preprocessing includes:

[0019] Align the driving time sequence data of each channel with timestamps, and unify the sampling frequency of the driving time sequence data of each channel to a set reference frequency;

[0020] The steering wheel angle is converted into the actual angle with direction indicator based on the steering wheel angle direction marking.

[0021] Preferably, sliding windows with different time lengths include:

[0022] Short-scale sliding window, with a window length of 50 to 200 sampling points;

[0023] Mesoscale sliding window, with a window length of 200 to 500 sampling points;

[0024] Long-scale sliding window, with a window length of 500 to 2000 sampling points;

[0025] In this configuration, the step size for all sliding windows is set to 20% of the window length.

[0026] Short-scale sliding windows are used to identify key events, which include: instantaneous steering, rapid acceleration, and sudden braking.

[0027] The mesoscale sliding window is used to identify lateral driving conditions, which include: lane changing, turning, and overtaking.

[0028] Long-scale sliding windows are used to identify longitudinal driving conditions, which include categories such as following other vehicles and free driving.

[0029] Preferred identification of longitudinal driving conditions includes:

[0030] The driving characteristic index is yaw rate. The average energy of the yaw rate of long-scale driving time-series data segments obtained by long-scale sliding window partitioning is calculated. for:

[0031] ;

[0032] in, Indicates the first The square of the yaw rate at the sampling time, where N represents the total number of sampling points. The average energy of the yaw rate The length of the energy calculation interval, , This indicates the length of the long-scale sliding window. The sampling interval represents the yaw rate;

[0033] When the average energy of the yaw rate If the energy level is below the set threshold, the corresponding time is considered to be a longitudinal driving condition.

[0034] Preferably, longitudinal driving conditions are classified according to following distance, including:

[0035] Set a following distance threshold, when the average energy of the yaw rate... If the energy level is below the set threshold and the actual following distance is less than the following distance threshold, it is judged as a following driving condition.

[0036] When the average energy of the yaw rate If the energy level is below the set threshold and the actual following distance is greater than the following distance threshold or there is no vehicle in front, it is judged as a free driving condition.

[0037] Preferably, the lateral driving condition template includes:

[0038] Templates for lane changing, turning, and overtaking.

[0039] Preferably, the identification and classification of lateral driving conditions includes:

[0040] The time series similarity measurement method is the dynamic time warping algorithm;

[0041] A dynamic threshold is set for each lateral driving condition template. If the DTW distance between a mesoscale driving time series data segment and any lateral driving condition template is less than the dynamic threshold of the lateral driving condition template, and the physical constraints of the lateral driving condition template are met, then the lateral driving condition category corresponding to the lateral driving condition template for the mesoscale driving time series data segment is determined.

[0042] If a mesoscale driving time series data segment simultaneously satisfies multiple lateral driving condition templates, the lateral driving condition template with the smallest DTW distance is selected as the lateral driving condition category corresponding to that mesoscale driving time series data segment.

[0043] Preferably, the formula for calculating DTW distance is:

[0044] ;

[0045] in, A sequence of one or more channels representing a segment of mesoscale driving time-series data. This represents one or more channel sequences for a specific lateral driving condition template. This indicates that the dynamic time warping algorithm obtains the optimal planned path. Represents distance metric, Indicates the first One or more channels of a mesoscale driving time-series data segment at the sampling time. Indicates the first One or more channels of a lateral driving condition template at a sampling time.

[0046] Preferred key driving events for calibrating longitudinal and lateral driving conditions include:

[0047] Key driving events include: instantaneous steering events, rapid acceleration events, and sudden braking events;

[0048] When a longitudinal acceleration greater than a set rapid acceleration threshold exists in a short-scale driving time-series data segment obtained by short-scale sliding window division, it is determined that there is a rapid acceleration event in the short-scale driving time-series data segment.

[0049] When there is a longitudinal acceleration less than a set emergency braking threshold in a short-scale driving time-series data segment obtained by short-scale sliding window division, it is determined that there is an emergency braking event in the short-scale driving time-series data segment.

[0050] When the speed of change of steering wheel angle in the short-scale driving time-series data segment obtained by short-scale sliding window division is greater than the set threshold for the speed of change of steering wheel angle, it is determined that there is an instantaneous steering event in the short-scale driving time-series data segment.

