Method for determining plugged vehicle, vehicle, processor and program product

Abstract

The invention discloses a method for determining a plugged vehicle, a vehicle, a processor and a program product. The method comprises the steps that in the current driving process of a vehicle, multiple pieces of sensing data collected by multiple sensors in the vehicle are obtained; the multiple pieces of sensing data are predicted, target driving data of the to-be-plugged vehicle in the future driving process are obtained, and the target driving data are used for representing the target driving state of the to-be-plugged vehicle in the future driving process; based on the target driving data, a target plugged vehicle is determined from the at least one to-be-plugged vehicle, and the cut-in success rate of the vehicle cutting into the lane where the target plugged vehicle is located is larger than the cut-in success rate of the to-be-plugged vehicles except the target plugged vehicle in the at least one to-be-plugged vehicle. The method and the device solve the technical problem of low accuracy of determining the plugged vehicle.

Description

Technical Field

[0001] This application relates to the field of vehicle control, and more specifically, to a method for determining vehicles that cut in line, a vehicle, a processor, and a program product. Background Technology

[0002] Currently, driver assistance systems typically rely on sensors such as cameras, millimeter-wave radar, and lidar to automatically execute a lane change maneuver after determining that there is sufficient safe space in the target lane. However, this method can only provide warnings of dangerous vehicles and cannot assist drivers in making decisions in complex dynamic situations, resulting in a technical problem of low accuracy in identifying vehicles cutting in front of them.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This application provides a method, vehicle, processor, and program product for determining vehicles that cut in line, in order to at least solve the technical problem of low accuracy in determining vehicles that cut in line.

[0005] According to one aspect of the embodiments of this application, a method for determining a vehicle that is about to cut in line is provided. The method may include: acquiring multiple perception data collected by multiple sensors in the vehicle during its current driving process, wherein the perception data characterizes the initial driving state of at least one vehicle that is about to cut in line during the current driving process, and the at least one vehicle is located behind the vehicle; predicting the multiple perception data to obtain target driving data of the vehicle that is about to cut in line during its future driving process, wherein the target driving data characterizes the target driving state of the vehicle that is about to cut in line during its future driving process; and determining a target vehicle that is about to cut in line from among the at least one vehicle that is about to cut in line, wherein the success rate of a vehicle cutting into the lane occupied by the target vehicle is greater than the success rate of other vehicles that are about to cut in line from among the at least one vehicle that are about to cut in line.

[0006] Optionally, based on target driving data, determining a target vehicle to cut in from at least one vehicle waiting to cut in includes: determining the relative distance between the vehicles waiting to cut in during future driving based on the target driving data; identifying the vehicle waiting to cut in as an initial vehicle waiting to cut in in response to the relative distance being greater than a safe distance threshold; determining the cutting-in friendliness of the initial vehicle waiting to cut in, wherein the cutting-in friendliness is related to the success rate of cutting in; and using the cutting-in friendliness to filter out the target vehicle to cut in from at least one initial vehicle waiting to cut in.

[0007] Optionally, determining the line-cutting friendliness of the initial vehicle to be cut in includes: converting the target driving data of the initial vehicle to be cut in to obtain the relative speed between the initial vehicle to be cut in and other vehicles; obtaining a first friendliness factor corresponding to the initial vehicle to be cut in based on the relative distance, relative speed and target driving data, wherein the first friendliness factor is associated with the target movement state of the initial vehicle to be cut in; and determining the line-cutting friendliness based on the first friendliness factor.

[0008] Optionally, determining the lane-cutting friendliness based on a first friendliness factor includes: in response to detecting the driving intention data of the initial vehicle waiting to cut in, determining the yielding willingness matching the driving intention data; determining a second friendliness factor corresponding to the initial vehicle waiting to cut in based on the yielding willingness, target driving data, and relative distance, wherein the second friendliness factor is used to characterize the yielding intention of the initial vehicle waiting to cut in; and performing weighted processing on the first friendliness factor and the second friendliness factor to obtain the lane-cutting friendliness.

[0009] Optionally, determining the lane-jumping friendliness based on the first friendliness factor includes: in response to the absence of detected driving intention data, determining the second friendliness factor corresponding to the initial lane-jumping vehicle based on the target driving data of the initial lane-jumping vehicle and the relative distance; and weighting the first friendliness factor and the second friendliness factor to obtain the lane-jumping friendliness.

[0010] Optionally, the target vehicle for cutting in line is selected from at least one initial vehicle waiting to cut in line using the vehicle's friendliness in cutting in line. This includes: selecting at least one initial vehicle waiting to cut in line using the vehicle's friendliness in cutting in line and a friendliness threshold; and identifying the initial vehicle waiting to cut in line with the highest friendliness in cutting in line among the selected vehicles as the target vehicle for cutting in line.

[0011] Optionally, the method further includes: controlling the display module in the vehicle to display the target vehicle that is cutting in line according to display requirements, and outputting a prompt message, wherein the prompt message is used to notify the vehicle of the target vehicle that is about to cut in line.

