A method, device, and system for identifying and processing abnormal vehicle events on highways.

Abstract

This invention belongs to the field of vehicle anomaly identification and processing technology. Specifically, it discloses a method, device, and system for identifying and processing abnormal vehicle events on highways. The method includes: identifying the existence of abnormal following events based on the speed and following distance of the target vehicle and the vehicle in front; if such events exist, analyzing the trend of following distance changes; when the trend decreases, determining a safe following distance based on vehicle speed and road adhesion coefficient; acquiring the speed and following distance of each following vehicle, predicting future changes in their following distance, calculating the rear-end collision risk coefficient, and determining the required deceleration; acquiring lane-changing behavior records and current motion states of vehicles in adjacent lanes, determining the lane-changing interference coefficient, and calculating the corrected safe distance and corrected deceleration accordingly; and controlling the target vehicle to reduce its speed according to the corrected deceleration until the following distance meets the requirements. This invention achieves integrated control and closed-loop safety adjustment of front, rear, and lateral risks under multi-vehicle interaction, improving highway driving safety.

Description

Technical Field

[0001] This invention belongs to the field of vehicle anomaly identification and processing technology, and relates to a method, device and system for identifying and processing abnormal vehicle events on highways. Background Technology

[0002] The identification and processing of abnormal vehicle events on highways is an important research area in the fields of intelligent traffic management and active safety control. During high-speed travel, factors such as following too closely, increased risk of rear-end collisions, and frequent lane changes by vehicles in adjacent lanes can easily lead to rear-end collisions and side collisions. Currently, some technical solutions use onboard sensors or roadside equipment to monitor vehicle driving status and achieve a certain degree of identification of abnormal driving behaviors.

[0003] For example, Chinese invention patent CN117152968A discloses a method, device, equipment, and medium for identifying abnormal vehicles on highways. The method includes: obtaining vehicle transaction data from a distributed message queue; integrating and archiving the transaction data using a streaming processing framework and statistically analyzing vehicle attribute information; establishing a highway network model based on digital twins; and using the highway network model combined with the statistically analyzed vehicle attribute information to trace and track vehicles, identifying vehicles with abnormal toll collection or behavior. This method combines big data analysis with a road network model, enabling rapid and accurate identification of abnormal vehicles and facilitating vehicle behavior reconstruction and toll collection.

[0004] However, the aforementioned existing technologies have at least two shortcomings: First, the above solutions mainly rely on digital twin road network models and historical vehicle flow data to trace and identify abnormal vehicle behaviors after they have occurred. As a result, they cannot respond in real time to the dynamically changing following risks while the vehicle is traveling at high speed. At the same time, they cannot issue active deceleration control commands to the target vehicle at the beginning of an abnormal event, thus making it difficult to effectively prevent rear-end collisions caused by insufficient following distance.

[0005] Second, the above-mentioned scheme only identifies anomalies based on the flow data of a single vehicle and the road network model. Its focus is limited to the driving parameters and historical trajectory of the target vehicle itself, ignoring the rear-end collision risk of vehicles behind and the impact of lane-changing interference from vehicles in adjacent lanes on the safe driving of the target vehicle. As a result, in complex multi-vehicle interaction environments, the vehicle cannot comprehensively weigh the risks of following the vehicle in front and the risks behind and to the sides, thus resulting in insufficient overall safety control.

[0006] Therefore, there is an urgent need for a method for identifying and processing abnormal vehicle events on highways that can integrate the risks of following other vehicles, rear-end collisions, and lateral lane change interference in real time under dynamic driving conditions on highways, and can achieve active closed-loop safety control of the target vehicle, in order to solve the above-mentioned technical problems. Summary of the Invention

[0007] In view of this, in order to solve the problems mentioned in the background art, a method, device and system for identifying and processing abnormal vehicle events on highways are proposed.

[0008] The objective of this invention can be achieved through the following technical solution: This invention provides a method for identifying and processing abnormal vehicle events on highways, including: identifying whether there is an abnormal following event based on the speed and following distance of the target vehicle and the vehicle in front; if so, identifying the trend of change based on the following distance within the preceding time window.

[0009] When the trend is decreasing, the safe following distance is determined based on the speed of the target vehicle and the vehicle in front, as well as the road adhesion coefficient.

[0010] The system obtains the speed and following distance of each vehicle within a preset distance range in the preceding time window, predicts the changes in speed and following distance of each vehicle in the future period, calculates the rear-end collision risk coefficient of the following vehicle, and determines the required deceleration in combination with the safe following distance.

[0011] The system acquires records of lane-changing behavior and current motion status of vehicles in adjacent lanes within a preceding time window, determines the lane-changing interference coefficient, and calculates the corrected safe distance and corrected deceleration based on the lane-changing interference coefficient.

[0012] Control the target vehicle to reduce its speed by a corrected deceleration until the following distance is greater than or equal to the real-time updated corrected safe distance.

[0013] The present invention also provides a highway vehicle abnormal event identification and processing device, which includes a processor and a memory. The memory stores a computer program, and the processor executes the computer program to implement a highway vehicle abnormal event identification and processing method.

