A collision analysis method and device for a vehicle trajectory, a vehicle, and a medium

Abstract

This application provides a method, apparatus, vehicle, and medium for collision analysis of vehicle trajectories. The method includes: acquiring a planned trajectory of a vehicle and at least one predicted target trajectory; for each trajectory in the planned trajectory and the predicted target trajectory, determining the spatial occupancy envelope of that trajectory at each time step based on the trajectory point parameters at each time step; determining the single-branch collision probability of the planned trajectory relative to the predicted target trajectory based on the overlap of the spatial occupancy envelopes of the planned trajectory and the predicted target trajectory at each time step; and determining the fusion collision probability of the planned trajectory based on the single-branch collision probabilities of the planned trajectory relative to each predicted target trajectory. This application can achieve comprehensive and effective analysis of vehicle trajectory collision risks, enhance the versatility of vehicle trajectory collision analysis in various road scenarios, significantly improve the foresight and robustness of vehicle trajectory collision analysis, and ensure vehicle driving safety.

Description

Technical Field

[0001] This application relates to the field of vehicle data processing technology, specifically to a collision analysis method, device, vehicle, and medium for vehicle trajectories. Background Technology

[0002] With the rapid development of vehicle automation technology, the decision-making and planning functions of autonomous driving systems are used to plan corresponding driving trajectories for vehicles. Moreover, for each planned trajectory, it is usually necessary to analyze whether there is any corresponding collision risk in order to ensure safe driving and efficient passage.

[0003] In related technologies, the real-time driving status of a vehicle at each moment in its planned trajectory is typically analyzed to predict instantaneous driving parameters at each moment, such as following distance, braking distance, or time-to-collision (TTC), thereby determining whether there is an instantaneous collision risk in the planned trajectory. However, existing vehicle trajectory collision analysis methods are too simplistic and cannot be effectively used in various complex real-world road scenarios, greatly reducing the accuracy and reliability of vehicle trajectory collision analysis. Summary of the Invention

[0004] This application provides a collision analysis method, device, vehicle, and medium for vehicle trajectory, enabling comprehensive and effective analysis of vehicle trajectory collision risks, providing a more accurate and reliable basis for collision analysis, and improving vehicle driving safety.

[0005] In a first aspect, embodiments of this application provide a collision analysis method for vehicle trajectories, the method comprising: Obtain the planned trajectory of the vehicle and at least one predicted trajectory of the target; For each trajectory in the planned trajectory of the autonomous vehicle and the predicted trajectory of the target, the spatial occupancy envelope of the trajectory at each time step is determined based on the trajectory point parameters at each time step. Based on the overlap of the spatial occupancy envelopes of the planned vehicle trajectory and the predicted target trajectory at each time step, the single-branch collision probability of the planned vehicle trajectory relative to the predicted target trajectory is determined. The fusion collision probability of the planned vehicle trajectory is determined based on the single-branch collision probability of the planned vehicle trajectory relative to each of the predicted target trajectories.

[0006] Secondly, embodiments of this application provide a collision analysis device for vehicle trajectory, the device comprising: The trajectory acquisition module is used to acquire the planned trajectory of the vehicle and at least one predicted trajectory of the target. The envelope determination module is used to determine the spatial occupancy envelope of each trajectory in the planned trajectory of the vehicle and the predicted trajectory of the target, based on the trajectory point parameters at each time step of the trajectory. The first collision analysis module is used to determine the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory based on the overlap of the spatial occupancy envelope of the self-planned trajectory and the target predicted trajectory at each time step. The second collision analysis module is used to determine the fusion collision probability of the self-planned trajectory based on the single-branch collision probability of the self-planned trajectory relative to each of the predicted trajectory of the target.

[0007] Thirdly, embodiments of this application provide a vehicle, the vehicle comprising: A processor and a memory, the memory being used to store a computer program, and the processor being used to call and run the computer program stored in the memory to execute the collision analysis method for vehicle trajectory provided in the first aspect of this application.

[0008] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements the collision analysis method for vehicle trajectories as provided in the first aspect of this application.

[0009] Fifthly, embodiments of this application provide a computer program product, including a computer program / instructions, which, when executed by a processor, implement the collision analysis method for vehicle trajectories as provided in the first aspect of this application.

[0010] The technical solution provided in this application acquires the planned trajectory of the vehicle and at least one predicted target trajectory in real time. For each trajectory, based on the trajectory point parameters at each time step, the spatial occupancy envelope of that trajectory at each time step is determined. This constructs a dynamically changing spatiotemporal safety boundary for the planned trajectory and the predicted target trajectory, more accurately representing their uncertainties. Then, by analyzing the overlap of the spatial occupancy envelopes of the planned trajectory and any predicted target trajectory at each time step, the single-branch collision probability of the planned trajectory relative to each predicted target trajectory is determined. This determines the fusion collision probability of the planned trajectory, enabling a comprehensive and effective analysis of vehicle trajectory collision risk. This further enhances the versatility of vehicle trajectory collision analysis across various road scenarios and optimizes instantaneous collision analysis into cumulative collision analysis over a continuous future time domain. This provides a more accurate and reliable basis for collision analysis and sufficient collision warning time, significantly improving the foresight and robustness of vehicle trajectory collision analysis and ensuring vehicle driving safety. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart illustrating a collision analysis method for vehicle trajectory provided in an embodiment of this application; Figure 2 A flowchart of another collision analysis method for vehicle trajectory provided in an embodiment of this application; Figure 3 A schematic block diagram of a collision analysis device for vehicle trajectory provided in an embodiment of this application; Figure 4 A schematic block diagram of a vehicle provided in an embodiment of this application. Detailed Implementation

