A vehicle collision warning method and electronic device in a mixed traffic environment

Abstract

Embodiments of the present application disclose a vehicle collision warning method and an electronic device in a mixed driving environment. The method comprises: obtaining human-driven vehicle trajectory information and autonomous vehicle trajectory information; determining first predicted motion state information according to the human-driven vehicle trajectory information, and determining second predicted motion state information according to the autonomous vehicle trajectory information; determining a motion collision index of the human-driven vehicle and the autonomous vehicle and a degree of advantage of a vehicle that reaches a predicted collision point first according to the first predicted motion state information and the second predicted motion state information; determining a collision probability of the human-driven vehicle and the autonomous vehicle according to the KCI and the degree of advantage; and performing different forms of warning on the human-driven vehicle and the autonomous vehicle according to the collision probability. The method can improve the accuracy of the warning and the user acceptance.

Description

Technical Field

[0001] This application relates to the field of road traffic safety, and in particular to a vehicle collision warning method and electronic equipment in mixed traffic environments. Background Technology

[0002] Vehicle collision warning systems are a key component of active safety in intelligent driving. Current vehicle collision warning systems primarily assess risk based on physical kinematic indicators such as time to collision (TTC), triggering a warning when the TTC falls below a set threshold.

[0003] However, this type of method has significant drawbacks in practical applications. For example, it ignores individual driver differences, leading to a high false alarm rate and poor model universality, and it is difficult to balance geometric modeling with real-time performance. Therefore, there is an urgent need for a vehicle collision warning method that integrates individual driving differences and possesses efficient geometric computation capabilities. Summary of the Invention

[0004] This application addresses the problems mentioned in the background art by providing a vehicle collision warning method and electronic device. This method quantifies the strength of a driver's intention to travel in a game-theoretic scenario by constructing a Kinetic Collision Index (KCI) and an Advantage Factor (AF), significantly reducing the false alarm rate caused by traditional single indicators. Simultaneously, it employs a model that integrates driving preference coefficients to proactively adapt to the differences in risk tolerance between aggressive and conservative drivers, solving the problem of poor model universality and improving the accuracy of warnings in complex interactive environments.

[0005] In a first aspect, a vehicle collision warning method for a mixed-traffic environment is provided, comprising: acquiring trajectory information of a human-driven vehicle and an autonomous vehicle during a first time period; determining, based on the trajectory information of the human-driven vehicle, a first predicted motion state of the human-driven vehicle during a second time period, and determining, based on the trajectory information of the autonomous vehicle, a second predicted motion state of the autonomous vehicle during the second time period, wherein the second time period follows the first time period; determining, based on the first and second predicted motion state information, a motion collision index (KCI) for the human-driven vehicle and the autonomous vehicle, and determining the degree of priority of the vehicle that arrives first at the predicted collision point, wherein the degree of priority is used to indicate the strength of the intention of the vehicle that arrives first at the predicted collision point to have priority passage; determining the collision probability of the human-driven vehicle and the autonomous vehicle based on the KCI and the degree of priority; and providing different forms of warnings to the human-driven vehicle and the autonomous vehicle based on the collision probability.

[0006] This scheme can effectively quantify psychological safety boundaries and game steady state, identify different game scenarios, and significantly improve the accuracy of early warning in complex interactive environments. At the same time, by integrating the degree of prior advantage and KCI, it can reduce false alarm rate and improve user experience.

[0007] In conjunction with the first aspect, in a possible implementation of the first aspect, determining the first predicted motion state information of the human-driven vehicle in the second time period based on the trajectory information of the human-driven vehicle includes: inputting the trajectory information of the human-driven vehicle and the driving style characteristics of the human-driven vehicle into a time-series prediction model to obtain the first predicted motion state information output by the time-series prediction model, wherein the driving style characteristics of the human-driven vehicle are extracted based on at least one of the following factors: historical driving behavior data of the human-driven vehicle, acceleration, steering angle change, and vehicle speed fluctuation.

[0008] This approach utilizes a time-series prediction model to uniformly process trajectory data of human-driven vehicles, ensuring consistency and computational efficiency in motion state prediction, and providing reliable input for subsequent collision risk assessment.

[0009] In conjunction with the first aspect, in a possible implementation of the first aspect, determining the KCI of the human-driven vehicle and the autonomous vehicle based on the first predicted motion state information and the second predicted motion state information includes: determining, based on the first predicted motion state information and the second predicted motion state information, that the human-driven vehicle and the autonomous vehicle have a potential collision risk; and determining the KCI based on the potential collision risk of the human-driven vehicle and the autonomous vehicle; wherein the KCI is determined according to the following formula:

[0010]

[0011] In the formula, This is the location of the vehicle being driven by the human. This is the location of the autonomous vehicle. It is the speed at which the human is driving the vehicle. It is the speed of the self-driving vehicle. It is the acceleration of the human driving the vehicle. It is the acceleration of the autonomous vehicle. It is the acceleration weight.

[0012] This solution accurately identifies high-risk scenarios, avoids missed reports, and enhances the reliability of the early warning system through potential collision risk verification.

[0013] In conjunction with the first aspect, in a possible implementation of the first aspect, determining the potential collision risk between the human-driven vehicle and the autonomous vehicle based on the first predicted motion state information and the second predicted motion state information includes: based on the first predicted motion state information and the second predicted motion state information, and based on the vehicle's circumcircle model, performing distance screening on the predicted positions of the human-driven vehicle and the autonomous vehicle, eliminating time steps without conflict risk, to obtain the preliminary predicted states of the human-driven vehicle and the autonomous vehicle; based on the preliminary predicted states of the human-driven vehicle and the autonomous vehicle and the vehicle's dual-circle model, determining the Euclidean distance between the front and rear center centers of the human-driven vehicle and the autonomous vehicle; and based on the Euclidean distance between the front and rear center centers of the human-driven vehicle and the autonomous vehicle and a preset distance threshold, determining the potential collision risk between the human-driven vehicle and the autonomous vehicle.