[0014] The present invention also provides a highway vehicle abnormal event identification and processing system, including: an abnormal trend identification module, which identifies whether there is an abnormal following event based on the speed and following distance of the target vehicle and the vehicle in front; if so, it identifies the changing trend of the following distance within the preceding time window.

[0015] The safe following distance determination module determines the safe following distance based on the speed of the target vehicle and the vehicle in front, as well as the road adhesion coefficient, when the trend of change is decreasing.

[0016] The rear vehicle risk calculation module obtains the speed and following distance of each vehicle within a preset distance range in the preceding time window, predicts the speed and following distance changes of each vehicle in the future period (e.g., 5 seconds), calculates the rear-end collision risk coefficient of the rear vehicle, and determines the required deceleration in combination with the safe following distance.

[0017] The lane change interference correction module acquires the lane change behavior records and current motion status of vehicles in adjacent lanes within the preceding time window, determines the lane change interference coefficient accordingly, and calculates the corrected safe distance and corrected deceleration based on the lane change interference coefficient.

[0018] The speed reduction control terminal controls the target vehicle to reduce its speed by a corrected deceleration until the following distance is greater than or equal to the corrected safe distance updated in real time.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention identifies abnormal following events based on the speed and following distance of the target vehicle and the vehicle in front, and further analyzes the trend of following distance change within the preceding time window when there is an abnormality, thereby realizing a closed-loop real-time response from abnormality identification to active intervention, which solves the problem that the prior art cannot provide real-time active protection by tracing the source after the fact.

[0020] (2) The present invention determines the safe following distance based on the speed of the target vehicle and the vehicle in front and the road adhesion coefficient when the trend of change is decreasing. This allows the safety boundary to be dynamically adjusted with the difference between the road adhesion coefficient and the vehicle speed. Thus, under the premise of ensuring safety, compared with the method of using a fixed safety distance threshold, it avoids unnecessary premature or excessive deceleration to a certain extent.

[0021] (3) By obtaining the speed and following distance of each vehicle within a preset distance range behind, the present invention predicts the changes in speed and following distance in the future period, calculates the rear-end collision risk coefficient and constrains the required deceleration of the target vehicle, and prevents the target vehicle from being rear-ended by the vehicle behind due to sudden deceleration, thus realizing the comprehensive and coordinated management of collision risks in the front and rear directions.

[0022] (4) This invention obtains the lane-changing behavior records and current motion status of vehicles in adjacent lanes, calculates the lane-changing frequency, following distance during lane changing and vehicle cutting probability to determine the lane-changing interference coefficient, and corrects the safe following distance and required deceleration accordingly, so that the target vehicle can actively adjust its safety strategy when facing frequent lane changes or sudden cutting by adjacent vehicles, thus improving the problem of insufficient consideration of lateral lane-changing interference risk in the prior art. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram showing the connections between the steps of the method of the present invention;

[0025] Figure 2This is a schematic diagram showing the connection steps for calculating the rear-end collision risk factor of the present invention;

[0026] Figure 3 This is a schematic diagram showing the connections of the various modules in the system of the present invention. Detailed Implementation

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

[0028] Please see Figure 1 As shown, the present invention provides a method for identifying and processing abnormal vehicle events on highways, which includes the following steps S1 to S5.

[0029] S1. Identify abnormal events and trends in vehicle following.

[0030] This step aims to determine whether the target vehicle is following abnormally in relation to the vehicle in front, and to further analyze the dynamic changes in the following distance in order to determine whether active safety control needs to be activated.

[0031] For example, identifying whether there is an abnormal following event includes: using the ratio of the following distance between the target vehicle and the vehicle in front to the speed of the target vehicle as the headway.

[0032] When the relative speed between the target vehicle and the vehicle in front is greater than 0, the ratio of the following distance to the relative speed is used as the collision time; if the relative speed is less than or equal to 0, the collision time is determined to be infinite, meaning there is no risk of collision.

[0033] The headway and collision time are compared with preset thresholds. When the headway is less than the preset minimum safe headway or the collision time is less than the preset minimum safe collision time, it is identified as an abnormal following event; otherwise, it is identified as no abnormal following event.

[0034] The minimum safe headway can be dynamically adjusted according to the current vehicle speed: at low speeds, the driver has sufficient reaction time, so a smaller value can be used; at high speeds, more reaction margin is needed. For example, 1.5 seconds is used when the vehicle speed is below 60 km / h, 2.0 seconds when it is between 60 and 100 km / h, and 2.5 seconds when it is above 100 km / h. The minimum safe collision time can also be dynamically adjusted according to the current vehicle speed: the higher the speed, the greater the safety margin required when the vehicle in front brakes suddenly. For example, 3 seconds is used below 60 km / h, 4 seconds between 60 and 100 km / h, and 5 seconds above 100 km / h.

[0035] For example, the identification of changing trends includes: constructing a time series sequence of following distances based on the following distances between the target vehicle and the preceding vehicle within a preceding time window.