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0015] To address the problem that existing vehicle trajectory collision analysis methods are overly simplistic and limited in effectiveness across complex real-world road scenarios, significantly reducing their accuracy and reliability, this application proposes a novel collision analysis scheme. This scheme constructs a spatial occupancy envelope for each time step of the planned vehicle trajectory and the predicted target trajectory, thereby more accurately representing the uncertainties of both. Then, by analyzing the overlap of the spatial occupancy envelopes of the planned vehicle trajectory and any predicted target trajectory at the same time step, the single-branch collision probability of the planned vehicle trajectory relative to each predicted target trajectory is determined. This, in turn, determines the fusion collision probability of the planned vehicle trajectory, achieving a comprehensive and effective analysis of vehicle trajectory collision risk. This enhances the versatility of vehicle trajectory collision analysis across various road scenarios, providing a more accurate and reliable basis for collision analysis and improving vehicle driving safety.

[0016] Figure 1 This is a flowchart illustrating a collision analysis method for vehicle trajectory provided in an embodiment of this application. This method can be executed by the collision analysis device for vehicle trajectory provided in this application. The collision analysis device for vehicle trajectory can be implemented using any software and / or hardware method. Exemplarily, this collision analysis device for vehicle trajectory can be applied to any vehicle, which can be a vehicle supporting autonomous driving.

[0017] Specifically, such as Figure 1 As shown, the method may include the following steps: S110, acquire the planned trajectory of the vehicle and at least one predicted trajectory of a target.

[0018] For vehicle automated driving technology, an autonomous driving system is usually pre-configured on the vehicle. This autonomous driving system has certain decision-making and planning functions, which are used to continuously plan and adjust the vehicle's driving trajectory in the future by analyzing the actual driving status of the vehicle in real time.

[0019] Therefore, in order to achieve autonomous driving, at each moment during the vehicle's operation, the decision-making and planning functions within the vehicle's autonomous driving system can be used to comprehensively analyze the vehicle's historical and current driving states to determine its current driving intention, such as accelerating straight ahead, turning, or changing lanes. Then, based on the vehicle's current driving intention, its driving trajectory for a future period of time is planned, which serves as the vehicle's planned trajectory in this application.

[0020] Furthermore, considering that vehicles often encounter dynamically changing obstacles such as other vehicles and pedestrians during operation, posing certain safety hazards, the aforementioned autonomous driving system equipped on the vehicle also possesses a trajectory prediction function to ensure driving safety. This function continuously predicts the trajectory of each dynamic obstacle in the future by analyzing the actual movement state of each obstacle around the vehicle in real time.

[0021] It is understood that the target in this application can be various dynamic obstacles that exist around the vehicle during its driving process, such as other vehicles or pedestrians around the vehicle.

[0022] Therefore, in order to ensure vehicle driving safety, at every moment during the vehicle's driving process, in addition to acquiring the vehicle's planned trajectory in real time, the trajectory prediction function in the autonomous driving system configured in the vehicle can be used to comprehensively analyze the historical and current motion states of each target around the vehicle, so as to determine the current motion intention of each target and predict the motion trajectory of each target in the future period of time, which is used as the target prediction trajectory in this application.

[0023] It should be noted that, for each target, this application can use the trajectory prediction function in the above-mentioned autonomous driving system to output multiple target prediction trajectories for each target, and output a corresponding prediction probability for each target prediction trajectory. The prediction probability of each target prediction trajectory can represent the probability that the target will actually move along this target prediction trajectory in the future.

[0024] In this context, the entire trajectory duration of both the planned trajectory of the vehicle and the predicted trajectory of each target is a pre-defined fixed duration, representing the driving situation of the vehicle and the target within the future time domain T. Moreover, the planned trajectory of the vehicle and the predicted trajectory of each target can be composed of a series of discrete spatiotemporal points at each time step within the future time domain T, and each spatiotemporal point in the planned trajectory of the vehicle and the predicted trajectory of each target represents a trajectory point at one time step.

[0025] S120: For each trajectory in the planned trajectory of the vehicle and the predicted trajectory of the target, determine the spatial occupancy envelope of the trajectory at each time step based on the trajectory point parameters at each time step of the trajectory.

[0026] Since both the planned trajectory of the vehicle and the predicted trajectory of the target are calculated based on the vehicle's historical driving data and are the most likely driving trajectories over a future period, rather than the vehicle's actual driving trajectory, the actual trajectory of the vehicle during driving may deviate from the planned trajectory, and the actual trajectory of the target during driving may also deviate from the predicted trajectory. Therefore, in order to accurately analyze whether there is a risk of collision between the vehicle and each target during driving, this application first needs to analyze the range of trajectory position deviation between the planned trajectory of the vehicle and the predicted trajectory at each time step, thereby accurately determining the probability of collision between the planned trajectory of the vehicle and the predicted trajectory at each time step.

[0027] Therefore, after obtaining the planned trajectory of the vehicle and the predicted trajectories of each target, this application can perform corresponding vehicle driving state analysis on each trajectory to obtain various driving parameters at each time step, which are then used as trajectory point parameters at that time step. Following this method, the trajectory point parameters of each trajectory in the planned trajectory of the vehicle and the predicted trajectories at each time step can be obtained.

[0028] The trajectory point parameters at each time step are used to characterize the vehicle's driving state at that time step, and may include, but are not limited to, various driving parameters such as driving position, speed, acceleration, and heading angle.

[0029] Then, for each trajectory in the planned trajectory and the predicted target trajectory, this application can analyze the trajectory point parameters at each time step to determine the vehicle's driving state at each time step, thereby analyzing the driving position deviation of the trajectory at each time step. Then, based on the trajectory point positions at each time step and the driving position deviation of the trajectory at that time step, the spatial occupancy envelope of the trajectory at each time step is constructed.