[0012] According to one aspect of the embodiments of this application, a device for determining a vehicle that is about to cut in line is provided. The device may include: an acquisition module, configured to acquire multiple perception data collected by multiple sensors in the vehicle during the current driving process of the vehicle, wherein the perception data is used to characterize the initial driving state of at least one vehicle that is about to cut in line during the current driving process, and the at least one vehicle that is about to cut in line is located behind the vehicle; a prediction module, configured to predict the multiple perception data to obtain target driving data of the vehicle that is about to cut in line during its future driving process, wherein the target driving data is used to characterize the target driving state of the vehicle that is about to cut in line during its future driving process; and a determination module, configured to determine a target vehicle that is about to cut in line from the at least one vehicle that is about to cut in line based on the target driving data, wherein the success rate of a vehicle cutting into the lane where the target vehicle is located is greater than the success rate of other vehicles that are about to cut in line from the at least one vehicle that is about to cut in line, excluding the target vehicle.

[0013] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to execute the method for determining vehicles cutting in line according to the embodiments of this application.

[0014] According to another aspect of the embodiments of this application, a processor is also provided for running a program, wherein the program is executed by the processor to perform the method for determining vehicles cutting in line according to the embodiments of this application.

[0015] According to another aspect of the embodiments of this application, a program product is also provided, the program product including computer instructions, wherein when the computer instructions are executed by a processor, they implement the method for determining vehicles cutting in line according to the embodiments of this application.

[0016] According to another aspect of the embodiments of this application, a vehicle is also provided, which can be used to perform the method for determining vehicles cutting in line according to the embodiments of this application.

[0017] In this embodiment, during the vehicle's current driving process, multiple perception data collected by multiple sensors in the vehicle are acquired. These perception data characterize the initial driving state of at least one vehicle vying to cut in, located behind the vehicle. The multiple perception data are then used to predict the target driving data of the vehicle vying to cut in during its future driving process. This target driving data characterizes the target driving state of the vehicle vying to cut in during its future driving process. Based on the target driving data, a target vehicle vying to cut in is determined from among the at least one vehicle vying to cut in. The success rate of a vehicle cutting into the lane of the target vehicle vying to cut in is greater than the success rate of other vehicles vying to cut in, excluding the target vehicle. In other words, this embodiment acquires the target driving data of the vehicles vying to cut in, determines at least one vehicle vying to cut in and its corresponding success rate based on the target driving data, and determines the target vehicle vying to cut in from among the at least one vehicle vying to cut in based on the success rate. This improves the accuracy of determining vehicles vying to cut in and solves the technical problem of low accuracy in determining vehicles vying to cut in. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0019] Figure 1 This is a flowchart of a method for determining vehicles that cut in line according to an embodiment of this application;

[0020] Figure 2 This is a flowchart of an optional method for determining vehicles that cut in line according to an embodiment of this application;

[0021] Figure 3 This is a schematic diagram of an optional vehicle cutting in line determination device according to an embodiment of this application. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0023] 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 apparatus 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 apparatus.

[0024] According to an embodiment of this application, an embodiment of a method for determining vehicles cutting in line is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0025] Figure 1 This is a flowchart of a method for determining vehicles cutting in line according to an embodiment of this application, as shown below. Figure 1 As shown, the method may include the following steps.

[0026] Step S102: During the current driving process of the vehicle, multiple perception data collected by multiple sensors in the vehicle are acquired. The perception data is used to characterize the initial driving state of at least one vehicle waiting to cut in during the current driving process. The at least one vehicle waiting to cut in is located behind the vehicle.

[0027] In the technical solution provided in step S102 of this application, the aforementioned perception data can be used to represent the initial driving state of at least one vehicle waiting to cut in line during its current driving process, such as the position information and speed information of the vehicle waiting to cut in line. The vehicle waiting to cut in line can be used to represent a vehicle located behind the vehicle, traveling in the same direction as the vehicle, and whose driver or autonomous driving system is considering performing a lane change operation in front of it. The aforementioned sensors can include different types of sensors deployed in the vehicle, including but not limited to: vehicle cameras, radar, etc.

[0028] Optionally, multiple perception data collected by multiple sensors in the vehicle can be acquired, including but not limited to images of vehicles moving behind the vehicle (i.e., vehicles waiting to cut in line) captured by an onboard camera. Based on these images, at least one vehicle waiting to cut in line can be identified and tracked using computer vision algorithms, and its position, speed, direction, type, and size can be determined. The relative position and speed information of vehicles behind, as well as changes in their acceleration, can be obtained using onboard radar (including millimeter-wave radar, ultrasonic radar, etc.).

[0029] Optionally, in addition to sensing information about vehicles waiting to cut in front of the vehicle, the aforementioned sensing data can also be used to improve vehicle safety during driving, including sensing the driving status of vehicles in front of the vehicle and vehicles in front of the vehicles waiting to cut in.

[0030] For example, information about traffic participants around a vehicle can be perceived through vehicle cameras, radar, etc. The perceived information from these sensors is then fused using Kalman filtering to output data such as the identification code (ID), position, speed, acceleration, heading angle, and category of each moving vehicle around the vehicle for each frame.

[0031] Through step S102 of this application, perception data of the surrounding traffic environment can be acquired comprehensively through multiple sensors. Based on this perception data, the status of surrounding vehicles can be determined in real time, including but not limited to the position, speed, acceleration, and heading angle of the surrounding vehicles.

[0032] Step S104: Predict multiple perception data to obtain the target driving data of the vehicle waiting to cut in line during its future driving process. The target driving data is used to characterize the target driving state of the vehicle waiting to cut in line during its future driving process.

[0033] In the technical solution provided by step S104 of this application, the target driving data can be used to represent the target driving state of the vehicle waiting to cut in line during future driving. It can be the acceleration value predicted in the future time period. For example, it can be the longitudinal acceleration prediction value at a time step of 0.5 seconds within the next 5 seconds, which can be represented by a1.