[0036] The time series is subjected to least-squares linear fitting to obtain the slope of the fitted line. If the slope is less than 0 and its absolute value exceeds a preset slope threshold, the trend is determined to be decreasing. The slope threshold can be obtained through calibration experiments: in a highway scenario, time series data of following distance during normal following traffic are collected, the absolute value of the least-squares fitted slope is calculated, and the 95th percentile of multiple sets of experimental data is taken as the slope threshold.

[0037] If the slope is greater than 0 and exceeds a preset slope threshold, the trend of change is determined to be increasing; otherwise, the trend of change is determined to be stable.

[0038] In a preferred embodiment, the length of the preceding time window is 3 to 10 seconds, the sampling frequency is 5 to 20 Hz, and the slope threshold is 0.1 to 0.5 m / s. For example, in a highway scenario, the window length can be set to 5 seconds, the sampling frequency to 10 Hz, and the slope threshold to 0.2 m / s.

[0039] When the trend is increasing, it indicates that the following distance is gradually increasing and the danger level is easing. At this time, the current speed of the target vehicle is maintained, active speed reduction control is not activated, and step S1 is repeated every preset monitoring cycle (e.g., 0.5 seconds) to continuously monitor the abnormal following status. If the abnormal following event persists for more than the preset first warning time window (e.g., 10 seconds) under the increasing trend, an auditory warning is issued to the driver, informing them that the following distance ahead is abnormal and requesting them to maintain a safe distance, but the vehicle is still not slowed down.

[0040] When the trend is stable, it indicates that the following distance has not changed significantly, but the abnormal following event may persist. To avoid the vehicle being in a dangerous following distance for a long time without active intervention, further judgment is made in the stable state: if the duration of the abnormal following event exceeds the preset second warning time window (e.g., 5 seconds), a mild active safety intervention is triggered, which includes: prompting the driver to increase the following distance; at the same time, the target vehicle is gradually decelerated at a preset mild intervention deceleration until one of the following exit conditions is met: a) the headway recovers to more than or equal to 1.2 times the minimum safe headway; b) the trend changes from stable to decreasing (at this time, step S2 is entered to perform normal active deceleration control).

[0041] During the aforementioned gradual deceleration process, continuous real-time monitoring of the following distance and its trend should be maintained. If the stable state continues but the headway remains below the safety threshold, the minimum deceleration control should be maintained, and braking intensity should not be increased to avoid the risk of rear-end collisions caused by excessive deceleration.

[0042] If the trend changes to decrease during the increasing or stable state, then immediately proceed to step S2 to execute active deceleration control.

[0043] It should be noted that when transitioning from a stable or increasing state to step S2, the deceleration command for the target vehicle should gradually transition from the current value to the corrected deceleration calculated in step S2 according to a preset rate limit, in order to avoid sudden changes in deceleration. The rate limit's upward slope is preset to 0.5 (m / s²) / s based on vehicle comfort requirements. If the corrected deceleration is less than the current command value, it can be updated directly.

[0044] S2. Determine a safe following distance.

[0045] When step S1 determines that the following distance is decreasing, this step calculates the dynamic safe following distance based on the speeds of the target vehicle and the vehicle in front, the road adhesion coefficient, and the preset safe distance. This distance is adjusted in real time according to the road surface slipperiness and speed difference to avoid unreasonable control caused by using a fixed threshold.

[0046] For example, determining the safe following distance includes: acquiring the speeds of the target vehicle and the vehicle in front, the road adhesion coefficient, and a preset safe parking distance. If the relative speed is greater than 0, the safe following distance is calculated. , In the formula For the target vehicle speed, The speed of the vehicle in front. The acceleration due to gravity (approximately) ), The road adhesion coefficient, This is a preset safe parking distance. Preferably, the safe parking distance... It is pre-calibrated and stored as a lookup table through actual vehicle braking tests. The input to the lookup table is the current vehicle speed. For example, when the vehicle speed is 60 km / h... When the distance is 2.5 meters and the vehicle speed is 120 km / h Take 3.5 meters. For specific calibration, on a dry, flat road surface, brake the test vehicle to a complete stop with different initial speeds, measure the remaining distance between the front of the test vehicle and the rear of the vehicle in front after stopping, and take the 90th percentile from multiple experiments as the distance at that speed. value.

[0047] Wherein, the road adhesion coefficient Real-time data is acquired through sensing devices deployed along the highway: First, the road surface type (e.g., asphalt, cement) and weather condition (dry, light rain, moderate rain, heavy rain, snow, ice) of the current road section are identified based on road surface condition cameras or meteorological data interfaces. A baseline value is obtained through a pre-stored road surface type-baseline adhesion coefficient mapping table (e.g., 0.8 for dry asphalt and 0.7 for dry cement). Then, a corresponding correction coefficient is selected based on the weather condition: 1 for dry, 0.8 for light rain (≤2.5 mm / h), 0.6 for moderate rain (2.5~7.5 mm / h), 0.4 for heavy rain (>7.5 mm / h), and 0.2 for snow or ice (snow thickness ≥5 mm or ice thickness ≥1 mm). The baseline value is then multiplied by the correction coefficient to obtain the road adhesion coefficient μ.

[0048] The aforementioned weather correction coefficients are calibrated as follows: On a standard test road, a friction coefficient measuring vehicle is used to measure the longitudinal adhesion coefficient under each working condition at 80 km / h. The dry working condition is taken as the baseline 1, and the ratio of the measured values ​​of the other working conditions to it is the corresponding correction coefficient.