[0051] Compared with the prior art, the present invention can achieve the following beneficial effects:

[0052] To address the issue of fragmented time scales in existing technologies, this invention constructs a multi-time-scale fusion framework. By setting a multi-scale sliding window with 50-2000 sampling points (corresponding to a time span of 0.5 to 20 seconds, based on a 100Hz sampling frequency), it achieves full coverage from instantaneous operations to continuous driving conditions. Furthermore, it integrates dynamic range adjustment and recursive segmentation matching techniques, along with a rule engine and time series similarity measurement algorithm, to achieve high-precision identification and segmentation of driving conditions. This effectively improves adaptability to complex scenarios and reduces the misjudgment rate for complex conditions. This avoids information omissions caused by fixed windows, which helps engineers take more effective measures when designing and maintaining intelligent driving systems, improving the system's adaptability and robustness to complex driving scenarios.

[0053] This invention, by integrating yaw rate energy threshold determination with the DTW dynamic time warping algorithm, can accurately identify longitudinal and lateral driving conditions. Compared with existing fixed-rule methods, this invention significantly reduces the misjudgment rate in complex scenarios (such as curves with gradually changing curvature), providing a high-quality data foundation for subsequent driving style analysis. Attached Figure Description

[0054] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0055] Figure 1This is a schematic diagram of a multi-resolution driving condition recognition and classification method provided in an embodiment of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and do not constitute a limitation thereof. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the invention. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, some operations related to the invention are not shown or described in the specification. This is to avoid obscuring the core parts of the invention with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0057] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined to form various implementations. Furthermore, the order of the steps or actions in the method description can be changed or adjusted in a manner readily apparent to those skilled in the art. Therefore, the various orders in the specification and drawings are merely for the clear description of a particular embodiment and do not imply a mandatory order, unless otherwise stated that a particular order must be followed.

[0058] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on this invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0059] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0060] The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0061] Please see Figure 1 In one embodiment of the present invention, a multi-resolution driving condition recognition and classification method is provided. By constructing a multi-resolution analysis system, this method solves the problems of high misjudgment rate and poor class data quality in traditional methods under complex scenarios. The method includes:

[0062] S1: Multi-channel driving time-series data streams are collected by onboard sensors.

[0063] Specifically, the process begins by acquiring the raw multi-channel driving time-series data stream from the vehicle's sensors. This data stream must include the following channels: steering wheel angle, steering wheel angle sign, vehicle speed, longitudinal acceleration, lateral acceleration, yaw rate, and timestamp (segment_id). In other words, the multi-channel driving time-series data stream consists of at least the following data: steering wheel angle data, steering wheel angle sign data, vehicle speed data, longitudinal acceleration data, lateral acceleration data, yaw rate data, and timestamp data.

[0064] Furthermore, after obtaining the multi-channel driving time-series data stream, data preprocessing is usually required. Preprocessing includes: aligning the driving time-series data of each channel with timestamps, and unifying the sampling frequency of the driving time-series data of each channel to a set reference frequency.

[0065] Specifically, since vehicle-mounted sensors may have different sampling frequencies and time delays, it is necessary to perform time synchronization processing on the multi-channel driving time-series data before windowing, unifying the sampling frequency to a set reference frequency. The sampling frequency unification method includes: ensuring the time consistency of driving time-series data in each channel through timestamp alignment, and setting a reference frequency. In this embodiment of the invention, the reference frequency is set to 100Hz. By resampling the channel data with different sampling rates and interpolating missing data, the continuity of the data is ensured, unifying the driving time-series data in each channel to 100Hz.

[0066] S2: Set multiple sliding windows of different time lengths and the working condition category to be identified for each sliding window. Use the sliding windows to divide the multi-channel driving time-series data stream to obtain a set of data segments containing driving time-series data segments of different time lengths.

[0067] Specifically, a dynamic sliding window method is used to segment driving time-series data. First, multiple time scales and sliding windows of different durations are set according to different analysis needs, and each sliding window of a certain duration is assigned a corresponding driving behavior category.

[0068] In this embodiment of the invention, three time scales are defined, corresponding to three sliding windows of different time lengths, including:

[0069] Short-scale sliding window, with a window length of 50 to 200 sampling points (window length corresponds to 0.5 to 2 seconds, sampling frequency 100Hz);

[0070] Mesoscale sliding window, with a window length of 200 to 500 sampling points (window length corresponds to 2 to 5 seconds);

[0071] A long-scale sliding window, with a window length of 500 to 2000 sampling points, corresponding to 5 to 20 seconds.