[0034] Optionally, during the current driving process of the vehicle, after acquiring multiple perception data collected by multiple sensors in the vehicle, the multiple perception data can be used to predict the target driving data of the vehicle to be cut in line during its future driving process.

[0035] Optionally, predictions based on multiple sensing data may include, but are not limited to: using a pre-trained deep learning model, such as a Long Short-Term Memory (LSTM) network or other time series prediction models, inputting the sensing data obtained above, to predict parameters such as the position, speed, acceleration, and heading angle of vehicles waiting to cut in line in the next few seconds.

[0036] For example, predicting the behavior of a vehicle attempting to cut in line can include: acquiring multiple perception data points, fusing these perception data points using a Kalman filter, and inputting the fused perception data into a bidirectional LSTM neural network to predict the current state of the vehicle attempting to cut in line. The training data for the bidirectional LSTM neural network can come from real-world driving datasets, such as the NuScenes Multimodal Autonomous Driving Dataset (nuScenes) or the High Density Trajectories for TrafficScene Understanding (HD-TDS). This bidirectional LSTM neural network can predict target driving data, for example, outputting a predicted longitudinal acceleration value a1=a1 / a2 at 0.5-second time steps over the next 5 seconds. (t+Δt).

[0037] Through step S104 of this application, visual information captured by a camera, distance and speed data acquired by millimeter-wave radar or lidar, and GPS positioning information can be obtained. A bidirectional LSTM neural network can be used to predict the driving status of vehicles waiting to cut in line over a future period of time.

[0038] Step S106: Based on the target driving data, determine the target lane-jumping vehicle from at least one waiting lane-jumping vehicle, wherein the success rate of a vehicle cutting into the lane where the target lane-jumping vehicle is located is greater than the success rate of other waiting lane-jumping vehicles excluding the target lane-jumping vehicle.

[0039] In the technical solution provided by step S106 of this application, the above-mentioned entry success rate can be used to represent the probability that the driver successfully merges into the lane where the vehicle waiting to cut in is located.

[0040] Optionally, after predicting multiple perception data to obtain the target driving data of the vehicle waiting to cut in line during its future driving process, the target vehicle to cut in line can be determined from at least one vehicle waiting to cut in line based on the target driving data.

[0041] Optionally, for each vehicle waiting to cut in line, an evaluation model for the success rate of cutting in can be established by using previously predicted target driving data (such as expected position, expected speed, expected acceleration, expected heading angle, etc.) combined with the vehicle's current state and road environment.

[0042] Optionally, the target driving data is input into the cut-in success rate evaluation model to calculate the cut-in success rate of each vehicle waiting to cut in. The cut-in success rates of each vehicle waiting to cut in are compared, and the vehicle with the highest success rate is selected as the target vehicle to cut in.

[0043] Through step S106 of this application, based on the analysis of target driving data and the evaluation of the lane-changing success rate, corresponding lane-changing strategies can be provided to the driver. Addressing the problems of low recognition efficiency, conservative lane-changing decisions, inability to understand other vehicles, and inability to identify whether the driver exhibits a potential yielding intention in the face of dense traffic flow, the aforementioned step S106 provides a method that can actively identify and recommend the "most likely to yield" vehicle to the driver. This method inputs target driving data into a lane-changing success rate evaluation model to obtain the lane-changing success rate of each vehicle waiting to cut in, and compares these success rates to determine the target vehicle for lane-changing. The lane-changing success rate evaluation model is based on the target driving data of the vehicle waiting to cut in during its future driving process, such as driving speed and acceleration, as well as the vehicle's current state and road environment, thereby assisting the driver in completing the lane-changing operation efficiently and safely.

[0044] In this embodiment, after identifying the target vehicle that is cutting in, it can be highlighted on the vehicle's display interface. If the vehicle is in autonomous driving mode, it can automatically move into the lane occupied by the target vehicle and move in front of it. If the vehicle is not in autonomous driving mode, it can obtain the user's selection command to determine whether to move into the lane occupied by the target vehicle. If the user selects to move into the target vehicle's lane, the vehicle's autonomous driving mode can be activated, and it can automatically move into the target vehicle's lane. Alternatively, the target vehicle can be highlighted on the vehicle's display interface to alert the user to the vehicle that is cutting in, and the user can then manually move into the target vehicle's lane according to their needs.

[0045] Optionally, the lane where the target lane-changing vehicle is located can be used to indicate the specific target location where the vehicle is performing the lane-changing operation.

[0046] Optionally, based on the target driving data, after determining the target vehicle to cut in from at least one vehicle waiting to cut in, the vehicle can be controlled to switch into the lane where the target vehicle is located, thereby controlling the switched vehicle to drive in front of the target vehicle.

[0047] Optionally, based on target driving data and the current vehicle status, a path that avoids vehicles waiting to cut in, pedestrians, and obstacles can be identified in real time. The vehicle's Electronic Stability Program (ESP), Lane Keeping Assist System (LKAS), or other yaw control systems can be used to automatically adjust the steering angle, guiding the vehicle smoothly into the target lane and maintaining its correct position within the lane. Adaptive Cruise Control (ACC) or similar functions can be used to automatically adjust the vehicle speed, ensuring the vehicle can complete lane changes at an appropriate speed and find a suitable position before the target vehicle cutting in.