[0030] The spatial occupancy envelope of the vehicle's planned trajectory at each time step characterizes the spatial range that the vehicle may occupy at that future time step. Similarly, the spatial occupancy envelope of the target's predicted trajectory at each time step characterizes the spatial range that the target may occupy at that future time step.

[0031] S130, based on the overlap of the spatial occupancy envelopes of the self-planned trajectory and the target predicted trajectory at each time step, determine the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory.

[0032] It is understandable that the entire trajectory duration of the self-planned trajectory and the predicted trajectory of each target is a fixed set duration, and has the same multiple time steps, so as to accurately analyze the collision risk of the self-planned trajectory and each target at each time.

[0033] Therefore, after obtaining the spatial occupancy envelope of each trajectory in the planned vehicle trajectory and the predicted target trajectory at each time step, this application can traverse each predicted target trajectory to sequentially determine the overlap between the spatial occupancy envelope of the planned vehicle trajectory at each time step and the spatial occupancy envelope of the currently traversed predicted target trajectory at the same time step. This allows for analysis of the probability of a collision between the planned vehicle trajectory and the currently traversed predicted target trajectory at each time step. Then, by comprehensively analyzing the probability of a collision between the planned vehicle trajectory and the currently traversed predicted target trajectory at each time step, the likelihood of a collision between the planned vehicle trajectory and the currently traversed predicted target trajectory in the future is determined. This determines the single-branch collision probability of the planned vehicle trajectory relative to the currently traversed predicted target trajectory, which is used to characterize the magnitude of the probability of a collision between the planned vehicle trajectory and the predicted target trajectory.

[0034] Following the same method described above, the single-branch collision probability of the vehicle's planned trajectory relative to each target predicted trajectory can be determined.

[0035] S140, determine the fusion collision probability of the self-planned trajectory based on the single-branch collision probability of the self-planned trajectory relative to the predicted trajectories of each target.

[0036] After obtaining the single-branch collision probability of the vehicle's planned trajectory relative to each target's predicted trajectory, this application can use the target situation around the vehicle and the relative driving state of the vehicle and each target to fuse the single-branch collision probability of the vehicle's planned trajectory relative to each target's predicted trajectory, thereby obtaining the fused collision probability of the vehicle's planned trajectory.

[0037] For example, this application can add the single-branch collision probabilities of the self-planned trajectory relative to the predicted trajectories of each target, and then average the sum of probabilities according to the number of targets around the self-planned trajectory to obtain the fused collision probability of the self-planned trajectory.

[0038] In some implementations, considering that multiple targets may exist around the vehicle, and that the trajectory prediction function in the vehicle's autonomous driving system may provide multiple predicted trajectories for each target, the vehicle will face a corresponding collision risk with each target. Therefore, this application can determine the fused collision probability of the vehicle's planned trajectory through the following steps: Step 1: Based on the single-branch collision probability of the vehicle's planned trajectory relative to each target's predicted trajectory, determine the individual collision probability of the vehicle's planned trajectory relative to each target.

[0039] Specifically, after obtaining the single-branch collision probability of the vehicle's planned trajectory relative to each target predicted trajectory, this application can determine which target each target predicted trajectory belongs to, thereby dividing each target predicted trajectory into a trajectory set corresponding to different targets. Then, for each target, this application can calculate the sum of the single-branch collision probabilities of the vehicle's planned trajectory relative to each target predicted trajectory in the trajectory set of the same target, to obtain the individual collision probability of the vehicle's planned trajectory relative to that target. Following the same method described above, the individual collision probability of the vehicle's planned trajectory relative to each target can be determined.

[0040] Step 2: Determine the fused collision probability of the vehicle's planned trajectory based on the collision confidence of each target and the individual collision probability of the vehicle's planned trajectory relative to each target; or, perform a union operation on the individual collision probabilities of the vehicle's planned trajectory relative to each target to obtain the fused collision probability of the vehicle's planned trajectory.

[0041] After obtaining the individual collision probabilities of the vehicle's planned trajectory relative to each target, the fused collision probability of the vehicle's planned trajectory can be calculated in the following two ways: Method 1: By analyzing the relative driving states between the vehicle and each target, the probability of collision between the vehicle and each target is determined, and an appropriate collision confidence level is assigned to each target. The collision confidence level of each target represents the likelihood of a collision between the vehicle and the target, and the sum of the collision confidence levels of all targets is 1. Then, the individual collision probabilities of the vehicle's planned trajectory relative to each target are weighted and summed using the collision confidence levels of each target to obtain the fused collision probability of the vehicle's planned trajectory.

[0042] Method 2: Calculate the individual non-collision probability of the vehicle's planned trajectory relative to each target by analyzing the individual collision probability of the vehicle's planned trajectory relative to each target. Then, use the product of the individual non-collision probabilities of the vehicle's planned trajectory relative to each target to calculate the union probability of the individual collision probabilities of the vehicle's planned trajectory relative to each target, and use this union probability as the fused collision probability of the vehicle's planned trajectory.

[0043] For example, the fused collision probability of the vehicle's planned trajectory can be calculated using the following formula:

[0044] in, The probability of collision between the planned trajectory of the vehicle and the j-th target.

[0045] In other embodiments, the decision-making and planning function of the autonomous driving system configured in the vehicle can plan multiple driving trajectories for the vehicle, serving as multiple candidate trajectories that the vehicle can refer to at each moment of travel. Therefore, the vehicle planning trajectory in this application can be at least one candidate trajectory planned for the vehicle. Thus, according to the overall calculation process of the fused collision probability of the vehicle planning trajectory provided in steps S110 to S140 above, a fused collision probability can be calculated for each candidate trajectory, which is used to characterize the probability of a collision when the vehicle travels along this candidate trajectory.