[0072] Short-scale sliding windows are used to identify key events, including: instantaneous steering, rapid acceleration, and sudden braking.

[0073] The mesoscale sliding window is used to identify lateral driving conditions, including: lane change, turning, and overtaking.

[0074] Long-scale sliding windows are used to identify longitudinal driving conditions, including: following other vehicles and free driving.

[0075] For each time scale, during the data stream partitioning process, the window movement step size for all sliding windows is set to 20% of the window length, achieving an 80% window overlap between adjacent sliding windows. This high overlap design ensures sufficient sampling of driving behavior and avoids missing key driving events. During window generation, the system simultaneously records metadata information for each window, including: window start index, end index, data segment identifier, and average vehicle speed within the window. This metadata provides crucial contextual information for subsequent condition identification and feature extraction.

[0076] The multi-channel driving time-series data stream is divided and classified using sliding windows of the three time lengths mentioned above. Each window division yields a driving time-series data segment of a corresponding time length, thus obtaining a set of driving time-series data segments of three different time lengths.

[0077] S3: Set driving characteristic indicators, and select driving time series data segments corresponding to longitudinal driving conditions from the data segment set based on the driving characteristic indicators to realize the identification and classification of longitudinal driving conditions.

[0078] Specifically, driving characteristic indicators and thresholds for driving indicators are set to extract longitudinal driving conditions. In this embodiment of the invention, yaw rate is selected as the driving characteristic indicator. The longitudinal driving condition is identified by calculating the average energy of the yaw rate of long-scale driving time-series data segments obtained by long-scale sliding window division, which serves as the basis for dividing the longitudinal and lateral driving conditions.

[0079] Specifically, the average energy of the yaw rate for:

[0080] ;

[0081] in, Indicates the first The square of the yaw rate at the sampling time, where N represents the total number of sampling points. The average energy of the yaw rate The length of the energy calculation interval, , This indicates the length of the long-scale sliding window. The sampling interval represents the yaw rate.

[0082] Furthermore, an energy setting threshold for judging longitudinal driving conditions is determined. Based on statistical analysis of a large amount of actual driving data, it was found that an energy setting threshold of 0.5 rad² / s for yaw rate can accurately distinguish between longitudinal and lateral driving conditions. In this embodiment of the invention, when the calculated average energy of the yaw rate... If the energy level is below the set threshold of 0.5 rad² / s, the corresponding time is considered to be a longitudinal driving condition.

[0083] Furthermore, longitudinal driving conditions are categorized based on following distance, including:

[0084] The following distance threshold is set at 50 meters, and the average energy of the yaw rate is... If the energy level is below the set threshold of 0.5 rad² / s and the actual following distance is less than 50 meters, it is classified as a following-vehicle driving condition. Under this condition, the driver mainly controls the longitudinal speed to maintain a safe distance from the vehicle in front.

[0085] When the average energy of the yaw rate If the energy level is below the set threshold and the actual following distance is greater than 50 meters or there is no vehicle in front, it is considered a free driving condition. Under this condition, the driver can freely choose the driving speed, which is mainly characterized by stable longitudinal driving.

[0086] Furthermore, additional constraints need to be set to ensure the continuity of the operating conditions and prevent noise from causing frequent switching between operating conditions. Specifically, a data segment is only marked as a longitudinal operating condition if 100 consecutive sampling points (1 second) meet the longitudinal operating condition conditions. For brief fluctuations in yaw rate energy, the system will perform smoothing processing to maintain the stability of the operating condition identification.

[0087] S4: Set a lateral driving condition template, and use the time series similarity measurement method to match the driving time series data segments in the data segment set with the lateral driving condition template, filter out the driving time series data segments corresponding to the lateral driving condition, and realize the identification and classification of lateral driving conditions.

[0088] Specifically, this invention provides three typical lateral driving scenario templates: lane-changing, turning, and overtaking. The lane-changing template is set based on typical steering wheel angle sequences extracted from actual lane-changing data, with a steering wheel turning time of 3 to 5 seconds, an angle change range of 15-90 degrees, and a sensitivity coefficient of 14. The turning template is set based on angle sequences extracted from intersection turning data, with a steering wheel turning time of 4 to 8 seconds, an angle change range of 25-120 degrees, and a sensitivity coefficient of 13. The overtaking template is set based on angle sequences extracted from highway overtaking data, with a steering wheel turning time of 5 to 10 seconds, an angle change range of 20-60 degrees, and a sensitivity coefficient of 11.