[0048] Through steps S102 to S106 of this application, during the current driving process of the vehicle, multiple perception data collected by multiple sensors in the vehicle are acquired. The perception data is used to characterize the initial driving state of at least one vehicle waiting to cut in during the current driving process, and the at least one vehicle waiting to cut in is located behind the vehicle. The multiple perception data are predicted to obtain the target driving data of the vehicle waiting to cut in during the future driving process. The target driving data is used to characterize the target driving state of the vehicle waiting to cut in during the future driving process. Based on the target driving data, a target vehicle is determined from the at least one vehicle waiting to cut in, wherein the success rate of the vehicle cutting into the lane where the target vehicle is located is greater than the success rate of the other vehicles waiting to cut in, excluding the target vehicle. In other words, in this embodiment of the application, the target driving data of the vehicle to be cut in line is obtained, at least one vehicle to be cut in line and the corresponding cut-in success rate of the vehicle to be cut in line are determined based on the target driving data, and the target vehicle to be cut in line is determined from the at least one vehicle to be cut in line based on the cut-in success rate, thereby achieving the technical effect of improving the accuracy of determining the vehicle to be cut in line, and thus solving the technical problem of low accuracy in determining the vehicle to be cut in line.

[0049] The method described in this embodiment will be further described below.

[0050] As an optional implementation, determining a target vehicle to cut in line from at least one vehicle waiting to cut in line, based on target driving data, includes: determining the relative distance between the vehicles waiting to cut in line during future driving based on the target driving data; determining the initial vehicle waiting to cut in line as the vehicle waiting to cut in line when the relative distance is greater than a safe distance threshold; determining the cutting-in friendliness of the initial vehicle waiting to cut in line, wherein the cutting-in friendliness is related to the success rate of cutting in; and using the cutting-in friendliness to filter out the target vehicle to cut in line from at least one initial vehicle waiting to cut in line.

[0051] In this embodiment of the application, the aforementioned relative distance can be used to represent the spatial interval between the vehicle and the vehicle waiting to cut in. The aforementioned safe distance threshold can be used to represent the minimum distance required to ensure that a collision does not occur between the vehicle and the vehicle waiting to cut in at different vehicle speeds. The aforementioned cut-in friendliness can be used to represent the probability that a vehicle will cut in front of a vehicle waiting to cut in. A higher success rate of the vehicle waiting to cut in corresponds to a higher cut-in friendliness.

[0052] Optionally, by using a predictive model and combining the vehicle's current speed, acceleration, and the speed change trend of the target vehicle, the future relative position change can be estimated, including but not limited to: calculating the relative distance between the vehicle and each vehicle waiting to cut in at each moment, which can be represented by ΔS.

[0053] Optionally, different safety distance thresholds can be set for different speed ranges to ensure sufficient safety buffer under various conditions. The predicted future relative distance ΔS of each vehicle waiting to cut in line can be compared with the safety distance threshold for the corresponding speed. If ΔS is greater than the safety distance threshold, the vehicle waiting to cut in line is identified as an initial vehicle that preliminarily meets the conditions for cutting in line.

[0054] Optionally, for each initial vehicle vying to cut in, the "friendliness" of its ability to cut in is analyzed based on physical conditions such as relative distance, relative speed, and acceleration. These initial vehicles are then ranked according to their friendliness, and a minimum threshold for friendliness is set. Among the ranked vehicles with a friendliness score greater than the minimum threshold, the initial vehicle with the highest friendliness score is selected as the target vehicle for cutting in.

[0055] Optionally, based on the target driving speed (i.e., the acceleration prediction result), the future gap change trend can be derived to obtain the relative distance between vehicles waiting to cut in: ΔS=ΔV t1+ΔV 2 / 2a1. Where ΔV represents the speed difference between the vehicle and the vehicle waiting to cut in, ΔV = Vvehicle - Vwaiting-to-cut vehicle; a1 represents the predicted acceleration of the vehicle waiting to cut in, a1 = a0... (t+Δt), where t1 can represent the time change, t1=t+Δt.

[0056] Furthermore, a safe distance threshold can be used to judge the relative distance. If the relative distance is greater than the safe distance threshold, the vehicle waiting to cut in line can be identified as an initial vehicle waiting to cut in line. Using this method, at least one initial vehicle waiting to cut in line can be identified. After identifying at least one initial vehicle waiting to cut in line, the "friendliness" of cutting in line for each initial vehicle can be determined. Using this friendliness, target vehicles can be selected from the at least one initial vehicle waiting to cut in line.

[0057] Using the above method, the change in relative distance at each moment can be estimated based on the speed difference and relative distance between the vehicle and the vehicle waiting to cut in, the predicted acceleration of the vehicle waiting to cut in, and the future speed change trend. Different speed ranges can be set with corresponding safe distance thresholds to confirm whether the relative distance ΔS between the vehicles waiting to cut in exceeds the safe distance threshold. If the relative distance is greater than the safe distance threshold, the vehicle waiting to cut in is identified as the initial vehicle to cut in. Based on the relative distance, relative speed, acceleration, and other states of each initial vehicle waiting to cut in, the willingness of the initial vehicle to cut in to yield is analyzed and judged to determine its "friendliness" in cutting in. A minimum threshold for friendliness is set, and the initial vehicle with the highest friendliness among vehicles whose friendliness is greater than the minimum threshold is selected as the target vehicle to cut in.