[0046] Therefore, in order to ensure the safe driving of the vehicle, this application can also determine the corresponding vehicle driving trajectory based on the fusion collision probability of each candidate trajectory.

[0047] Specifically, to ensure safe vehicle operation, this application can pre-set a collision probability threshold to characterize the minimum probability of a collision occurring to the vehicle. After obtaining the fused collision probability of each candidate trajectory, this application can determine whether a collision will occur when the vehicle travels along a candidate trajectory by analyzing whether the fused collision probability of each candidate trajectory exceeds the preset collision probability threshold.

[0048] If the fused collision probability of each candidate trajectory is greater than or equal to the aforementioned collision probability threshold, it indicates that the probability of a collision for each candidate trajectory is relatively high. In this case, this application can generate a replanning request and send it to the autonomous driving system to replan the corresponding candidate trajectory for the vehicle through the decision-making and planning function of the autonomous driving system. Alternatively, in this case, this application can directly generate a warning message and output it to the driver to alert the driver of the potential collision ahead.

[0049] If at least one candidate trajectory has a fusion collision probability lower than the aforementioned collision probability threshold, it indicates that the probability of a collision with this candidate trajectory is low. In this case, this application can select a subset of candidate trajectories whose fusion collision probability is lower than the aforementioned collision probability threshold. Then, using the fusion collision probabilities of the selected candidate trajectories, and combining them with other cost information such as comfort cost, progress cost, and traffic rule cost, the overall driving cost of the selected candidate trajectories is comprehensively analyzed. This allows for further selection of the optimal candidate trajectory from among the selected candidate trajectories, which serves as the vehicle's driving trajectory in this application, ensuring the safety of the vehicle's driving.

[0050] The technical solution provided in this application acquires the planned trajectory of the vehicle and at least one predicted target trajectory in real time. For each trajectory, based on the trajectory point parameters at each time step, the spatial occupancy envelope of that trajectory at each time step is determined. This constructs a dynamically changing spatiotemporal safety boundary for the planned trajectory and the predicted target trajectory, more accurately representing their uncertainties. Then, by analyzing the overlap of the spatial occupancy envelopes of the planned trajectory and any predicted target trajectory at each time step, the single-branch collision probability of the planned trajectory relative to each predicted target trajectory is determined. This determines the fusion collision probability of the planned trajectory, enabling a comprehensive and effective analysis of vehicle trajectory collision risk. This further enhances the versatility of vehicle trajectory collision analysis across various road scenarios and optimizes instantaneous collision analysis into cumulative collision analysis over a continuous future time domain. This provides a more accurate and reliable basis for collision analysis and sufficient collision warning time, significantly improving the foresight and robustness of vehicle trajectory collision analysis and ensuring vehicle driving safety.

[0051] As an optional implementation scheme in this application, in order to provide a more accurate and reliable collision analysis basis and sufficient collision warning time for vehicle trajectory, it is first necessary to ensure the accuracy of the spatial occupancy envelope of each trajectory in the self-planned trajectory and the target predicted trajectory at each time step, and the single-branch collision probability of the self-planned trajectory relative to each target predicted trajectory. The following combines... Figure 2 This application can provide a detailed explanation of the spatial occupancy envelope of each trajectory in the self-planned trajectory and the target predicted trajectory involved in step S120 at each time step, as well as the specific determination process of the single-branch collision probability of the self-planned trajectory relative to each target predicted trajectory involved in step S130.

[0052] Figure 2 A flowchart of another collision analysis method for vehicle trajectory provided in this application embodiment, the method specifically may include the following steps: S210, obtain the planned trajectory of the vehicle and at least one predicted trajectory of the target.

[0053] S220: For each trajectory in the planned trajectory of the vehicle and the predicted trajectory of the target, determine the envelope reference size of the trajectory at each time step based on the physical size of the trajectory object and the heading angle at each time step in the trajectory. The trajectory object is the vehicle or the target.

[0054] Because the vehicle and various targets may have different dimensions, the actual space occupied by the vehicle and targets during travel will also differ. Furthermore, the actual driving state of the vehicle will be analyzed based on changes in its position and direction of travel within the same reference coordinate system. This reference coordinate system can be the world coordinate system or a pre-defined third-party coordinate system for any third-party object.

[0055] Therefore, in order to accurately analyze whether there is a collision risk between the vehicle and various targets during the driving process, this application first needs to analyze the trajectory position deviation range of the vehicle's planned trajectory and the target's predicted trajectory at each time step, so as to accurately determine the collision probability of the vehicle's planned trajectory and the target's predicted trajectory at each time step. At this time, for each trajectory in the vehicle's planned trajectory and each target's predicted trajectory, this application first determines whether the trajectory object is the vehicle itself or a target around the vehicle, and obtains the actual length and width of the trajectory object.

[0056] Furthermore, because the vehicle's direction of travel constantly changes during actual driving, the basic spatial dimensions that the vehicle may occupy in the horizontal and vertical axes of the reference coordinate system will also change. Therefore, for each trajectory in the self-planned trajectory and the predicted trajectories of each target, this application can also determine the heading angle at each time step of the trajectory by analyzing the trajectory points at each time step. This heading angle can represent the angle between the travel direction of the trajectory object and the vertical axis of the reference coordinate system.

[0057] Therefore, for each trajectory in the planned trajectory of the autonomous vehicle and the predicted trajectories of each target, this application can, according to the heading angle at each time step of the trajectory, sequentially distribute the physical dimensions of the trajectory object onto the horizontal and vertical axes of the reference coordinate system at each time step, thereby obtaining the envelope reference dimensions at each time step of the trajectory. The envelope reference dimensions can include the horizontal axis reference dimensions and the vertical axis reference dimensions of the trajectory object, corresponding to the physical dimensions of the trajectory object distributed onto the horizontal and vertical axes of the reference coordinate system.