[0089] This invention employs time series similarity metrics to filter driving time series data segments corresponding to lateral driving conditions from a set of data segments. Optional time series similarity metrics include Euclidean distance, Dynamic Time Warping (DTW), Time Warp Edit Distance (TWED), cosine similarity, Pearson correlation coefficient, and Symbolic Aggregate Approximation (SAX). In this embodiment, Dynamic Time Warping is specifically chosen. To improve computational efficiency, a GPU-based batch parallel DTW algorithm is used, parallelizing the traditional serial DTW computation and simultaneously processing the matching calculations of multiple windows and templates. Specifically, the window data to be matched (i.e., the mesoscale driving time series data segments obtained by mesoscale sliding window partitioning) and the lateral driving condition template data are first transferred to GPU memory. CUDA tensor operations are used to calculate the point-to-point distance matrix between all windows and templates. Then, a dynamic programming method is used to solve for the optimal alignment path, calculating the DTW distance between the mesoscale driving time series data segments obtained by mesoscale sliding window partitioning and each lateral driving condition template. The batch size is set to 1024, which can be dynamically adjusted according to the GPU memory capacity. Then, for each horizontal sub-case, the dynamic time warping distance (DTW) between the template and the corresponding scale fragment set is calculated. The DTW algorithm calculates the optimal alignment path and cumulative distance for each pair of channels. Typically, the distances of multiple channels are fused (e.g., weighted summation, taking the maximum value) to obtain a comprehensive DTW distance value, which is calculated as follows:

[0090] ;

[0091] in, A sequence of one or more channels representing a segment of mesoscale driving time-series data. This represents one or more channel sequences for a specific lateral driving condition template. This indicates that the dynamic time warping algorithm obtains the optimal planned path. Represents distance metric, Indicates the first One or more channels of a mesoscale driving time-series data segment at the sampling time. Indicates the first One or more channels of a lateral driving condition template at a sampling time.

[0092] Furthermore, a dynamic threshold is set for each lateral driving condition template. The setting of the dynamic threshold needs to take into account the noise level of each template. The specific dynamic threshold calculation process is as follows:

[0093] First, the templates are Z-score standardized, and the median of the absolute value of the first-order difference of the standardized templates is used as an estimate of the noise level. Then, the noise level, template duration, and configuration sensitivity coefficient are multiplied to obtain the dynamic threshold. The dynamic threshold calculated by this method can adapt to the characteristics of different templates. Considering that left turns and right turns exhibit mirror symmetry in steering wheel angle, this invention generates a corresponding mirror template for each basic template. The mirror template is obtained by inverting the angle value of the original template. This design allows the system to simultaneously recognize symmetrical driving actions without defining a separate template for each direction. The dynamic thresholds corresponding to lane-changing, turning, and overtaking scenarios can be calculated using the above method. For any mesoscale driving time-series data segment, its DTW distance with the lateral driving scenario template is calculated. If the calculated DTW distance is less than the dynamic threshold set by the corresponding lateral driving scenario template, the mesoscale driving time-series data segment is determined to be a lateral scenario and belongs to the driving scenario represented by that lateral driving scenario template.

[0094] Furthermore, a composite condition verification based on physical constraints is introduced to improve recognition accuracy. The requirements are set as follows: lane changing requires a steering wheel angle change between 15-90 degrees; turning requires 25-120 degrees; overtaking requires 20-60 degrees; the average vehicle speed for lane changing is required to be no less than 20 m / s; for turning, no less than 15 m / s; and for overtaking, no less than 25 m / s. In addition, duration constraints can be set to ensure that the duration of each condition is within a reasonable range, avoiding misidentification of transient disturbances as complete conditions. Only when the data from the sampling points of the mesoscale driving time-series data segment simultaneously meets the above physical constraints will the corresponding mesoscale driving time-series data segment be labeled as the corresponding type of lateral condition.

[0095] S5: Calibrates key driving events in driving time sequence data segments under longitudinal and lateral driving conditions, providing refined event-level data support for subsequent tasks such as driving behavior intensity assessment and driving style analysis.