[0058] As an optional implementation, determining the line-cutting friendliness of the initial vehicle to be cut in includes: converting the target driving data of the initial vehicle to be cut in to obtain the relative speed between the initial vehicle to be cut in and other vehicles; obtaining a first friendliness factor corresponding to the initial vehicle to be cut in based on the relative distance, relative speed and target driving data, wherein the first friendliness factor is associated with the target movement state of the initial vehicle to be cut in; and determining the line-cutting friendliness based on the first friendliness factor.

[0059] In this embodiment of the application, the aforementioned relative speed can be used to represent the relative motion rate between the vehicle and the initial vehicle waiting to cut in, and can be represented by ΔV. It can be used to determine the speed difference between the vehicle and the initial vehicle waiting to cut in from the rear. The aforementioned first friendliness factor can be used to represent the feasibility of cutting in based on physical conditions. It can be a physical friendliness factor Fp, which can be calculated based on a physical safety model and can be used to reflect the space and speed conditions that the vehicle objectively provides for cutting in.

[0060] Optionally, based on the target driving data, the predicted position of the initial vehicle waiting to cut in line can be determined at a future time, thus obtaining the position information of the vehicle waiting to cut in line at that future time. The position information of the initial vehicle waiting to cut in line is then transformed into a vehicle-related coordinate system to calculate the relative speed.

[0061] For example, a new coordinate system can be established using the current position of the vehicle as the origin, with the front and left directions as positive. The position of the initial vehicle waiting to cut in can be represented as its coordinates relative to the current vehicle. Based on the changes in these transformed coordinates relative to the vehicle, the relative distance between the current vehicle and the initial vehicle waiting to cut in can be determined. It should be noted that the above method for calculating relative distance is merely an example and does not impose specific limitations on the method used to calculate relative distance.

[0062] Optionally, based on the target driving data, features closely related to the lane-cutting operation are determined, including but not limited to the relative distance ΔS and relative speed ΔV between the initial vehicle to be cut off, and the relative acceleration Δa between the vehicle and the vehicle to be cut off. A first friendliness factor Fp is obtained by combining and calculating the weighted parameters such as the relative distance ΔS, relative speed ΔV, and relative acceleration Δa. The lane-cutting friendliness is then determined based on the first friendliness factor.

[0063] For example, the target driving data of the initial vehicles waiting to cut in line is transformed to obtain relative speed. Physical parameters such as relative distance ΔS, relative speed ΔV, and relative acceleration Δa can then be input into the physical safety model. This physical safety model can calculate the first friendliness factor using the following formula:

[0064] Fp=w1 f(ΔS) + w2 f(ΔV)+w2 f(Δa)

[0065] Where w1 can be the weighting coefficient corresponding to the relative distance ΔS, w2 can be the weighting coefficient corresponding to the relative velocity ΔV, w3 can be the weighting coefficient corresponding to the relative acceleration Δa, and f, g, and h are normalization functions, such as the Sigmoid function. The above normalization functions can map the physical parameters to the interval [0, 1].

[0066] Optionally, for the first friendliness factor Fp, the greater the distance between the vehicle and the vehicle cutting in, the smaller the speed difference (the speed difference may be a positive value), and the negative acceleration (deceleration) of the vehicle cutting in, the higher the Fp score.

[0067] Optionally, the friendliness of cutting in line can be determined based on the first friendliness factor.

[0068] Using the above method, weighting coefficients can be set for relative distance ΔS, relative velocity ΔV, and relative acceleration Δa. The weighted relative distance ΔS, relative velocity ΔV, and relative acceleration Δa are then summed to obtain a first friendliness factor Fp that reflects the physical feasibility of the blocking operation.

[0069] As an optional implementation, determining the lane-cutting friendliness based on a first friendliness factor includes: in response to detecting the driving intention data of the initial vehicle waiting to cut in, determining the yielding willingness matching the driving intention data; determining a second friendliness factor corresponding to the initial vehicle waiting to cut in based on the yielding willingness, target driving data, and relative distance, wherein the second friendliness factor is used to characterize the yielding intention of the initial vehicle waiting to cut in; and performing weighted processing on the first friendliness factor and the second friendliness factor to obtain the lane-cutting friendliness.

[0070] In this embodiment, the aforementioned driving intention data can be determined based on the potential behavioral tendencies of the driver of the initial vehicle waiting to cut in line. It can be used to determine the driver's driving intention and can be determined based on signals such as acceleration, deceleration, maintaining a constant speed, lane-changing intention, and use of turn signal signals of the vehicle waiting to cut in line. These signals can be parsed from V2X messages. The aforementioned yielding willingness can be used to represent the likelihood that the driver of the initial vehicle waiting to cut in line will actively or passively provide space for the vehicle to cut in line. The aforementioned second friendliness factor can be used to represent the feasibility of the cutting-in operation predicted based on the driving intention data and yielding willingness; it can be represented as the yielding friendliness factor, which can be denoted by Fw.

[0071] Optionally, signals received via V2X communication, such as the "yield flag" signal, or signals that detect changes in vehicle acceleration, the frequency and pattern of turn signal usage, and inter-vehicle communication interactions, can be analyzed to determine the potential behavioral tendencies of the target vehicle's driver. For driving intention data contained in the V2X communication, explicit yield signals are directly read. If the target vehicle's "yield flag" is "true," it indicates that the target vehicle's driver has a clear intention to yield.

[0072] Optionally, a fixed high score can be set for the received explicit yield signal P1 (e.g., the "yield flag" is "true" in V2X communication).