[0058] S230, based on the envelope reference size and multi-dimensional driving state quantity at each time step in the trajectory, determine the spatial occupancy envelope of the trajectory at each time step.

[0059] For each trajectory in the planned vehicle trajectory and the predicted trajectories for each target, after obtaining the envelope reference size at each time step of the trajectory, this application further obtains various driving parameters at each time step of the trajectory as multi-dimensional driving state quantities at that time step. Then, by analyzing the multi-dimensional driving state quantities at each time step of the trajectory, the vehicle driving state at each time step of the trajectory is determined, thereby analyzing the driving position deviation of the trajectory at each time step. Then, based on the envelope reference size at each time step of the trajectory, combined with the driving position deviation of the trajectory at that time step, the spatial occupancy envelope of the trajectory at each time step is constructed.

[0060] In some implementations, the aforementioned multidimensional driving state quantities may include at least trajectory point position, velocity, acceleration, and heading angle. Therefore, by comprehensively analyzing the trajectory point position, velocity vector, and acceleration vector at each time step in the trajectory, the vehicle's driving state at each time step can be determined, thereby analyzing the deviation of the driving position at each time step. For example, this application can set the shape of the space occupancy envelope to be elliptical. Then, for each trajectory in the vehicle's planned trajectory and each target prediction trajectory, the envelope reference size at each time step of the trajectory can represent the length of the major axis and minor axis of the ellipse of the space occupancy envelope of the trajectory at that time step.

[0061] Therefore, the major and minor axes of the spatial occupancy envelope of this trajectory at each time step can be dynamically determined by the following formula:

[0062] in, , The physical dimensions of the trajectory object in either the planned trajectory of the vehicle or the predicted trajectory of the target are the horizontal and vertical reference dimensions distributed along the horizontal and vertical axes of the reference coordinate system at any time step t.

[0063] The velocity vector of the trajectory object at any time step t for either the planned trajectory of the vehicle or the predicted trajectory of the target.

[0064] , These are the lateral and longitudinal accelerations of the trajectory object at any time step t, representing either the planned trajectory of the vehicle or the predicted trajectory of the target.

[0065] , These are the lateral and longitudinal standard deviations of any trajectory in the planned trajectory and the predicted trajectory of the target vehicle at any time step t, respectively. They are used to characterize the lateral and longitudinal uncertainties of the trajectory points at any time step t. They can be determined by Bayesian filtering or neural networks outputting the covariance matrix of the trajectory points at any time step t.

[0066] , , , , These are the pre-set weighting coefficients for each of the above parameters. and They can be the same or different. and They can be the same or different. This application addresses... , , , , The specific value is not limited and can be adaptively configured according to actual needs.

[0067] Therefore, by using the above formula, the lengths of the major and minor axes of the ellipse of any trajectory in the planned trajectory of the vehicle and the predicted trajectory of the target can be calculated at each time step t, thereby accurately constructing the spatial occupancy envelope of this trajectory at each time step t.

[0068] S240, based on the overlapping area and total area of ​​the spatial occupancy envelopes of the self-planned trajectory and the target predicted trajectory at the same time step, determine the envelope overlap degree of the self-planned trajectory and the target predicted trajectory at that time step.

[0069] After obtaining the spatial occupancy envelope of each trajectory in the planned vehicle trajectory and the predicted target trajectory at each time step, this application can, for each same time step in the planned vehicle trajectory and any predicted target trajectory, traverse each predicted target trajectory to sequentially analyze the specific size of the two spatial occupancy envelopes at the same time step, thereby determining the actual area of ​​these two spatial occupancy envelopes. Furthermore, by analyzing the specific overlap between these two spatial occupancy envelopes, the overlapping area of ​​these two spatial occupancy envelopes can be determined.

[0070] For example, the shape of the spatial occupancy envelope in this application can be elliptical. Therefore, for the overlapping area of ​​the spatial occupancy envelopes of the planned vehicle trajectory and any predicted target trajectory at the same time step, this application can use an elliptical intersection area analytical algorithm or a rasterization approximation method to calculate it, ensuring that the calculation effect of the overlapping area can meet the real-time requirements of vehicle trajectory collision analysis.

[0071] Then, by using the actual areas of the two spatial occupancy envelopes of the planned vehicle trajectory and any target predicted trajectory at the same time step, as well as the overlapping area of ​​these two spatial occupancy envelopes, the total area of ​​the two spatial occupancy envelopes can be determined. Furthermore, by calculating the ratio between the overlapping area of ​​the two spatial occupancy envelopes of the planned vehicle trajectory and any target predicted trajectory at the same time step and the total area of ​​the two spatial occupancy envelopes, the envelope overlap degree of the planned vehicle trajectory and the target predicted trajectory at that time step can be obtained.

[0072] Therefore, following the same method described above, for each target predicted trajectory, the envelope overlap between the vehicle's planned trajectory and the target predicted trajectory at each time step can be determined, so as to analyze the overall collision situation between the vehicle's planned trajectory and the target predicted trajectory in subsequent analysis.

[0073] S250 determines the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory based on the prediction confidence of the target predicted trajectory, the preset time decay coefficient, and the envelope overlap between the self-planned trajectory and the target predicted trajectory at each time step.