[0049] If the relative speed is less than or equal to 0, meaning the target vehicle's speed is no higher than the vehicle in front, although no additional braking distance is needed, a safe following distance based on the current speed should still be maintained to prepare for the possibility of the vehicle in front suddenly braking. Therefore, .in The preset safe headway ranges from 1.5 seconds to 2.5 seconds.

[0050] S3. Calculate the rear-end collision risk coefficient and determine the required deceleration.

[0051] This step considers both forward safety and the risk of rear-end collisions with vehicles behind, preventing the target vehicle from being hit by a rear-end collision due to excessive deceleration. It quantifies the risk of a rear-end collision by predicting future changes in the following vehicle's speed and following distance, and then adjusts the required deceleration based on a rear-end collision risk coefficient.

[0052] Please see Figure 2 As shown, for example, the calculation of the rear-end collision risk coefficient of the following vehicle includes: S3-1, obtaining the status data of the following vehicle.

[0053] For each vehicle within a preset distance range (e.g., 150 meters) behind the target vehicle, obtain the vehicle speed, following distance, and continuous speed sequence within a preceding time window (e.g., 5 seconds) for each vehicle behind it.

[0054] S3-2, Predicting the future speed of the vehicle behind.

[0055] Based on the vehicle speed sequence, the average relative acceleration of the following vehicle relative to the target vehicle is calculated using least squares linear fitting, and the vehicle speed at each time point in the future period is predicted based on the average relative acceleration.

[0056] Specifically, the fitting calculation process for the average relative acceleration is as follows: First, extract the vehicle speeds of the following vehicle and the target vehicle at each sampling moment from the vehicle speed sequences of the following vehicle and the target vehicle within the preceding time window (e.g., 5 seconds), and calculate the relative speeds of the two at each moment.

[0057] Then, using the time series as the independent variable and the relative velocity at each moment as the dependent variable, a least squares linear fitting algorithm is used to fit a straight line, and the slope of the line is taken as the average relative acceleration of the following vehicle relative to the target vehicle. When the average relative acceleration is positive, it indicates that the following vehicle is accelerating towards the target vehicle; when the average relative acceleration is negative, it indicates that the following vehicle is decelerating away.

[0058] Based on the aforementioned average relative acceleration, the speed of the following vehicle at each time point within the future time period is predicted. The prediction consists of two steps: First, assuming the target vehicle maintains its current constant speed throughout the future time period, the future speed of the target vehicle is predicted; simultaneously, based on the acceleration obtained by fitting the preceding time window of the following vehicle (assuming this acceleration remains constant throughout the future time period), the future speed of the following vehicle is predicted. This yields the predicted speeds of the target vehicle and the following vehicle at each discrete time point within the future time period.

[0059] Then, based on the predicted vehicle speed, the relative speed is calculated, and the predicted following distance is obtained by integrating over time, thereby calculating the initial rear-end collision risk coefficient of the following vehicle.

[0060] Based on the initial rear-end collision risk coefficient, the demand deceleration is further determined. As a preferred implementation, the calculated demand deceleration can be re-introduced into the above prediction process for iterative calculation: the future predicted speed of the target vehicle is updated using the demand deceleration, the rear-end collision risk coefficient is recalculated, and if the difference from the initial value exceeds a preset threshold, the demand deceleration is adjusted again until convergence. If convergence is not achieved after more than 5 iterations, the average of the two most recent demand decelerations is taken as the final value, and an anomaly flag is recorded. Simultaneously, the current demand deceleration is limited to no more than 2.0 m / s², and a system calculation anomaly warning is issued to the target vehicle via the roadside communication unit, requesting a safe following distance. Relevant status data (including the vehicle's and the vehicle in front's speed, following distance, rear vehicle data sequence, and iteration process) is uploaded to the roadside management platform or cloud for subsequent algorithm optimization.

[0061] S3-3, Predicting Future Following Distance

[0062] Based on the predicted speeds and following distances of the target vehicle and the following vehicle, the following distance at each time point in the future period is predicted by integrating the relative speed over time.

[0063] Specifically, predicting the following distance at each moment within the future time period involves using the current following distance between the following vehicle and the target vehicle as an initial value, integrating the predicted relative speeds of the two vehicles over the future time period, and adding the integrated result to the initial value to obtain the predicted following distance at the future moment. If the integration result is negative, it indicates that the following vehicle is approaching, and the following distance decreases.

[0064] S3-4. Extract the minimum following distance and corresponding relative speed.

[0065] Obtain the minimum following distance and the relative speed at the time corresponding to the minimum following distance from the following distance at each time point within the future time period.

[0066] S3-5. Rear-end Collision Risk Assessment

[0067] The minimum following distance is compared with a preset safe following distance threshold, which represents the minimum safe distance that should be maintained between the following vehicle and the target vehicle. If the minimum following distance is less than or equal to the preset safe following distance threshold, it indicates that the following vehicle will enter the dangerous following area within the predicted time period, posing a high risk of rear-end collision. This threshold can be determined according to highway driving safety regulations, with a typical range of 15 to 30 meters. For example, in this embodiment, 20 meters is used; in this case, the rear-end collision risk coefficient of the following vehicle is recorded as 1.