[0014] This approach reduces computational burden and improves system real-time performance by using efficient geometric modeling methods, while ensuring accurate fitting of the vehicle body contour to reduce the risk of misjudgment.

[0015] In conjunction with the first aspect, in a possible implementation of the first aspect, the degree of advance advantage is the ratio of the length of the collision point already passed by the vehicle arriving at the collision point first to the first equivalent perceived length. The first equivalent perceived length is determined based on the physical vehicle length and dynamic safety margin of the vehicle arriving at the predicted collision point first in the human-driven vehicle and the autonomous vehicle. The dynamic safety margin is an increasing function of velocity and acceleration, obtained by the following formula:

[0016]

[0017]

[0018] In the formula, The degree of first-mover advantage of the vehicle that arrives at the potential collision point first; The degree of priority for vehicles arriving at the point later; The length that the vehicle that arrives at the potential collision point first has already traversed from that potential collision point. This refers to the total length of the vehicle body.

[0019] This scheme utilizes dynamic safety margins to simulate the nonlinear perception characteristics of human drivers regarding risk, making the degree of advance advantage more consistent with the logic of real driving decisions and improving the accuracy of game intent representation.

[0020] In conjunction with the first aspect, in a possible implementation of the first aspect, determining the collision probability between the human-driven vehicle and the autonomous vehicle based on the KCI and the degree of prior advantage includes: obtaining the driving preference coefficient of the driver of the human-driven vehicle. The driving preference coefficient is used to quantify the driver's perceived tolerance to spatiotemporal risks. Based on the driving preference coefficient, the KCI, and the degree of precedence advantage, the collision probability between the human-driven vehicle and the autonomous vehicle is determined. This collision probability is calculated using the following formula:

[0021]

[0022] In the formula, the random intercept term Used to characterize unobserved heterogeneity between different vehicle pairs , representing the random intercept term between different vehicle pairs, The following formula is used for calculation:

[0023] When the degree of leading advantage is equal to the degree of leading advantage of the autonomous vehicle,

[0024]

[0025] When the degree of leading advantage is equal to the degree of leading advantage of the human-driven vehicle,

[0026]

[0027] In the formula, AF represents the degree of prior advantage. For the intercept term, , , These are the regression coefficients of KCI, prior advantage degree, and their interaction term, respectively.

[0028] This scheme uses a driving preference coefficient to dynamically adjust the intensity of risk response, reducing false alarms for aggressive drivers and amplifying warnings for conservative drivers, thus eliminating the impact of individual differences on the model's universality.

[0029] In conjunction with the first aspect, in possible implementations of the first aspect, different forms of warnings are issued to the human-driven vehicle and the autonomous vehicle based on the collision probability, including: determining that the collision probability of the human-driven vehicle is greater than a first threshold and less than a second threshold, and that it has no advantage in passing, and thus issuing a warning; or determining that the collision probability of the human-driven vehicle is greater than the second threshold, and that it has no advantage in passing, and thus forcibly braking and decelerating; wherein, the first threshold... Second threshold , This serves as the first threshold based on the autonomous vehicle's operating system and also as the first warning threshold for autonomous vehicles. The system is based on a second threshold, which is also a second warning threshold for autonomous vehicles; and determines that the collision probability of the autonomous vehicle is greater than the first warning threshold and less than the second warning threshold, and that it has no advantage in passing, and then issues an alarm to remind surrounding vehicles to pay attention to the distance; determines that the collision probability of the autonomous vehicle is greater than the second warning threshold, and that it has no advantage in passing, and then forcibly brakes to decelerate.

[0030] This solution implements tiered warnings based on collision probability and game state, avoiding excessive interference with drivers and significantly improving user acceptance.

[0031] In a second aspect, an electronic device is provided, comprising: one or more processors; one or more memories; the one or more memories storing one or more computer programs, the one or more computer programs including instructions that, when executed by the one or more processors, cause the method of the first aspect or any possible implementation of the first aspect to be performed.

[0032] Thirdly, a computer-readable storage medium is provided that stores computer instructions that, when executed on a computer, cause the method of the first aspect or any possible implementation thereof to be performed.

[0033] Fourthly, a computer program product is provided that, when the computer program product is run on a computer, causes the computer to perform the method of the first aspect or any possible implementation of the first aspect. Attached Figure Description

[0034] Figure 1 A schematic flowchart of the vehicle collision warning method provided in an embodiment of this application is shown;

[0035] Figure 2 The system architecture diagram of the vehicle collision warning system provided in the embodiment of this application is shown;

[0036] Figure 3 A schematic diagram of the double-circle model provided in an embodiment of this application is shown;

[0037] Figure 4 A schematic diagram of a collision warning in a scenario with no signalized intersections provided in an embodiment of this application is shown;

[0038] Figure 5 This is a structural schematic diagram of a device provided in an embodiment of this application;

[0039] Figure 6 This is a structural schematic diagram of a system on a chip (SoC) provided in an embodiment of this application. Detailed Implementation

[0040] The technical solution of this application is described below with reference to the accompanying drawings.

[0041] As mentioned in the background section, current vehicle collision warning systems primarily rely on physical kinematic indicators such as TTC (Traffic Tactical Compatibility) for collision risk assessment. This method ignores individual driver differences, leading to a high false alarm rate and poor model universality. Furthermore, it is difficult to balance geometric modeling with real-time performance.