[0074] Considering that the trajectory prediction function in the vehicle-configured autonomous driving system typically outputs multiple predicted trajectories for each target, and only one of these multiple predicted trajectories for the same target is a reference to the actual trajectory, this application, when outputting multiple predicted trajectories for each target through the trajectory prediction function of the vehicle-configured autonomous driving system, also outputs a prediction probability for each predicted trajectory and uses this probability as the prediction confidence level of that predicted trajectory, representing the degree of certainty that the target will follow that predicted trajectory in the future.

[0075] Furthermore, considering that this application performs a trajectory prediction process for each target around the vehicle at every moment, the predicted trajectory changes in real time. Therefore, the impact of the collision risk in the predicted trajectory further into the future on the final collision analysis result decreases. Thus, to accurately analyze the collision risk of the vehicle trajectory, this application pre-sets a time decay coefficient, the value of which continuously decreases over time. This coefficient enhances the influence of near-term collision risk in the predicted trajectory on the final collision analysis result and reduces the influence of long-term collision risk in the predicted trajectory on the final collision analysis result.

[0076] Therefore, for each predicted target trajectory, this application can analyze the envelope overlap between the planned vehicle trajectory and the predicted target trajectory at each time step to determine the probability of a collision between them at each time step. Then, using the prediction confidence of the predicted target trajectory and a preset time decay coefficient, a comprehensive analysis of the probability of a collision between the planned vehicle trajectory and the predicted target trajectory at each time step is performed to determine the overall probability of a collision between them in the future, thereby determining the single-branch collision probability of the planned vehicle trajectory relative to the predicted target trajectory.

[0077] Following the same method described above, the single-branch collision probability of the vehicle's planned trajectory relative to each target predicted trajectory can be determined.

[0078] In some implementations, the single-branch collision probability of the planned vehicle trajectory relative to the target predicted trajectory can be determined by the following methods: determining the instantaneous collision probability of the planned vehicle trajectory relative to the target predicted trajectory at each time step based on the prediction confidence of the target predicted trajectory and the envelope overlap between the planned vehicle trajectory and the target predicted trajectory at each time step; and determining the single-branch collision probability of the planned vehicle trajectory relative to the target predicted trajectory based on a preset time decay coefficient and the instantaneous collision probability of the planned vehicle trajectory relative to the target predicted trajectory at each time step.

[0079] Specifically, after obtaining the envelope overlap between the planned trajectory of the vehicle and the predicted trajectory of any target at each time step, this application can use the prediction confidence of this predicted trajectory to normalize the envelope overlap between the planned trajectory of the vehicle and the predicted trajectory of the target at each time step, thereby obtaining the instantaneous collision probability of the planned trajectory of the vehicle relative to this predicted trajectory of the target at each time step.

[0080] For example, the instantaneous collision probability of the vehicle's planned trajectory relative to the predicted trajectory at each time step can be calculated using the following formula:

[0081] in, Let be the instantaneous collision probability of the planned trajectory of the vehicle relative to the predicted trajectory of the i-th target at any time step t. Let be the prediction confidence of the i-th target trajectory.

[0082] The overlapping area of ​​the spatial occupancy envelopes of the planned trajectory of the vehicle and the predicted trajectory of the i-th target at any time step t.

[0083] The actual area of ​​the spatial occupancy envelope of the planned trajectory of the vehicle at any time step t.

[0084] Let be the actual area of ​​the spatial occupancy envelope of the predicted trajectory of the i-th target at any time step t.

[0085] Then, after obtaining the instantaneous collision probability of the self-planned trajectory relative to the target predicted trajectory at each time step, this application can use a preset time decay coefficient to perform time integration on the instantaneous collision probability of the self-planned trajectory relative to the target predicted trajectory at each time step to determine the overall probability of the self-planned trajectory and the target predicted trajectory colliding in the future, thereby obtaining the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory.

[0086] For example, the single-branch collision probability of the vehicle's planned trajectory relative to any target's predicted trajectory can be calculated using the following formula:

[0087] in, The single-branch collision probability of the planned trajectory of the vehicle relative to the predicted trajectory of the i-th target.

[0088] The preset time decay coefficient, It will decrease over time.

[0089] S260, based on the single-branch collision probability of the self-planned trajectory relative to the predicted trajectories of each target, determine the fusion collision probability of the self-planned trajectory.

[0090] The technical solution provided in this application acquires the planned trajectory of the vehicle and at least one predicted target trajectory in real time. For each trajectory, based on the trajectory point parameters at each time step, the spatial occupancy envelope of that trajectory at each time step is determined. This constructs a dynamically changing spatiotemporal safety boundary for the planned trajectory and the predicted target trajectory, more accurately representing their uncertainties. Then, by analyzing the overlap of the spatial occupancy envelopes of the planned trajectory and any predicted target trajectory at each time step, the single-branch collision probability of the planned trajectory relative to each predicted target trajectory is determined. This determines the fusion collision probability of the planned trajectory, enabling a comprehensive and effective analysis of vehicle trajectory collision risk. This further enhances the versatility of vehicle trajectory collision analysis across various road scenarios and optimizes instantaneous collision analysis into cumulative collision analysis over a continuous future time domain. This provides a more accurate and reliable basis for collision analysis and sufficient collision warning time, significantly improving the foresight and robustness of vehicle trajectory collision analysis and ensuring vehicle driving safety.

[0091] Figure 3 This is a schematic block diagram of a collision analysis device for vehicle trajectory provided in an embodiment of this application. Figure 3 As shown, the device 300 may include: The trajectory acquisition module 310 is used to acquire the planned trajectory of the vehicle and at least one predicted trajectory of a target. Envelope determination module 320 is used to determine the spatial occupancy envelope of each trajectory at each time step for each trajectory in the planned trajectory of the vehicle and the predicted trajectory of the target, based on the trajectory point parameters at each time step of the trajectory. The first collision analysis module 330 is used to determine the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory based on the overlap of the spatial occupancy envelope of the self-planned trajectory and the target predicted trajectory at each time step. The second collision analysis module 340 is used to determine the fusion collision probability of the self-planned trajectory based on the single-branch collision probability of the self-planned trajectory relative to each of the target predicted trajectories.