[0073] Optionally, when driving intention data can be monitored, the second friendliness factor can be determined based on the displayed signal, target driving data, and relative distance. That is, the "yield intention flag" can be parsed from the V2X message to obtain the driving intention data. If the flag is "true," a high score is assigned. Simultaneously, an implicit behavior P2 can be set to continuously monitor the vehicle's acceleration a. If a vehicle waiting to cut in is detected to have significant deceleration (a < negative threshold), and its relative distance to the vehicle in front (the main vehicle) increases within a safe range, it is determined that the vehicle waiting to cut in has an implicit yielding intention, and implicit behavior P2 is assigned a medium score. If there are no yielding signs, the score is 0. The yielding friendliness factor Fw can be determined using the above method. Here, Fw = P1 + P2. P1 and P2 can be weighted and summed, or averaged, to obtain the yielding friendliness factor, i.e., the second friendliness factor. After obtaining the first friendliness factor and the second friendliness factor, the first friendliness factor and the second friendliness factor can be weighted to obtain the friendliness score.

[0074] Optionally, weighting coefficients α and β are assigned to the first friendliness factor Fp and the second friendliness factor Fw, respectively, where α + β = 1. The first friendliness factor Fp and the second friendliness factor Fw are then weighted and summed according to the pre-set weighting coefficients α and β to obtain the friendliness score Y. The calculation formula is Y = α Fp+β Fw.

[0075] By introducing the first friendliness factor Fp and the second friendliness factor Fw, the decision-making process is no longer limited to physical conditions but can comprehensively consider the interaction between vehicles and the driver's subjective intentions. For example, regarding Fw, the implicit intention of a vehicle waiting to cut in line can be determined by monitoring driver intention data and setting implicit behaviors, thus making the decision more accurate and improving driving safety. By monitoring changes in the acceleration and distance of the target vehicle, potential dangerous situations can be detected and avoided, reducing the risk of safety accidents caused by misjudgment.

[0076] As an optional implementation, determining the lane-jumping friendliness based on a first friendliness factor includes: in response to the absence of detected driving intention data, determining a second friendliness factor corresponding to the initial lane-jumping vehicle based on the target driving data of the initial lane-jumping vehicle and the relative distance; and weighting the first friendliness factor and the second friendliness factor to obtain the lane-jumping friendliness.

[0077] Optionally, in the absence of an explicit yield signal P1, the target vehicle's willingness to yield can be inferred by analyzing implicit behaviors. These implicit behaviors can refer to vehicle acceleration, deceleration, etc. For example, by continuously monitoring the target vehicle's acceleration, if deceleration is observed (especially when the vehicle approaches the target vehicle), it can be determined that the target vehicle may have a tendency to yield.

[0078] Optionally, the deceleration tendency of the target vehicle can be analyzed based on the change in its acceleration *a*. An implicit behavior *P2* and a threshold *a_threshold* are set to continuously monitor the acceleration *a* of the initial vehicle attempting to cut in. If the initial vehicle attempting to cut in shows significant deceleration (*a* < negative threshold *a_threshold*), and the relative distance between the vehicle attempting to cut in and other vehicles increases within a safe range, it is determined that the vehicle attempting to cut in has an implicit intention to yield, and *P2* is assigned a moderate score. If there are no signs of yielding, the score is 0.

[0079] Optionally, if no driving intention data is detected, it can be determined whether the initial vehicle intending to cut in intends to yield based on the target speed and relative distance, thereby determining a second friendliness factor corresponding to the target speed and relative distance. The first and second friendliness factors can be weighted to obtain the cutting-in friendliness score.

[0080] For example, based on the initial vehicle's driving data, including its position, speed, acceleration, and relative distance to other vehicles, we can infer the vehicle's current driving state and potential behavior in the next few seconds. We can continuously monitor the relative distance ΔS between the vehicle and the initial vehicle attempting to cut in. If ΔS shows an increasing trend in a short period, especially when the vehicle is approaching the target vehicle, it may indicate that the target vehicle is decelerating, creating favorable conditions for cutting in. In this case, a relatively high score can be assigned to the second friendliness factor Fw. We can analyze the target vehicle's acceleration change 'a'. If the acceleration 'a' is negative (i.e., the target vehicle is decelerating), even without an explicit yielding signal, it can be assumed that the target vehicle may be willing to yield, and the value of Fw can be increased accordingly.

[0081] Using the methods described above, the ability to intelligently assess the lane-cutting friendliness of a target vehicle can be achieved even in the absence of direct driving intention data. When direct driving intention data is unavailable, the potential behavior of the target vehicle can be indirectly estimated by analyzing target driving data (such as acceleration and relative distance changes), ensuring that lane-cutting assistance suggestions can be provided under various driving conditions.

[0082] As an optional implementation, the target vehicle for cutting in line is selected from at least one initial vehicle waiting to cut in line by utilizing the friendliness of cutting in line, including: using at least one friendliness of cutting in line corresponding to at least one initial vehicle waiting to cut in line and a friendliness threshold to filter at least one initial vehicle waiting to cut in line; and determining the initial vehicle waiting to cut in line with the highest friendliness of cutting in line among the filtered at least one initial vehicle waiting to cut in line as the target vehicle for cutting in line.