[0042] To address the above technical issues, this application proposes a vehicle collision warning method 100 for mixed traffic environments. Figure 1 A schematic flowchart of a vehicle collision warning method 100 provided in an embodiment of this application is shown. Figure 1 As shown, method 100 includes steps S110 to S150. In method 100, by constructing KCI and prior advantage degree, the strength of the driver's intention to travel in a game scenario is quantified, significantly reducing the false alarm rate caused by the traditional single TTC indicator. At the same time, a model that integrates driving preference coefficients is adopted to actively adapt to the differences in risk tolerance between aggressive and conservative drivers, solving the problem of poor model universality and improving the accuracy of warnings in complex interactive environments.

[0043] Step S110: Obtain the trajectory information of the human-driven vehicle in the first time period and the trajectory information of the autonomous vehicle in the first time period.

[0044] For example, the trajectory information of both human-driven and autonomous vehicles may include motion state data such as position coordinates, speed, acceleration, and heading angle. This data can be collected in real time using onboard sensors and V2V communication modules.

[0045] Step S120: Based on the trajectory information of the human-driven vehicle, determine the first predicted motion state information of the human-driven vehicle in the second time period, and based on the trajectory information of the autonomous vehicle, determine the second predicted motion state information of the autonomous vehicle in the second time period.

[0046] The second time period is after the first time period.

[0047] Optionally, in embodiments of this application, the trajectory data of human-driven vehicles can be uniformly processed using a time-series prediction model to obtain predicted motion state information, ensuring the consistency of motion state prediction. Specifically, in step S120, the trajectory information of the human-driven vehicle and the driving style characteristics of the human-driven vehicle can be input into the time-series prediction model to obtain the first predicted motion state information output by the time-series prediction model.

[0048] In the embodiments of this application, the time-series prediction model can be constructed based on a Long Short-Term Memory (LSTM) network or a Kalman filter algorithm. Driving style characteristics can be parameters that quantify the driver's behavioral tendencies, such as aggressive or conservative coefficients extracted based on historical driving behavior data.

[0049] This approach utilizes a time-series prediction model to uniformly process the trajectory data of two vehicles, ensuring consistency in motion state prediction and computational efficiency, and providing reliable input for subsequent collision risk assessment.

[0050] Step S130: Based on the first predicted motion state information and the second predicted motion state information, determine the KCI of the human-driven vehicle and the autonomous vehicle, and determine the degree of priority advantage of the vehicle that arrives at the predicted collision point first among the human-driven vehicle and the autonomous vehicle.

[0051] KCI is used to quantify collision urgency, and the degree of priority advantage is used to indicate the strength of the intention of the vehicle that arrives at the predicted collision point first to pass.

[0052] Optionally, in step S130, based on the first predicted motion state information and the second predicted motion state information, it is determined that there is a potential collision risk between the first vehicle and the second vehicle; based on the potential collision risk between the first vehicle and the second vehicle, the KCI is determined. The following embodiments will describe this process in detail, and it will not be repeated here.

[0053] For example, the process of determining potential collision risk includes: performing distance screening on the predicted positions of the first vehicle and the second vehicle based on the vehicle's circumcircle model, eliminating time steps without collision risk; calculating the Euclidean distance between the combined centers of the front and rear circles of the two vehicles based on the predicted state after initial screening and the vehicle's dual-circle model; if any center distance is less than a set threshold, a potential collision risk is determined to exist. The following embodiments will describe this process in detail, and will not be repeated here.

[0054] Optionally, in step S130, the degree of advance advantage can be defined as the ratio of the length of the vehicle passing the collision point first to the first equivalent sensing length, the first equivalent sensing length being determined based on the physical length of the vehicle passing the collision point first and the dynamic safety margin, the dynamic safety margin being an increasing function of speed and acceleration.

[0055] By introducing a dynamic margin, the calculation of the leading advantage degree is made more in line with the risk sensitivity of human drivers in high-speed scenarios. The following examples will describe this process in detail, and will not be repeated here.

[0056] Step S140: Determine the collision probability between the human-driven vehicle and the autonomous vehicle based on the KCI and the degree of precedence advantage.

[0057] Optionally, in step S140, the process of determining the collision probability can incorporate a driving preference coefficient as an adjustment factor to accommodate differences in risk perception among different drivers. Specifically, this process includes: obtaining the driving preference coefficient of the driver of the human-driven vehicle; and determining the collision probability of the first vehicle and the second vehicle based on the driving preference coefficient, the KCI, and the degree of precedence advantage. This process will be described in detail below.

[0058] Step S150: Based on the collision probability, issue different forms of warnings to the human-driven vehicle and the autonomous vehicle.

[0059] For example, in an embodiment of this application, the warning response can employ a tiered triggering strategy, combining the game steady-state verification results to output differentiated instructions. Specifically, the process includes: issuing different forms of warnings to the human-driven vehicle and the autonomous vehicle based on the collision probability, including determining that the collision probability of the human-driven vehicle is greater than a first threshold and less than a second threshold, and that it has no advantage in passing, and thus issuing a warning; or determining that the collision probability of the human-driven vehicle is greater than the second threshold, and that it has no advantage in passing, and thus forcibly braking and decelerating; wherein, the first threshold can be determined based on the first warning threshold of the autonomous vehicle, and the second threshold can be determined based on the second warning threshold of the autonomous vehicle; and determining that the collision probability of the autonomous vehicle is greater than the first warning threshold and less than the second warning threshold, and that it has no advantage in passing, and thus issuing an alarm to remind surrounding vehicles to pay attention to the distance; determining that the collision probability of the autonomous vehicle is greater than the second warning threshold, and that it has no advantage in passing, and thus forcibly braking and decelerating. The process will be described in detail below.