[0096] Specifically, for short-scale driving time-series data segments obtained by short-scale sliding window partitioning, since the time length of these segments is relatively short, they will be distributed within the intervals of long-scale or medium-scale driving time-series data segments. In other words, the short-scale driving time-series data segments are located in lateral or longitudinal driving conditions. By calibrating the key driving events of the short-scale driving time-series data segments, the key driving events of both longitudinal and lateral driving conditions can be calibrated.

[0097] Three types of key driving events are defined: instantaneous steering events, rapid acceleration events, and emergency braking events. A rapid acceleration event is determined to exist in a short-scale driving time-series data segment obtained by dividing a short-scale sliding window if the longitudinal acceleration exceeds a set rapid acceleration threshold; an emergency braking event is determined to exist in a short-scale driving time-series data segment if the longitudinal acceleration is less than a set emergency braking threshold; and an instantaneous steering event is determined to exist in a short-scale driving time-series data segment if the speed of change of steering wheel angle exceeds a set steering wheel angle change speed threshold.

[0098] Specifically, rapid acceleration events are defined as moments when longitudinal acceleration exceeds 2.5 m / s², and the number, duration, and frequency of rapid acceleration events are recorded for each operating condition segment. Emergency braking events are defined as moments when longitudinal acceleration exceeds -2.5 m / s², and the number, duration, and frequency of emergency braking events are recorded for each operating condition segment. Instantaneous steering events are defined as moments when the absolute value of the steering wheel angular velocity exceeds 200° / s, etc.

[0099] The three key driving events mentioned above can be further flagged as indicators of dangerous driving behavior. The number of events is further calculated based on the dangerous driving behavior event indicator function. :

[0100] ;

[0101] Event frequency per unit time :

[0102] ;

[0103] in, This indicates the total duration of the segment in question.

[0104] Event Time Percentage :

[0105] ;

[0106] in, This indicates that the sampling point at that time was marked as a critical driving event.

[0107] Key driving events reflect a driver's emergency avoidance or aggressive driving behavior. Based on event statistics, the system calculates derived characteristics such as event frequency (events / s) and event duration, comprehensively characterizing the degree of aggressive driving.

[0108] Through the above-mentioned multi-resolution driving condition recognition and classification, driving data can be classified in detail at multiple time scales, providing important data support for subsequent driving style analysis. This solves the problems of coarse classification in traditional driving data processing, significant misjudgment rate in complex scenarios (such as curves with gradually changing curvature), inability to distinguish between multiple driving conditions existing simultaneously, or dangerous driving behaviors in driving conditions.

[0109] In summary, the above description is merely a preferred embodiment of this specification and is not intended to limit the scope of protection of this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of protection of this specification.

[0110] The systems, apparatuses, modules, or units described in one or more of the above embodiments may be implemented by a computer chip or entity, or by a product having a certain function. A typical implementation device is a computer. Specifically, a computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0111] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0112] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