[0068] S3-6, Calculation of the product of rear-end collision risk coefficients

[0069] If the minimum following distance is greater than the preset rear safe following distance threshold, the ratio of the relative speed to the following vehicle's speed is used as the speed risk factor. If the relative speed is less than or equal to 0, the speed risk factor is set to 0. The ratio of the rear safe following distance threshold to the minimum following distance is used as the distance risk factor. The product of the speed risk factor and the distance risk factor is then used as the rear-end collision risk coefficient of the following vehicle.

[0070] For example, determining the required deceleration includes: calculating the basic deceleration required to avoid rear-ending the vehicle in front based on the target vehicle speed, the speed of the vehicle in front, the following distance, and the safe following distance.

[0071] Specifically, base deceleration Calculated using the kinematic formulas for uniform deceleration: In the formula The physical meaning of this formula for following distance is: when the target vehicle brakes with uniform deceleration, at a relative distance... The speed of the car will be reduced from Down to The minimum deceleration required. It should be noted that when... When the denominator is not positive, this formula is not applicable. In this case, the basic deceleration is directly taken as the maximum braking deceleration that the vehicle can achieve (e.g., However, this method usually already satisfies the requirements when an abnormal following event is triggered. Otherwise, the system will immediately perform emergency braking.

[0072] When the rear-end collision risk coefficient is less than 1, the value 1 is subtracted from the rear-end collision risk coefficient to obtain the reduction coefficient. Then, the reduction coefficient is multiplied with the base deceleration to calculate the required deceleration.

[0073] When the rear-end collision risk coefficient is equal to 1, the base deceleration is compared with the preset rear-end risk safety deceleration threshold, and the smaller value is taken as the required deceleration.

[0074] The rear-end risk safety deceleration threshold is used to limit the deceleration of the target vehicle when the risk of a rear-end collision is high, in order to avoid being rear-ended by the following vehicle due to sudden deceleration. This threshold can be calibrated based on factors such as the relative speed between the target vehicle and the following vehicle, and the braking ability of the following vehicle, with a value range of 1.5 to 3.0 m / s². In this embodiment, it is set to 2.0 m / s².

[0075] When the risk of a rear-end collision reaches its peak, priority must be given to avoiding a rear-end collision caused by the vehicle's sudden deceleration. Therefore, the required deceleration is limited to below a low safety threshold (e.g., 2.0 m / s²). At the same time, braking cannot be completely abandoned to prevent the risk of following the vehicle ahead from worsening and escalating into a collision. Therefore, by taking the smaller value between the base deceleration and the safety threshold, a compromise is achieved that ensures the vehicle does not decelerate excessively while retaining forward avoidance capabilities, thus achieving a balance between front and rear risks.

[0076] S4. Determine the lane change interference coefficient and calculate the corrected safety distance and corrected deceleration.

[0077] This step quantifies the impact of adjacent lane vehicle lane-changing behavior on the target vehicle's safety. A basic disturbance coefficient is obtained by statistically analyzing historical lane-changing events, and then combined with the current vehicle's cut-in probability to obtain a dynamic disturbance coefficient. Finally, a lane-changing disturbance coefficient is generated, which is used to correct the safe following distance and required deceleration.

[0078] For example, determining the lane change interference coefficient includes: obtaining from the lane change behavior record the lane change occurrence time and lane change start time of each lane change event in the adjacent lane of the target vehicle within the preceding time window, and the following distance between the lane change vehicle and the target vehicle.

[0079] Based on the following distance between the lane-changing vehicle and the target vehicle at the start of the lane change, a risk weight is assigned to each lane change according to a preset mapping relationship, where the smaller the following distance, the higher the weight is assigned.

[0080] The above-mentioned pre-defined mapping relationship is obtained by collecting traffic accidents or dangerous events caused by lane changes in highway scenarios, statistically analyzing the correlation between following distance at the start of lane changes and accident incidence, and obtaining weight curves after normalizing the incidence rates. For example, based on 1000 valid samples, the relative frequency of danger is calculated to be 80% when the following distance is less than or equal to 10 meters, 40% when the following distance is between 10 and 20 meters, and 15% when the following distance is greater than 20 meters. The weights are then set to 0.8, 0.4, and 0.15, respectively.

[0081] The number of lane changes within the preceding time window is counted. The total number of lane changes is divided by the length of the preceding time window to obtain the lane change frequency. The risk weights of each lane change are summed and then divided by the total number of lane changes to obtain the average risk weight. Finally, the lane change frequency is multiplied by the average risk weight to obtain the basic interference coefficient.

[0082] In one specific embodiment, the preceding time window length is 60 seconds. If three lane change events are detected within the window, the lane change frequency is calculated. , If the risk weights for each lane change are 0.5, 1.0, and 0.5 respectively, then the average risk weight... , Basic interference coefficient , The basic interference coefficient reflects the combined impact of the frequency and danger of historical lane changes.