[0060] This solution implements tiered warnings based on collision probability and game state, avoiding excessive interference with drivers and significantly improving user acceptance.

[0061] This application reduces the false alarm rate by constructing a game-theoretic risk assessment model by integrating the prior advantage index with the KCI (Knowledge, Indication, and Compatibility). Furthermore, it reduces the calculation error of the prior advantage index by simulating the nonlinear characteristics of human risk perception through dynamic safety margin. In addition, this application lowers the false alarm rate for collision warnings for drivers with different driving preferences by adjusting the model's fixed effects using a driving preference coefficient.

[0062] The following is an introduction Figure 1 Detailed embodiments of each step shown.

[0063] Figure 2 A system architecture diagram of the vehicle collision warning system provided in an embodiment of this application is shown. Figure 2 As shown, the vehicle collision warning system includes a data perception and vehicle-to-vehicle (V2V) unit, a trajectory prediction unit, a spatiotemporal index calculation unit, a collision probability assessment unit, and a graded warning unit.

[0064] The data sensing and V2V unit can be used to collect real-time motion status data of the vehicle and surrounding traffic participants through onboard sensors and V2V communication modules, such as position coordinates, speed, acceleration and heading angle, as well as the vehicle's historical driving characteristics data.

[0065] The trajectory prediction unit can predict the vehicle's trajectory sequence within a preset time window based on historical motion state data. For example, it can use a Bi-LSTM network to predict the trajectories of the vehicle and other vehicles and determine the distance between them in the future time period.

[0066] The spatiotemporal index calculation unit can be used to construct a vehicle geometry model that adapts to the vehicle's motion state, calculate potential conflict points based on the predicted driving trajectory, and calculate the KCI and the degree of advance advantage that characterizes the driver's game interaction intention.

[0067] The collision probability assessment unit can be configured with a probability assessment model that incorporates individual differences. It takes KCI, priority advantage degree and their interaction terms as input features, and combines them with the driver's style feature parameters to output the collision probability of the current vehicle.

[0068] The graded early warning unit can be used to determine the risk level based on the vehicle's collision probability, and trigger differentiated vehicle early warning or control commands based on the game stability verification results and early warning compliance strategy.

[0069] The following is based on Figure 2 The system architecture shown is an example of method 100.

[0070] First, corresponding to step S110, this embodiment of the application can acquire vehicle trajectory data in real time.

[0071] For example, vehicle A (such as the human-driven vehicle in step S110) can obtain its own motion state vector in real time through onboard sensors. (i.e., the trajectory information of the human-driven vehicle in step S110), where For position coordinates, For speed, For acceleration, The heading angle is given. Simultaneously, vehicle A receives a broadcast signal from left-turning vehicle B (such as the autonomous vehicle in step S110) via the V2V communication module, and obtains the motion state vector of vehicle B. And vehicle size information (i.e., the trajectory information of the autonomous vehicle in step S110), such as vehicle length and car width .

[0072] It is worth noting that vehicle A can be an autonomous vehicle, and vehicle B can be a human-driven vehicle. Both of these scenarios can be used in the detailed vehicle warning process described below.

[0073] Subsequently, corresponding to step S120, this embodiment of the application can use a time-series prediction model to predict the vehicle's motion state within a future time window. Furthermore, based on the vehicle's circumcircle model, a preliminary distance screening is performed on the predicted position to eliminate time steps without conflict risk.

[0074] For example, the trajectory prediction unit can input the historical trajectory data sequence of vehicle A and vehicle B within a past time period (i.e., the first time period in step S120) into a pre-trained Bi-LSTM model. Then, the Bi-LSTM model outputs the predicted trajectory sequence of the two vehicles within a future preset time window (i.e., the second time period in step S120) (i.e., the first and second predicted motion state information in step S120). This step not only utilizes historical location information but also captures the acceleration and deceleration intentions of the vehicles during their approach to the intersection through the gating mechanism of the LSTM.

[0075] Subsequently, in this embodiment of the application, a dual-circle model of the vehicle can be constructed based on the predicted state that has passed the initial screening, and the Euclidean distance between the combined center points of the two vehicles can be calculated. If the distance between any two center points is less than a set threshold, a potential collision risk is determined to exist.

[0076] For example, the spatiotemporal index calculation unit can use a vehicle dual-circle geometric model to process the predicted trajectory. Figure 3 A schematic diagram of the double-circle model provided in an embodiment of this application is shown. Figure 3 As shown in the embodiment of this application, the rectangular outlines of vehicles A and B can be mapped to two tangent circles (front circle and rear circle) respectively, and the radius of the double-circle model can be calculated. The formula for calculating the radius R is as follows:

[0077]

[0078] Based on the above formula, calculate each prediction time step. The Euclidean distance between the centers of the double circles of vehicle A and the centers of the double circles of vehicle B. Figure 4 This illustration shows a collision warning diagram for a scenario involving an unsignalized intersection, as provided in an embodiment of this application. Figure 4 As shown, the system identifies the geometric intersection area between the straight-ahead trajectory of vehicle A and the left-turning trajectory of vehicle B, defining it as a potential conflict point. Furthermore, a collision is determined when any of the following conditions are met.

[0079]

[0080]

[0081]

[0082] In other words, during this process, a collision is determined to have occurred when the dual-circle model meets a specific geometric distance condition. This approach allows the dual-circle model to more accurately cover the vehicle's outline than a point mass model, while also being more computationally efficient than a rectangular model.