Claims

1. A multi-resolution driving condition recognition and classification method, characterized in that, include: S1: Multi-channel driving time-series data streams are collected by onboard sensors; S2: Set multiple sliding windows of different time lengths and the operating condition category to be identified for each sliding window; use the sliding windows to divide the multi-channel driving time-series data stream to obtain a data segment set containing driving time-series data segments of different time lengths; the multiple sliding windows of different time lengths include: Short-scale sliding window, with a window length of 50 to 200 sampling points; Mesoscale sliding window, with a window length of 200 to 500 sampling points; Long-scale sliding window, with a window length of 500 to 2000 sampling points; In this configuration, the step size for all sliding windows is set to 20% of the window length. The short-scale sliding window is used to identify key events, and the categories of key events include: instantaneous steering, rapid acceleration, and sudden braking; The mesoscale sliding window is used to identify lateral driving conditions, and the categories of lateral driving conditions include: lane changing, turning, and overtaking; The long-scale sliding window is used to identify longitudinal driving conditions, and the categories of longitudinal driving conditions include: following other vehicles and free driving; S3: Set driving characteristic indicators, and based on these indicators, filter driving time-series data segments corresponding to longitudinal driving conditions from the data segment set to achieve the identification and classification of longitudinal driving conditions; the identification of longitudinal driving conditions includes: The driving characteristic index is yaw rate, and the average energy of the yaw rate of the long-scale driving time-series data segment obtained by the long-scale sliding window division is calculated. for: ； in, Indicates the first The square of the yaw rate at the sampling time, where N represents the total number of sampling points. The average energy of the yaw rate The length of the energy calculation interval, , This represents the length of the long-scale sliding window. The sampling interval represents the yaw rate; When the average energy of the yaw rate If the energy level is below the set threshold, the corresponding time is considered to be a longitudinal driving condition. S4: Set a lateral driving condition template, and use a time series similarity measurement method to match the driving time series data segments in the data segment set with the lateral driving condition template, filtering out the driving time series data segments corresponding to the lateral driving conditions, thereby realizing the identification and classification of lateral driving conditions; the lateral driving condition template includes: lane changing condition template, turning condition template, and overtaking condition template; the identification and classification of lateral driving conditions includes: The time series similarity measurement method is the dynamic time warping algorithm; The Dynamic Time Warping (DTW) distance between the mesoscale driving time-series data segments obtained by mesoscale sliding window partitioning and each lateral driving condition template is calculated using the Dynamic Time Warping (DTW) algorithm. A dynamic threshold is set for each lateral driving condition template. If the DTW distance between a mesoscale driving time series data segment and any lateral driving condition template is less than the dynamic threshold of the lateral driving condition template, and the physical constraints of the lateral driving condition template are met, then the lateral driving condition category corresponding to the lateral driving condition template for the mesoscale driving time series data segment is determined. If a mesoscale driving time series data segment simultaneously satisfies multiple lateral driving condition templates, the lateral driving condition template with the smallest DTW distance shall be selected as the lateral condition category corresponding to the mesoscale driving time series data segment. S5: Calibrate key driving events in driving time sequence data segments under longitudinal and lateral driving conditions.

2. The multi-resolution driving condition recognition and classification method according to claim 1, characterized in that, The multi-channel driving time-series data stream includes: steering wheel angle, steering wheel angle direction indicator, vehicle speed, longitudinal acceleration, lateral acceleration, yaw rate, and timestamp.

3. The multi-resolution driving condition recognition and classification method according to claim 2, characterized in that, After the multi-channel driving time-series data stream is acquired by the vehicle-mounted sensors, the multi-channel driving time-series data stream is first preprocessed, and the preprocessing includes: Align the driving time sequence data of each channel with timestamps, and unify the sampling frequency of the driving time sequence data of each channel to a set reference frequency; The steering wheel angle is converted into the actual angle with direction indicator based on the steering wheel angle direction marking.

4. The multi-resolution driving condition recognition and classification method according to claim 1, characterized in that, Based on following distance, longitudinal driving conditions are classified into the following categories: Set a following distance threshold, when the average energy of the yaw rate... If the energy level is below the set threshold and the actual following distance is less than the following distance threshold, it is determined to be a following driving condition. When the average energy of the yaw rate If the energy level is below the set threshold and the actual following distance is greater than the set following distance threshold or there is no vehicle in front, it is determined to be a free driving condition.

5. The multi-resolution driving condition recognition and classification method according to claim 1, characterized in that, The formula for calculating the DTW distance is: ； in, A sequence of one or more channels representing a segment of mesoscale driving time-series data. This represents one or more channel sequences for a specific lateral driving condition template. This indicates that the dynamic time warping algorithm obtains the optimal planned path. Represents distance metric, Indicates the first One or more channels of a mesoscale driving time-series data segment at the sampling time. Indicates the first One or more channels of a lateral driving condition template at a sampling time.

6. The multi-resolution driving condition recognition and classification method according to claim 1, characterized in that, The key driving events for calibrating longitudinal and lateral driving conditions include: The key driving events include: instantaneous steering events, rapid acceleration events, and emergency braking events; When a longitudinal acceleration greater than a set rapid acceleration threshold exists in a short-scale driving time-series data segment obtained by short-scale sliding window division, it is determined that there is a rapid acceleration event in the short-scale driving time-series data segment. When there is a longitudinal acceleration less than a set emergency braking threshold in a short-scale driving time-series data segment obtained by short-scale sliding window division, it is determined that there is an emergency braking event in the short-scale driving time-series data segment. When the speed of change of steering wheel angle in the short-scale driving time-series data segment obtained by short-scale sliding window division is greater than the set threshold for the speed of change of steering wheel angle, it is determined that there is an instantaneous steering event in the short-scale driving time-series data segment.

Images