[0092] In some implementations, the envelope determination module 320 can be specifically used for: Based on the physical dimensions of the trajectory object and the heading angle at each time step in the trajectory, the envelope reference size at each time step of the trajectory is determined, wherein the trajectory object is a vehicle or a target. Based on the envelope reference size and multidimensional driving state quantity at each time step in the trajectory, the spatial occupancy envelope of the trajectory at each time step is determined.

[0093] In some implementations, the multidimensional driving state quantities include at least the trajectory point position, velocity, acceleration, and heading angle.

[0094] In some implementations, the first collision analysis module 330 may include: The envelope overlap determination unit is used to determine the envelope overlap of the planned vehicle trajectory and the predicted target trajectory at the same time step based on the overlapping area and total envelope area of ​​the spatial occupancy envelopes of the planned vehicle trajectory and the predicted target trajectory at the same time step. The first collision analysis unit is used to determine the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory based on the prediction confidence of the target predicted trajectory, the preset time decay coefficient, and the envelope overlap between the self-planned trajectory and the target predicted trajectory at each time step.

[0095] In some implementations, the first collision analysis unit can be specifically used for: Based on the prediction confidence of the target predicted trajectory and the envelope overlap between the vehicle planned trajectory and the target predicted trajectory at each time step, the instantaneous collision probability of the vehicle planned trajectory relative to the target predicted trajectory at each time step is determined. Based on the preset time decay coefficient and the instantaneous collision probability of the planned trajectory of the vehicle relative to the predicted trajectory of the target at each time step, the single-branch collision probability of the planned trajectory of the vehicle relative to the predicted trajectory of the target is determined.

[0096] In some implementations, the second collision analysis module 340 can be specifically used for: Based on the single-branch collision probability of the planned trajectory of the vehicle relative to each of the predicted trajectories of the targets, the individual collision probability of the planned trajectory of the vehicle relative to each target is determined. The fused collision probability of the planned vehicle trajectory is determined based on the collision confidence of each target and the individual collision probability of the planned trajectory relative to each target; or, The fused collision probability of the vehicle's planned trajectory is obtained by performing a union operation on the individual collision probabilities of the vehicle's planned trajectory relative to each target.

[0097] In some implementations, the planned vehicle trajectory includes at least one candidate trajectory planned for the vehicle, and the collision analysis device 300 for the vehicle trajectory may further include a vehicle trajectory determination module. This vehicle trajectory determination module can be used for: The corresponding vehicle trajectory is determined based on the fusion collision probability of each candidate trajectory.

[0098] In this embodiment, the planned trajectory of the vehicle and at least one predicted target trajectory are acquired in real time. For each trajectory, the spatial occupancy envelope at each time step is determined based on the trajectory point parameters at each time step. This constructs a dynamically changing spatiotemporal safety boundary for the planned trajectory and the predicted target trajectory, more accurately representing their uncertainties. Then, by analyzing the overlap of the spatial occupancy envelopes of the planned trajectory and any predicted target trajectory at each time step, the single-branch collision probability of the planned trajectory relative to each predicted target trajectory is determined. This determines the fusion collision probability of the planned trajectory, enabling a comprehensive and effective analysis of vehicle trajectory collision risk. This further enhances the versatility of vehicle trajectory collision analysis across various road scenarios and optimizes instantaneous collision analysis into cumulative collision analysis over a continuous future time domain. This provides a more accurate and reliable basis for collision analysis and sufficient collision warning time, significantly improving the foresight and robustness of vehicle trajectory collision analysis and ensuring vehicle driving safety.

[0099] It should be understood that the apparatus embodiments and method embodiments can correspond to each other, and similar descriptions can be found in the method embodiments. To avoid repetition, further details are omitted here. Specifically, Figure 3 The apparatus 300 shown can execute any of the method embodiments provided in this application, and the foregoing and other operations and / or functions of each module in the apparatus 300 are respectively for implementing the corresponding processes in the various methods of the embodiments of this application. For the sake of brevity, they will not be described in detail here.

[0100] The apparatus 300 of this application embodiment has been described above from the perspective of functional modules in conjunction with the accompanying drawings. It should be understood that this functional module can be implemented in hardware, in software instructions, or in a combination of hardware and software modules. Specifically, the steps of the method embodiments in this application can be completed by integrated logic circuits in the processor's hardware and / or by software instructions. The steps of the method disclosed in this application embodiment can be directly embodied as being executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. Optionally, the software module can be located in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps in the above method embodiments.

[0101] Figure 4 A schematic block diagram of a vehicle provided in an embodiment of this application.

[0102] like Figure 4As shown, the vehicle 400 may include: The system includes a memory 410 and a processor 420. The memory 410 stores computer programs and transfers the program code to the processor 420. In other words, the processor 420 can retrieve and run the computer program from the memory 410 to implement the methods described in the embodiments of this application.

[0103] For example, the processor 420 can be used to execute the above-described method embodiments according to instructions in the computer program.

[0104] In some embodiments of this application, the processor 420 may include, but is not limited to: General-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0105] In some embodiments of this application, the memory 410 includes, but is not limited to: Volatile memory and / or non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0106] In some embodiments of this application, the computer program may be divided into one or more modules, which are stored in the memory 410 and executed by the processor 420 to perform the method provided in this application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the vehicle.

[0107] like Figure 4 As shown, the vehicle may also include: Transceiver 430, which can be connected to processor 420 or memory 410.