[0083] Then, corresponding to step S130, this embodiment of the application calculates KCI for potential collision risks and determines the degree of prior advantage characterizing the driver's intention.

[0084] For example, KCI is determined according to the following formula:

[0085]

[0086] In the formula, This refers to the location of a human-driven vehicle (such as vehicle A mentioned above). This refers to the location of autonomous vehicles (such as vehicle B mentioned above). It is the speed at which humans drive vehicles. It is the speed of autonomous vehicles. It is the acceleration of a human driving a vehicle. It is the acceleration of autonomous vehicles. It is the acceleration weight.

[0087] The degree of advance advantage is defined as follows: In a virtual collision scenario, if both vehicles continue moving in their current state, when the vehicle that arrives later reaches the potential collision point, a portion of the body of the vehicle that arrived earlier has already passed the potential collision point. The degree of advance advantage is the process simulated by the system where both vehicles maintain their current state of motion until they reach the potential collision point.

[0088] For example, the degree of first-mover advantage is the ratio of the length already passed by the first vehicle at the point to its total vehicle length. This indicator quantifies the order of arrival at the conflict point, and is calculated using the following formula:

[0089]

[0090]

[0091] In the formula, The degree of first-mover advantage of the vehicle that arrives at the potential collision point first; The degree of priority for vehicles arriving at the point later; The length of the collision point already passed by the first arriving vehicle. This refers to the total length of the vehicle body.

[0092] Furthermore, the calculation process for the degree of leading advantage incorporates a dynamic speed compensation factor. Consequently, the vehicle's equivalent perceived length is linearly superimposed from the vehicle's physical length and the dynamic safety margin that varies with speed. It is an increasing function of velocity and acceleration, used to simulate the nonlinear cognitive characteristics of human drivers regarding dominance under different dynamic states.

[0093] Then, corresponding to step S140, this embodiment of the application inputs the KCI and the degree of precedence advantage into the probability assessment model to output the current vehicle collision probability. .

[0094] For example, the probability assessment model can be a generalized linear mixed model (GLMM), which includes a fixed effects part and a random effects part. The fixed effects part includes an interaction effect term consisting of KCI and AF, which is used to characterize the moderating effect of the driver's perception of the advantage of passage on the sensitivity to spatiotemporal urgency. The random effects part introduces vehicle pair changes as a random intercept term, which is used to characterize the unobserved heterogeneity between different vehicle pairs.

[0095] For example, GLMM can be output according to the following formula Specifically, when the degree of advance advantage is equal to the degree of advance advantage of the autonomous vehicle, that is, when the autonomous vehicle arrives at the collision point first, The following formula is used for calculation:

[0096]

[0097] The formula includes a driving style recognition module, a fixed effects component, and a random effects component. The driving style recognition module extracts historical driving characteristic data of the vehicle in real time, including average acceleration, historical vehicle spacing, and the KCI threshold at historical braking triggers. A mapping model is pre-trained using the dataset to output driving preference coefficients. .

[0098] in, This is used to quantify drivers' perceptual tolerance to spatiotemporal risks; the fixed effects part includes an interaction effect term consisting of KCI and AF, which is used to characterize the moderating effect of drivers' perception of traffic advantages on their sensitivity to spatiotemporal urgency; the random effects part introduces vehicle pair changes as a random intercept term, which is used to characterize the unobserved heterogeneity between different vehicle pairs.

[0099] For example, when the driver is aggressive, This is used to compress the risk response intensity to reduce false alarms; when identified as a conservative driver... This is used to amplify the intensity of the risk response; when the driver is identified as a normal type... ; For the intercept term, , , These are the regression coefficients of KCI, prior advantage degree, and their interaction term, respectively.

[0100] Furthermore, interactive items It played a key regulatory role. In smaller cases, if It's quite large; the interactive items will be corrected. This reduces the final collision probability, thereby avoiding false alarms. A random intercept term is introduced. The current vehicle interaction pairs are modified to characterize the unobserved heterogeneity between different vehicle pairs.

[0101] Furthermore, when the degree of priority is equal to the degree of priority of the human-driven vehicle, that is, when the human-driven vehicle arrives at the collision point first, The following formula is used for calculation:

[0102]

[0103] In other words, for autonomous vehicles, the interaction object is a human-driven vehicle, therefore it is necessary to consider the driver's driving style recognition. For human-driven vehicles, the interaction object is an autonomous vehicle, therefore it is not necessary to consider driving style recognition.

[0104] Then, the vehicle collision probability is calculated using the following formula for the Sigmoid function, which transforms the linear prediction value into... Collision probability between As shown in the following formula:

[0105]

[0106] Finally, corresponding to step S150, the vehicle collision probability obtained by real-time monitoring in this embodiment of the application is... When the risk exceeds the preset high-risk threshold, the risk is marked and an early warning response strategy is triggered.

[0107] For example, the tiered early warning unit calculates the probability. and degree of first-mover advantage Implement differentiated strategies. For example, in scenario 1 (high-risk overtaking), if... and The system determines that vehicle A has no clear advantage and is classified as a Level 3 emergency risk. Vehicle A immediately triggers an audible and visual alarm and sends a braking command to the vehicle controller to prompt the driver to slow down and avoid the obstacle. For example, in scenario 2 (dominant right-of-way), if... It falls within the middle range (e.g., 0.4~0.8), but Vehicle A has a very high advantage, which the system determines is a winning position in the game. At this point, a warning suppression strategy is implemented, displaying only a yellow icon on the instrument panel without emitting a beeping sound to avoid unnecessary startling the driver. For example, in scenario 3 (safety interaction), if... The system remains in silent monitoring mode.