[0083] The lateral velocity, longitudinal distance, and relative velocity of each adjacent vehicle relative to the target lane are obtained from the motion state. The probability of each adjacent vehicle cutting into the target lane in the future time period is calculated, and then the maximum probability of all adjacent vehicles is taken as the dynamic disturbance coefficient.

[0084] The basic interference coefficient and the dynamic disturbance coefficient are summed and limited to the range of 0 to 1 to obtain the lane change interference coefficient. Specifically, when the dynamic disturbance coefficient exceeds a preset high-risk threshold (e.g., 0.8), the lane change interference coefficient is directly assigned a value of 1 to unconditionally activate the most conservative safety strategy. It should be noted that when the sum of the basic interference coefficient and the dynamic disturbance coefficient exceeds 1, it indicates that the lateral intrusion risk has reached an extremely high level. At this point, the system no longer needs to further distinguish subtle differences; setting the lane change interference coefficient to 1 will trigger the most conservative safety strategy (i.e., maximizing the safe distance and minimizing deceleration).

[0085] It should be noted that the dynamic disturbance coefficient is used to quantify the subjective probability and urgency of vehicles in adjacent lanes cutting into this lane at the current moment. For each vehicle in an adjacent lane, the following three factors are calculated, and their product is taken as the vehicle's cutting probability; the maximum value of all vehicle cutting probabilities is taken as the dynamic disturbance coefficient. In a specific embodiment, the calculation process is as follows.

[0086] (1) Lateral velocity factor

[0087] Based on the vehicle's lateral speed (positive in the direction of the target lane). When the lateral speed is less than or equal to the first preset threshold (e.g., 0.1 m / s), it indicates no obvious intention to change lanes, and the lateral speed factor is set to 0. When the lateral speed is greater than or equal to the second preset threshold (e.g., 0.6 m / s), it indicates a rapid lane change, and the lateral speed factor is set to 1. Intermediate values ​​are calculated using linear interpolation. This piecewise linear mapping is used to avoid false triggers caused by normal minor adjustments made by the vehicle within the lane.

[0088] (2) Vertical distance factor

[0089] Based on the longitudinal distance between the vehicle and the target vehicle, an exponential decay model or equivalent segmented values ​​are used: when the longitudinal distance is less than or equal to 5 meters, the longitudinal distance factor is set to 1; when the longitudinal distance is greater than or equal to 30 meters, the longitudinal distance factor is set to 0; the intermediate distances decrease inversely proportionally or exponentially. The longitudinal distance factor reflects the accelerating growth of the risk of entry as the distance increases.

[0090] (3) Relative velocity factor

[0091] Let the relative velocity be denoted as... ,when When the relative speed factor is 1, it indicates that the adjacent vehicle is significantly faster and approaching from behind, with a strong motivation to change lanes; when the relative speed factor is 1. When the target vehicle is significantly faster and adjacent vehicles have no intention of changing lanes, the relative speed factor is taken as 0.3; when At that time, the relative velocity factor is calculated using linear interpolation.

[0092] Multiply the above three factors to obtain the cut-in probability of each adjacent vehicle. Iterate through all adjacent vehicles in all adjacent lanes and take the largest cut-in probability as the dynamic disturbance coefficient. If there are no vehicles in the adjacent lane, the dynamic disturbance coefficient is 0.

[0093] All the preset parameters mentioned above (such as speed threshold, distance threshold, and interpolation method) can be adjusted according to the actual vehicle calibration or system design requirements.

[0094] For example, the calculation of the corrected safe distance and the corrected deceleration includes: multiplying the safe following distance by (1 plus the lane change interference coefficient) to obtain the corrected safe distance.

[0095] Multiply the required deceleration by (1 minus the lane change interference coefficient) to obtain the preliminary corrected deceleration. Then compare the preliminary corrected deceleration with the preset minimum allowable deceleration and take the larger value as the corrected deceleration.

[0096] The preset minimum permissible deceleration represents the lowest deceleration value allowed for the target vehicle when performing deceleration control. When the lane change interference coefficient is high (i.e., the risk of lateral cut-in is high), the value of 1 minus the lane change interference coefficient may approach 0 or even be negative, resulting in an excessively small or zero initial correction deceleration. By setting the minimum permissible deceleration, it is ensured that the target vehicle can maintain effective braking control under any circumstances to ensure forward safety. The calibration of the minimum permissible deceleration should comprehensively consider two aspects: first, the lower limit of the driver's perceived comfort of deceleration (generally considered to be almost imperceptible below 0.5 m / s², but weak braking force may lead to a continuously too close following distance); second, the minimum deceleration required for the vehicle to overcome the component of gravity on common slopes (such as a maximum slope of 3% on highways), for example, in this embodiment, is taken as... .

[0097] It should be noted that the corrected deceleration, as the final control command of the target vehicle, will also be fed back to the rear vehicle risk prediction in step S3 to update the future predicted speed of the target vehicle and recalculate the rear-end collision risk coefficient to ensure the self-consistency of the front and rear risks.

[0098] S5. Control the target vehicle to reduce its speed until a safe distance is maintained.