[0108] The processor 420 can control the transceiver 430 to communicate with other devices; specifically, it can send information or data to other devices or receive information or data sent by other devices. The transceiver 430 may include a transmitter and a receiver. The transceiver 430 may further include antennas, and the number of antennas may be one or more.

[0109] It should be understood that the various components in the vehicle are connected through a bus system, which includes a data bus, a power bus, a control bus, and a status signal bus.

[0110] This application also provides a computer storage medium storing a computer program thereon, which, when executed by a computer, enables the computer to perform the methods of the above-described method embodiments. Alternatively, this application also provides a computer program product containing instructions that, when executed by a computer, cause the computer to perform the methods of the above-described method embodiments.

[0111] When implemented using software, it can be implemented entirely or partially as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0112] 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.

[0113] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0114] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. For example, the functional modules in the various embodiments of this application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0115] The above are merely specific embodiments 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.

Claims

1. A collision analysis method for vehicle trajectories, characterized in that, include: Obtain the planned trajectory of the vehicle and at least one predicted trajectory of the target; For each trajectory in the planned trajectory of the autonomous vehicle and the predicted trajectory of the target, the spatial occupancy envelope of the trajectory at each time step is determined based on the trajectory point parameters at each time step. Based on the overlap of the spatial occupancy envelopes of the planned vehicle trajectory and the predicted target trajectory at each time step, the single-branch collision probability of the planned vehicle trajectory relative to the predicted target trajectory is determined. The fusion collision probability of the planned vehicle trajectory is determined based on the single-branch collision probability of the planned vehicle trajectory relative to each of the predicted target trajectories.

2. The method according to claim 1, characterized in that, The step of determining the spatial occupancy envelope of the trajectory at each time step based on the trajectory point parameters at each time step includes: Based on the physical dimensions of the trajectory object and the heading angle at each time step in the trajectory, the envelope reference size at each time step of the trajectory is determined, wherein the trajectory object is a vehicle or a target. Based on the envelope reference size and multidimensional driving state quantity at each time step in the trajectory, the spatial occupancy envelope of the trajectory at each time step is determined.

3. The method according to claim 2, characterized in that, The multidimensional driving state quantities include at least the trajectory point position, speed, acceleration, and heading angle.

4. The method according to claim 1, characterized in that, The step of determining the single-branch collision probability of the planned vehicle trajectory relative to the predicted target trajectory based on the overlap of the spatial occupancy envelopes of the planned vehicle trajectory and the predicted target trajectory at each time step includes: Based on the overlapping area and total area of ​​the spatial occupancy envelopes of the planned vehicle trajectory and the predicted target trajectory at the same time step, the envelope overlap degree of the planned vehicle trajectory and the predicted target trajectory at that time step is determined. Based on the prediction confidence of the target predicted trajectory, the preset time decay coefficient, and the envelope overlap between the vehicle's planned trajectory and the target predicted trajectory at each time step, the single-branch collision probability of the vehicle's planned trajectory relative to the target predicted trajectory is determined.

5. The method according to claim 4, characterized in that, The step of determining the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory based on the prediction confidence of the target predicted trajectory, a preset time decay coefficient, and the envelope overlap between the self-planned trajectory and the target predicted trajectory at each time step includes: Based on the prediction confidence of the target predicted trajectory and the envelope overlap between the vehicle planned trajectory and the target predicted trajectory at each time step, the instantaneous collision probability of the vehicle planned trajectory relative to the target predicted trajectory at each time step is determined. Based on the preset time decay coefficient and the instantaneous collision probability of the planned trajectory of the vehicle relative to the predicted trajectory of the target at each time step, the single-branch collision probability of the planned trajectory of the vehicle relative to the predicted trajectory of the target is determined.

6. The method according to claim 1, characterized in that, The step of determining the fusion collision probability of the self-planned trajectory based on the single-branch collision probability of the self-planned trajectory relative to each of the predicted target trajectories includes: Based on the single-branch collision probability of the planned trajectory of the vehicle relative to each of the predicted trajectories of the targets, the individual collision probability of the planned trajectory of the vehicle relative to each target is determined. The fused collision probability of the planned vehicle trajectory is determined based on the collision confidence of each target and the individual collision probability of the planned trajectory relative to each target; or, The fused collision probability of the vehicle's planned trajectory is obtained by performing a union operation on the individual collision probabilities of the vehicle's planned trajectory relative to each target.

7. The method according to claim 1, characterized in that, The vehicle-planned trajectory includes at least one candidate trajectory planned for the vehicle, and the method further includes: The corresponding vehicle trajectory is determined based on the fusion collision probability of each candidate trajectory.

8. A collision analysis device for vehicle trajectory, characterized in that, include: The trajectory acquisition module is used to acquire the planned trajectory of the vehicle and at least one predicted trajectory of the target. The envelope determination module is used to determine the spatial occupancy envelope of each trajectory in the planned trajectory of the vehicle and the predicted trajectory of the target, based on the trajectory point parameters at each time step of the trajectory. The first collision analysis module is used to determine the single-branch collision probability of the self-planned trajectory relative to the target predicted trajectory based on the overlap of the spatial occupancy envelope of the self-planned trajectory and the target predicted trajectory at each time step. The second collision analysis module is used to determine the fusion collision probability of the self-planned trajectory based on the single-branch collision probability of the self-planned trajectory relative to each of the predicted trajectory of the target.

9. A vehicle, characterized in that, include: A processor and a memory, the memory for storing a computer program, the processor for calling and running the computer program stored in the memory to perform the collision analysis method for vehicle trajectory according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instruction is executed by the processor, it implements the collision analysis method for the vehicle trajectory as described in any one of claims 1-7.

Images