[0108] As another example, embodiments of this application may also set different thresholds for human-driven vehicles and autonomous vehicles.

[0109] For example, the first threshold for human-driven vehicles is determined according to the following formula:

[0110]

[0111] In the formula, This serves as the first threshold based on the data, and also as the first warning threshold for autonomous vehicles.

[0112] The second threshold for human-driven vehicles is determined according to the following formula:

[0113]

[0114] This serves as the second threshold based on the existing threshold, and also as the second warning threshold for autonomous vehicles.

[0115] Furthermore, in this embodiment of the application, after determining that the collision probability of the human-driven vehicle is greater than the first threshold and less than the second threshold, and there is no advantage to passage, a warning is issued to the human-driven vehicle; or, after determining that the collision probability of the human-driven vehicle is greater than the second threshold, and there is no advantage to passage, the human-driven vehicle is forced to brake and decelerate.

[0116] Furthermore, in this embodiment of the application, if it is determined that the collision probability of the autonomous vehicle is greater than the first warning threshold and less than the second warning threshold, and there is no advantage to passing, an alarm is issued to remind surrounding vehicles to pay attention to the distance; or, if it is determined that the collision probability of the autonomous vehicle is greater than the second warning threshold, and there is no advantage to passing, the vehicle is forced to brake and decelerate.

[0117] In summary, this embodiment of the application integrates the degree of prior advantage with KCI by introducing the prior advantage metric. By examining the interaction between the degree of prior advantage and KCI, the model can identify safety scenarios where "although the KCI is short, one party has established a clear prior advantage," thereby effectively reducing the false alarm rate. Furthermore, this embodiment of the application not only focuses on "physical collision avoidance" but also on "social interaction." In other words, traditional KCI only reflects "whether a collision will occur," while this embodiment of the application, combined with the degree of prior advantage, reflects "whether a collision should occur," making the warning logic more consistent with the cognitive habits of human drivers. In addition, this embodiment of the application uses a dual-circle model for collision detection, which significantly reduces computational power consumption compared to fine polygon detection, while ensuring coverage of the actual vehicle body volume compared to a point mass model, making it very suitable for the real-time computing needs of in-vehicle embedded systems.

[0118] This application also provides a computer program product that, when run on an electronic device, causes the electronic device to execute the technical solutions described in the above embodiments. Its implementation principle and technical effects are similar to those of the related embodiments described above, and will not be repeated here.

[0119] This application also provides a readable storage medium containing instructions that, when executed by an electronic device, cause the electronic device to perform the technical solutions described in the above embodiments. The implementation principle and technical effects are similar and will not be repeated here.

[0120] This application also provides a chip for executing instructions. When the chip is running, it executes the technical solutions described in the above embodiments. Its implementation principle and technical effects are similar and will not be repeated here.

[0121] The hardware module of this application is described below, which can be used to implement the aforementioned method 100.

[0122] Now for reference Figure 5 The diagram shows a block diagram of a device 500 according to one embodiment of this application. Device 500 may include one or more processors 501 coupled to a controller hub 503. In at least one embodiment, the controller hub 503 communicates with the processor 501 via a multi-branch bus such as a front side bus (FSB), a point-to-point interface such as a quickpath interconnect (QPI), or a similar connection 510. The processor 501 executes instructions controlling general types of data processing operations. In one embodiment, the controller hub 503 includes, but is not limited to, a graphics memory controller hub (GMCH) (not shown) and an input / output hub (IOH) (which may be on a separate chip) (not shown), wherein the GMCH includes memory and a graphics controller and is coupled to the IOH.

[0123] Device 500 may also include a coprocessor 502 and a memory 504 coupled to a controller hub 503. Alternatively, one or both of the memory and the GMCH may be integrated within the processor, with memory 504 and coprocessor 502 directly coupled to processor 501 and controller hub 503, which resides on a single chip with the IOH. Memory 504 may be, for example, dynamic random access memory (DRAM), phase change memory (PCM), or a combination of both. In one embodiment, coprocessor 502 is a dedicated processor, such as, for example, a high-throughput MIC processor (many integrated core, MIC), a network or communication processor, a compression engine, a graphics processor, a general-purpose computing on GPU (GPGPU), or an embedded processor, etc. Optional properties of coprocessor 502 are indicated by dashed lines. Figure 5 middle.

[0124] Memory 504, as a computer-readable storage medium, may include one or more tangible, non-transitory computer-readable media for storing data and / or instructions. For example, memory 504 may include any suitable non-volatile memory such as flash memory and / or any suitable non-volatile storage device such as one or more hard-disk drives (HDD(s)), one or more compact disc (CD) drives, and / or one or more digital versatile disc (DVD) drives.

[0125] In one embodiment, device 500 may further include a network interface controller (NIC) 506. NIC 506 may include a transceiver for providing a radio interface to device 500, enabling communication with any other suitable device, such as a front-end module, antenna, etc. In various embodiments, NIC 506 may be integrated with other components of device 500. NIC 506 can implement the functionality of the communication unit in the above embodiments.

[0126] Device 500 may further include input / output (I / O) devices 505. I / O 505 may include: a user interface designed to enable a user to interact with device 500; a peripheral component interface designed to enable peripheral components to also interact with device 500; and / or sensors designed to determine environmental conditions and / or location information related to device 500.

[0127] It is worth noting that, Figure 6 This is merely an example. That is, although... Figure 6 The diagram shows that device 600 includes multiple devices such as processor 601, controller hub 603, and memory 604. However, in actual applications, devices using the methods of this application may include only a portion of the devices in device 600. For example, it may include only processor 601 and NIC 606. Figure 6 The properties of the optional devices are shown in dashed lines. According to some embodiments of this application, the memory 604, which is a computer-readable storage medium, stores instructions that, when executed on a computer, cause the device 600 to perform the methods according to the above embodiments. Specific details can be found in the methods of the above embodiments, and will not be repeated here.