[0099] The target vehicle actively reduces its speed using the corrected deceleration calculated in step S4. During the deceleration process, the following distance between the target vehicle and the vehicle in front, as well as various dynamic parameters (vehicle speed, road adhesion coefficient, following vehicle status, lane change interference coefficient, etc.), are updated in real time, and the corrected safe distance (i.e., the real-time updated corrected safe distance) is recalculated. The current following distance is continuously compared with the corrected safe distance until the following distance is greater than or equal to the real-time updated corrected safe distance, at which point the deceleration control is exited, and normal driving resumes.

[0100] If the deceleration control duration exceeds the preset limit (e.g., 30 seconds), or the target vehicle speed is lower than the minimum safe speed threshold (e.g., 40 km / h), the deceleration control will be forcibly disengaged and a warning will be issued.

[0101] After exiting the speed reduction control, the system returns to step S1 and continues to monitor the following status at a preset monitoring cycle (e.g., 0.5 seconds). If an abnormal following event is detected again and the trend is decreasing or stable (exceeding the duration threshold), steps S2 to S5 are executed again to form a continuous safety monitoring closed loop.

[0102] The present invention also provides a highway vehicle abnormal event identification and processing device, which includes a processor and a memory. The memory stores a computer program, and the processor executes the computer program to implement a highway vehicle abnormal event identification and processing method.

[0103] Please see Figure 3 As shown, the present invention also provides a highway vehicle abnormal event identification and processing system, which includes: an abnormal trend identification module, a safe distance determination module, a following vehicle risk calculation module, a lane change interference correction module, and a speed reduction control terminal.

[0104] In the above, the safe distance determination module is connected to the abnormal trend identification module and the following vehicle risk calculation module, respectively, and the lane change interference correction module is also connected to the following vehicle risk calculation module and the speed reduction control terminal, respectively.

[0105] The abnormal trend identification module identifies whether there are any abnormal following events based on the speed and following distance of the target vehicle and the vehicle in front. If so, it identifies the trend of change based on the following distance within the preceding time window.

[0106] The safe following distance determination module determines the safe following distance based on the speed of the target vehicle and the vehicle in front, as well as the road adhesion coefficient, when the trend of change is decreasing.

[0107] The rear vehicle risk calculation module obtains the speed and following distance of each vehicle within a preset distance range in the preceding time window, predicts the speed and following distance changes of each vehicle in the future period, calculates the rear-end collision risk coefficient of the rear vehicle, and determines the required deceleration in combination with the safe following distance.

[0108] The lane change interference correction module acquires the lane change behavior records and current motion status of vehicles in adjacent lanes within the preceding time window, determines the lane change interference coefficient accordingly, and calculates the corrected safety distance and corrected deceleration based on the lane change interference coefficient.

[0109] The speed reduction control terminal controls the target vehicle to reduce its speed by a corrected deceleration until the following distance is greater than or equal to the corrected safe distance updated in real time.

[0110] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0111] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0112] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0113] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0114] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying and processing abnormal vehicle events on highways, characterized in that: The method includes: The system identifies whether there are any abnormal following events based on the speed and following distance of the target vehicle and the vehicle in front. If so, it identifies the trend of change in following distance within the preceding time window. When the trend of change is decreasing, the safe following distance is determined based on the speed of the target vehicle and the vehicle in front and the road adhesion coefficient. The system obtains the speed and following distance of each vehicle within a preset distance range in the preceding time window, predicts the changes in speed and following distance of each vehicle in the future period, calculates the rear-end collision risk coefficient of the following vehicle, and determines the required deceleration in combination with the safe following distance. The system acquires records of lane-changing behavior and current motion status of vehicles in adjacent lanes within a preceding time window, determines the lane-changing interference coefficient, and calculates the corrected safe distance and corrected deceleration based on the lane-changing interference coefficient. Control the target vehicle to reduce its speed by a corrected deceleration until the following distance is greater than or equal to the real-time updated corrected safe distance.

2. The method for identifying and processing abnormal vehicle events on highways according to claim 1, characterized in that: The identification of whether an abnormal following event exists includes: The ratio of the following distance between the target vehicle and the vehicle in front to the speed of the target vehicle is used as the headway. The speed difference between the target vehicle and the vehicle in front is taken as the relative speed, and the ratio of the following distance to the relative speed is taken as the collision time. The headway and collision time are compared with preset thresholds. When the headway is less than the preset minimum safe headway or the collision time is less than the preset minimum safe collision time, it is identified as an abnormal following event; otherwise, it is identified as no abnormal following event.

3. The method of claim 1, wherein: The identified trends include: Based on the following distance between the target vehicle and the preceding vehicle within the preceding time window, a time series sequence of following distance is constructed. The time series is subjected to least squares linear fitting to obtain the slope of the fitted line. If the slope is less than 0 and its absolute value exceeds a preset slope threshold, the trend of change is determined to be decreasing. If the slope is greater than 0 and exceeds a preset slope threshold, the trend of change is determined to be increasing; otherwise, the trend of change is determined to be stable.