[0083] Optionally, all possible initial vehicles waiting to cut in line are identified from the surrounding environment, and a cutting-in friendliness score Y is calculated for each vehicle, wherein the cutting-in friendliness score Y can be obtained by weighted summation of the first friendliness factor Fp and the second friendliness factor Fw.

[0084] Optionally, a minimum cut-in friendliness threshold value Y_threshold is defined to filter out vehicles with too low a score, thus deeming them unsuitable for cut-in operations. The cut-in friendliness score Y of each initial vehicle to be cut in is compared with Y_threshold. Only when Y >= Y_threshold will the initial vehicle to be cut in be included in the candidate list for further analysis.

[0085] Optionally, the selected candidate vehicles can be sorted in descending order of their "line-cutting friendliness" (Y). A higher Y value indicates better conditions and lower risk for lane-changing. The vehicle with the highest "line-cutting friendliness" (Y) can be selected from the sorted candidate list and designated as the target vehicle for lane-cutting in the current scenario.

[0086] Using the above method, intelligent screening can be performed on all possible initial vehicles waiting to cut in line from the surrounding environment, based on the vehicle's "cut-in friendliness" score. First, the "cut-in friendliness" score Y for each vehicle is calculated, using a weighted sum of the physical friendliness factor Fp and the willingness friendliness factor Fw. By comparing the "cut-in friendliness" score Y of each vehicle with Y_threshold, vehicles with scores higher than or equal to the threshold are selected as candidates for further analysis. From the candidate vehicle list, the initial vehicle waiting to cut in line with the highest score Y is selected, and this initial vehicle is determined as the target vehicle for cutting in line.

[0087] As an optional implementation, the method further includes: controlling the display module in the vehicle to display the target vehicle cutting in line according to display requirements, and outputting prompt information, wherein the prompt information is used to prompt the vehicle about the target vehicle cutting in line.

[0088] In this embodiment, the display module can be used to represent various information display devices inside the vehicle, such as a head-up display (HUD), a central touchscreen, and an instrument panel. The aforementioned prompt information can be used to indicate a clear instruction generated by the system for a target vehicle cutting in line.

[0089] Optionally, the vehicle's head-up display can be used to graphically highlight the outline of the target vehicle cutting in line, and the target vehicle can be indicated via voice prompts, text messages, or by pointing an arrow at the target vehicle on the central touchscreen. The instrument panel can display the target vehicle's location and a cutting-in friendliness rating via indicator lights or digital information.

[0090] By using the methods described above, drivers can clearly and intuitively understand the target vehicles recommended by the system and their corresponding lane-cutting friendliness before performing a lane-cutting operation, thus making safer and more confident decisions.

[0091] In this embodiment, during the vehicle's current driving process, multiple perception data collected by multiple sensors in the vehicle are acquired. These perception data characterize the initial driving state of at least one vehicle vying to cut in, located behind the vehicle. The multiple perception data are then used to predict the target driving data of the vehicle vying to cut in during its future driving process. This target driving data characterizes the target driving state of the vehicle vying to cut in during its future driving process. Based on the target driving data, a target vehicle vying to cut in is determined from among the at least one vehicle vying to cut in. The success rate of a vehicle cutting into the lane of the target vehicle vying to cut in is greater than the success rate of other vehicles vying to cut in, excluding the target vehicle. In other words, this embodiment acquires the target driving data of the vehicles vying to cut in, determines at least one vehicle vying to cut in and its corresponding success rate based on the target driving data, and determines the target vehicle vying to cut in from among the at least one vehicle vying to cut in based on the success rate. This improves the accuracy of determining vehicles vying to cut in and solves the technical problem of low accuracy in determining vehicles vying to cut in.

[0092] The technical solutions of the embodiments of this application will be illustrated below with reference to preferred embodiments.

[0093] Currently, the core of related technologies lies in determining whether a safety risk exists, that is, whether a safe physical space exists. This is typically achieved using hard safety thresholds based on kinematic models, such as "Time to Collision" (TTC).

[0094] Current technical solutions suffer from several shortcomings, including low decision-making efficiency and excessive conservatism. For example, existing ALC systems, in order to ensure absolute safety, typically set safety gap thresholds far exceeding the reasonable gaps that human drivers would judge based on experience in daily driving. Under heavy traffic conditions (such as during urban rush hour), the system is almost unable to find gaps that meet the criteria, leading to frequent rejections of lane-change requests or prolonged waiting times, thus negating its auxiliary function. Ultimately, the driver still needs to take over the vehicle and rely on personal driving skills to complete the maneuver, which actually increases the risk.

[0095] The existing technical solutions lack an understanding of the "social attributes" of other drivers. For example, traffic behavior is essentially a complex social game. When a human driver cuts in, they subconsciously observe the behavior of vehicles behind them, such as whether the other driver slightly slows down, flashes their lights, or turns their head to check. These subtle signals are key to judging the other driver's "willingness to yield." The existing technologies treat other vehicles as purely physical obstacles, and the models lack consideration for this "driving tacit understanding" or "game intent," thus failing to predict that the other driver might proactively create space in response to the driver's lane-changing intention.

[0096] The human-computer interaction experience of related technical solutions is poor. For example, existing ALC systems are usually "black box" decisions: either the system successfully changes lanes, or it doesn't work at all. It cannot tell the driver "which car behind might be a lane-changing opportunity." This "one-size-fits-all" interaction method deprives the driver of the final decision-making power and fails to help the driver improve their perception ability in complex environments.