[0128] Now for reference Figure 6 The diagram shown is a block diagram of a SoC 600 according to an embodiment of this application. Figure 6 In the diagram, similar components share the same reference numerals. Additionally, dashed boxes are an optional feature for more advanced SoCs. Figure 6 In this SoC 600, the following are included: an interconnect unit 650 coupled to an application processor 610; a system proxy unit 680; a bus controller unit 690; an integrated memory controller unit 640; a group or one or more coprocessors 620, which may include integrated graphics logic, an image processor, an audio processor, and a video processor; a static random access memory (SRAM) unit 630; and a direct memory access (DMA) unit 660. In one embodiment, the coprocessor 620 includes a dedicated processor, such as, for example, a network or communication processor, a compression engine, a GPGPU, a high-throughput MIC processor, or an embedded processor.

[0129] The static random-access memory (SRAM) unit 630 may include one or more computer-readable media for storing data and / or instructions. The computer-readable storage medium may store instructions, specifically, temporary and permanent copies of those instructions. These instructions may include, when executed by at least one unit in the processor, causing the SoC 600 to perform the vehicle collision warning method according to the above embodiments, as detailed in the methods described above, which will not be repeated here.

[0130] Various embodiments of the mechanisms disclosed in this application can be implemented in hardware, software, firmware, or combinations of these implementation methods. Embodiments of this application can be implemented as computer programs or program code executable on a programmable system, the programmable system including at least one processor, a storage system (including volatile and non-volatile memory and / or storage elements), at least one input device, and at least one output device.

[0131] Program code can be applied to input instructions to execute the functions described in this application and generate output information. The output information can be applied to one or more output devices in a known manner. For the purposes of this application, the processing system includes any system having a processor such as, for example, a digital signal processor (DSP), a microcontroller, an application-specific integrated circuit (ASIC), or a microprocessor.

[0132] The program code can be implemented using a high-level procedural language or an object-oriented programming language to communicate with the processing system. Assembly language or machine language can also be used when needed. In fact, the mechanisms described in this application are not limited to any particular programming language. In either case, the language can be a compiled language or an interpreted language.

[0133] In some cases, the disclosed embodiments may be implemented in hardware, firmware, software, or any combination thereof. The disclosed embodiments may also be implemented as instructions carried or stored thereon on one or more temporary or non-temporary machine-readable (e.g., computer-readable) storage media, which may be read and executed by one or more processors. For example, the instructions may be distributed via a network or through other computer-readable media. Therefore, machine-readable media may include any mechanism for storing or transmitting information in a machine-readable (e.g., computer-readable) form, including but not limited to floppy disks, optical disks, CD-ROMs, compact disc read-only memory (CD-ROMs), magneto-optical disks, read-only memory (ROM), random-access memory (RAM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic cards or optical cards, flash memory, or tangible machine-readable storage for transmitting information (e.g., carrier waves, infrared signals, digital signals, etc.) using the Internet in the form of electrical, optical, acoustic, or other forms of propagated signals. Therefore, machine-readable media includes any type of machine-readable medium suitable for storing or transmitting electronic instructions or information in a machine-readable (e.g., computer-readable) form.

[0134] In the accompanying drawings, some structural or methodological features may be shown in a specific arrangement and / or order. However, it should be understood that such a specific arrangement and / or order may not be necessary. Rather, in some embodiments, these features may be arranged in a manner and / or order different from that shown in the accompanying drawings. Furthermore, including structural or methodological features in a particular figure does not imply that such features are required in all embodiments, and in some embodiments, these features may be omitted or may be combined with other features.

[0135] It should be noted that all units / modules mentioned in the device embodiments of this application are logical units / modules. Physically, a logical unit / module can be a physical unit / module, a part of a physical unit / module, or a combination of multiple physical units / modules. The physical implementation of these logical units / modules themselves is not the most important factor; the combination of functions implemented by these logical units / modules is the key to solving the technical problems proposed in this application. Furthermore, to highlight the innovative aspects of this application, the above-described device embodiments of this application have not introduced units / modules that are not closely related to solving the technical problems proposed in this application. This does not mean that the above-described device embodiments do not contain other units / modules.

[0136] It should be noted that in the examples and description of this patent, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, 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 limitations, 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.

[0137] Although this application has been illustrated and described with reference to certain preferred embodiments thereof, those skilled in the art should understand that various changes in form and detail may be made thereto without departing from the spirit and scope of this application.

Claims

1. A vehicle collision warning method in a mixed-traffic environment, characterized in that, include: Acquire the trajectory information of human-driven vehicles and autonomous vehicles in the first time period. Based on the trajectory information of the human-driven vehicle, a first predicted motion state information of the human-driven vehicle in a second time period is determined, and based on the trajectory information of the autonomous vehicle, a second predicted motion state information of the autonomous vehicle in the second time period is determined, wherein the second time period is after the first time period. Based on the first predicted motion state information and the second predicted motion state information, the motion collision index KCI of the human-driven vehicle and the autonomous vehicle is determined, and the degree of priority advantage of the vehicle that arrives at the predicted collision point first among the human-driven vehicle and the autonomous vehicle is determined. The KCI is used to indicate the risk of spatiotemporal overlap, and the degree of priority advantage is used to indicate the strength of the intention of the vehicle that arrives at the predicted collision point first among the human-driven vehicle and the autonomous vehicle to pass first. Based on the KCI and the degree of precedence advantage, the collision probability between the human-driven vehicle and the autonomous vehicle is determined; Based on the collision probability, different forms of warnings are issued to the human-driven vehicle and the autonomous vehicle.