4. The method of claim 1, wherein: Determining a safe following distance includes: Obtain the speed of the target vehicle and the vehicle in front, the road adhesion coefficient, and the preset safe parking distance; If the relative speed is greater than 0, calculate the safe following distance. , In the formula For the target vehicle speed, The speed of the vehicle in front. It is the acceleration due to gravity. The road adhesion coefficient, This is the preset safe parking distance; If the relative speed is less than or equal to 0, the dynamic safe distance is calculated based on the target vehicle speed and the preset safe headway, and the sum of the dynamic safe distance and the parking safe distance is taken as the safe following distance.

5. The method for identifying and processing abnormal vehicle events on highways according to claim 1, characterized in that: The calculation of the rear-end collision risk coefficient includes: Obtain the speed, following distance, and continuous speed sequence of each following vehicle within the preceding time window; Based on the vehicle speed sequence, the average relative acceleration of the following vehicle relative to the target vehicle is calculated using least squares linear fitting, and the vehicle speed at each time point in the future period is predicted based on the average relative acceleration. Based on the predicted speeds and following distances of the target vehicle and the following vehicle, the following distances at various times in the future period are predicted by integrating the relative speed over time. Obtain the minimum following distance and the relative speed at the time corresponding to the minimum following distance from the following distance at each time point within the future time period; The minimum following distance is compared with the preset rear safe following distance threshold. If the minimum following distance is less than or equal to the preset rear safe following distance threshold, the rear-end collision risk coefficient of the following vehicle is recorded as 1. If the minimum following distance is greater than the preset safe following distance threshold, the ratio of the relative speed to the following vehicle's speed will be used as the speed risk factor, and the ratio of the safe following distance threshold to the minimum following distance will be used as the distance risk factor. The product of the speed risk factor and the distance risk factor will then be used as the rear-end collision risk coefficient of the following vehicle.

6. The method of claim 5, wherein: The determination of demand reduction includes: Calculate the basic deceleration required to avoid rear-ending the vehicle in front, based on the target vehicle's speed, the speed of the vehicle in front, the following distance, and the safe following distance. When the rear-end collision risk coefficient is less than 1, the value 1 is subtracted from the rear-end collision risk coefficient to obtain the reduction coefficient. Then, the reduction coefficient is multiplied with the base deceleration to calculate the required deceleration. When the rear-end collision risk coefficient is equal to 1, the base deceleration is compared with the preset rear-end risk safety deceleration threshold, and the smaller value is taken as the required deceleration.

7. The method of claim 1, wherein: The determination of the lane change interference coefficient includes: From the lane change behavior records, obtain the time of lane change occurrence and the following distance between the lane-changing vehicle and the target vehicle at the start time of each lane change event that has completed a lane change towards the target lane in the adjacent lane of the target vehicle within the preceding time window. Based on the following distance between the lane-changing vehicle and the target vehicle at the start of the lane change, a risk weight is assigned to each lane change according to a preset mapping relationship. The number of lane changes within the preceding time window is counted. The total number of lane changes is divided by the length of the preceding time window to obtain the lane change frequency. The risk weights of each lane change are summed and then divided by the total number of lane changes to obtain the average risk weight. Finally, the lane change frequency is multiplied by the average risk weight to obtain the basic interference coefficient. The lateral velocity, longitudinal distance, and relative velocity of each adjacent vehicle relative to the target lane are obtained from the motion state. The probability of each adjacent vehicle cutting into the target lane in the future time period is calculated, and then the maximum probability of all adjacent vehicles is taken as the dynamic disturbance coefficient. The basic interference coefficient and the dynamic disturbance coefficient are summed and then limited to the range of 0 to 1 to obtain the lane change interference coefficient.

8. The method of claim 6, wherein: The calculation of the corrected safety distance and corrected deceleration includes: Multiply the safe following distance by (1 plus the lane change interference factor) to obtain the corrected safe distance; Multiply the required deceleration by (1 minus the lane change interference coefficient) to obtain the preliminary corrected deceleration. Then compare the preliminary corrected deceleration with the preset minimum allowable deceleration and take the larger value as the corrected deceleration.

9. An expressway vehicle abnormal event recognition processing device characterized by comprising: The processing device includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1-8.

10. A highway vehicle abnormal event recognition processing system characterized by comprising: The system includes: The abnormal trend recognition module identifies whether there are any abnormal following events based on the speed and following distance of the target vehicle and the vehicle in front. If so, it identifies the trend of change based on the following distance within the preceding time window. The safe following distance determination module determines the safe following distance based on the speed of the target vehicle and the vehicle in front, as well as the road adhesion coefficient, when the trend of change is decreasing. The rear vehicle risk calculation module obtains the speed and following distance of each vehicle within a preset distance range in the preceding time window, predicts the speed and following distance changes of each vehicle in the future period, calculates the rear-end collision risk coefficient of the rear vehicle, and determines the required deceleration in combination with the safe following distance. The lane change interference correction module acquires the lane change behavior records and current motion status of vehicles in adjacent lanes within the preceding time window, determines the lane change interference coefficient accordingly, and calculates the corrected safety distance and corrected deceleration based on the lane change interference coefficient. The speed reduction control terminal controls the target vehicle to reduce its speed by a corrected deceleration until the following distance is greater than or equal to the corrected safe distance updated in real time.

Images