2. The method according to claim 1, characterized in that, The step of determining the first predicted motion state information of the human-driven vehicle in the second time period based on the trajectory information of the human-driven vehicle includes: The trajectory information of the human-driven vehicle and the driving style characteristics of the human-driven vehicle are input into the time series prediction model to obtain the first predicted motion state information output by the time series prediction model. The driving style characteristics of the human-driven vehicle are extracted based on at least one of the following factors: historical driving behavior data of the human-driven vehicle, acceleration, steering angle change, and vehicle speed fluctuation.

3. The method according to claim 1 or 2, characterized in that, Determining the KCI of the human-driven vehicle and the autonomous vehicle based on the first predicted motion state information and the second predicted motion state information includes: Based on the first predicted motion state information and the second predicted motion state information, it is determined that there is a potential collision risk between the human-driven vehicle and the autonomous vehicle. The KCI is determined based on the potential collision risks of the human-driven vehicle and the autonomous vehicle. The KCI is determined according to the following formula: In the formula, This refers to the location of the human-driven vehicle. This refers to the location of the autonomous vehicle. It refers to the speed at which humans drive vehicles. It is the speed of the autonomous vehicle. It is the acceleration of the human-driven vehicle. It is the acceleration of the autonomous vehicle. It is the acceleration weight.

4. The method according to claim 3, characterized in that, The step of determining a potential collision risk between the human-driven vehicle and the autonomous vehicle based on the first predicted motion state information and the second predicted motion state information includes: Based on the first predicted motion state information and the second predicted motion state information, and based on the vehicle circumcircle model, the predicted positions of the human-driven vehicle and the autonomous vehicle are screened by distance, and time steps without conflict risk are eliminated to obtain the preliminary predicted states of the human-driven vehicle and the autonomous vehicle. Based on the predicted states of the human-driven vehicle and the autonomous vehicle after the initial screening, as well as the vehicle double-circle model, the Euclidean distance between the front and rear center centers of the human-driven vehicle and the autonomous vehicle is determined. Based on the Euclidean distance between the front and rear centers of the human-driven vehicle and the autonomous vehicle, and a preset distance threshold, it is determined that there is a potential collision risk between the human-driven vehicle and the autonomous vehicle.

5. The method according to claim 1 or 2, characterized in that, The prior advantage is the ratio of the length of the collision point already passed by the vehicle that arrives first to the first equivalent perceived length. The first equivalent perceived length is determined based on the physical vehicle length and dynamic safety margin of the vehicle that arrives first at the predicted collision point in the human-driven vehicle and the autonomous vehicle. The dynamic safety margin is an increasing function of speed and acceleration. The prior advantage is obtained according to the following formula: In the formula, The degree of first-mover advantage of the vehicle that arrives at the potential collision point first; The degree of priority for vehicles arriving at the point later; The length that the vehicle that arrived at the collision point first has already passed through the collision point. The total length of the vehicle body is... This refers to the dynamic safety margin.

6. The method according to claim 1 or 2, characterized in that, Determining the collision probability between the human-driven vehicle and the autonomous vehicle based on the KCI and the prior advantage degree includes: Obtain the driving preference coefficient of the driver of the human-driven vehicle. The driving preference coefficient is used to quantify the driver's emotional tolerance to spatiotemporal risks. The collision probability between the human-driven vehicle and the autonomous vehicle is determined based on the driving preference coefficient, the KCI, and the degree of precedence advantage. The collision probability is calculated according to the following formula: In the formula, the random intercept term Used to characterize unobserved heterogeneity between different vehicle pairs , representing the random intercept term between different vehicle pairs, The following formula is used for calculation: When the degree of leading advantage is equal to the degree of leading advantage of the autonomous vehicle, When the degree of prior advantage is equal to the degree of prior advantage of the human-driven vehicle, In the formula, AF represents the degree of prior advantage. For the intercept term, , , These are the regression coefficients of KCI, prior advantage degree, and their interaction term, respectively.

7. The method according to claim 1 or 2, characterized in that, The provision of different forms of warnings for the human-driven vehicle and the autonomous vehicle based on the collision probability includes: If the collision probability of the human-driven vehicle is determined to be greater than a first threshold and less than a second threshold, and there is no advantage to passage, a warning is issued; or, If the probability of a collision with the human-driven vehicle is determined to be greater than the second threshold and there is no advantage to passing, then the vehicle is forced to brake and decelerate. Wherein, the first threshold is The second threshold is , This serves as the first threshold based on the autonomous vehicle's operating system and also as the first warning threshold for autonomous vehicles. This serves as the second threshold based on the existing threshold, and also as the second warning threshold for autonomous vehicles. as well as, If the collision probability of the autonomous vehicle is determined to be greater than the first warning threshold and less than the second warning threshold, and there is no advantage to pass, an alarm is issued to remind surrounding vehicles to pay attention to the distance. If the probability of collision of the autonomous vehicle is determined to be greater than the second warning threshold and there is no advantage to pass, then the vehicle will be forced to brake and decelerate.

8. An electronic device, characterized in that, It includes one or more processors; one or more memories; said one or more memories storing one or more computer programs, said one or more computer programs including instructions that, when executed by said one or more processors, cause the method of any one of claims 1 to 7 to be performed.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed on a computer, cause the method as described in any one of claims 1 to 7 to be performed.

10. A computer program product, characterized in that, When the computer program product is run on a computer, it causes the computer to perform the method as described in any one of claims 1 to 7.

Images