[0097] The root of these problems lies in the fact that related technologies typically "find physical safety gaps" rather than "predict and exploit cooperative driving behavior." The difficulty in solving these problems lies in how to quantify and predict human drivers' "willingness to yield," but traditional rule-based algorithms are ill-equipped to do so.

[0098] This application discloses a method and system for recommending vehicles prone to cutting in, based on environmental perception. In congested urban areas and when merging from highway ramps into main roads, drivers face significant challenges when changing lanes (commonly known as "cutting in"). Drivers need to simultaneously observe the distance and speed of vehicles approaching from behind in the target lane and attempt to determine the intentions of other drivers. This process is fraught with uncertainty and safety risks, easily leading to driver stress, operational errors, and even traffic accidents. Traditional driver assistance systems, such as blind spot monitoring, only provide warnings of dangerous vehicles and cannot offer decision support to drivers in complex dynamic situations. This application aims to address these pain points by using multi-sensor fusion to perceive the environment and innovatively introducing a "yielding willingness" assessment model to proactively filter and recommend the easiest target vehicle to cut in and the available space in front of it, improving the efficiency, safety, and comfort of lane-changing operations.

[0099] This application belongs to the technical field of Advanced Driver-Assistance Systems (ADAS), specifically involving vehicle environmental perception, driving behavior prediction, and human-machine interaction technologies. With the development of automotive electronics and artificial intelligence technologies, ADAS systems have evolved from initial low-level driver assistance (such as ACC adaptive cruise control and LCC lane keeping assist) to high-level driver assistance (such as highway NOA and city NOA).

[0100] To address the problems of inefficiency, conservative decision-making, and inability to understand the potential yielding intentions of other drivers in dense traffic flow in related technologies, this application aims to provide a method that can proactively identify and recommend vehicles "most likely to yield" to the driver, assisting the driver in completing lane-changing operations efficiently and safely. This application does not replace driver operation, but rather serves as a decision-making aid tool, combining human experience and judgment with the precise perception of machines.

[0101] Compared with related technologies, this application introduces the core concept of "lane-cutting feasibility scoring model", which brings the following significant benefits: improved efficiency. For example, in congested traffic conditions, this application can identify opportunities where there is little physical space but a tendency to slow down and yield, reducing the average lane-changing waiting time by 30%-70% and significantly improving traffic efficiency.

[0102] In this embodiment, by accurately capturing minute dynamics of the target vehicle (e.g., changes in acceleration), the potential behavior of the target vehicle can be predicted, preventing the driver from accidentally entering a narrowing, dangerous gap, or misjudging a seemingly spacious but unyielding situation where the following vehicle shows no intention to yield, thus enhancing safety. The "feelings" of a human driver can be quantified into reliable data support.

[0103] In this embodiment, the final decision-making power can be returned to the driver, but clear and reliable suggestions are provided. Recommending vehicles that are easy to cut in through the instrument panel and large-screen HMI greatly reduces the driver's observation burden and psychological pressure when changing lanes, making the driving process easier and more confident, thereby improving driving comfort and overall experience.

[0104] In this embodiment, the core innovation lies in the algorithm level, while the hardware requirements are relatively flexible. Good results can be achieved on a basic ADAS hardware platform consisting only of cameras and millimeter-wave radar, without incurring additional costs, and it can be applied to a wide range of vehicle models.

[0105] Currently, higher-level autonomous driving systems (L2+ and above) have begun to have Automatic Lane Change (ALC) functionality. These systems are typically based on sensors such as cameras, millimeter-wave radar, and lidar, and automatically execute a complete lane change maneuver after determining that there is sufficient safe space in the target lane.

[0106] Optionally, information about traffic participants around the vehicle can be perceived through vehicle cameras, radar, etc., to achieve perception data fusion and tracking. Kalman filtering is used to fuse the sensor input perception data, outputting data such as the ID, position, speed, acceleration, heading angle, and category of the target vehicle (vehicles moving around it) for each frame, thus achieving data fusion and target tracking.

[0107] In this embodiment, the current state of the target vehicle and V2X intent signals (if any) are input into a pre-built model in the target vehicle behavior prediction module. This model can employ a bidirectional LSTM neural network, and the training data is constructed from real-world driving datasets (such as nuScenes or HighD). The model can output a predicted longitudinal acceleration value a1=a1 / a2 at 0.5-second time steps over the next 5 seconds. (t+Δt). Based on the above acceleration prediction, the future gap change trend ΔS can be derived. Where ΔS=ΔV t1 + ΔV^2 / 2a1. ΔV can be the speed difference between the vehicle and the approaching target vehicle, ΔV = Vvehicle - Vtarget; t1 can be the time variation, t1 = t + Δt. If ΔS > the safety threshold within the next 3 seconds, then cutting in is considered feasible, and the next step of the model is to determine the specific vehicle that can cut in.

[0108] In this embodiment, a lane-cutting friendliness model can be designed, which can calculate the lane-cutting friendliness score Y of a vehicle. This score is obtained by weighting a physical friendliness factor and a willingness-to-be-friendliness factor.

[0109] Optionally, the aforementioned physical friendliness factor Fp can be calculated based on a physical safety model and can be used to reflect the objective space and speed conditions provided by the vehicle for cutting in. Physical parameters including relative distance ΔS, relative speed ΔV (negative values ​​indicate rapid approach by vehicles behind, poor conditions), and acceleration Δa can be input into the aforementioned physical safety model. The calculation based on the physical safety model can include Fp=w1 f(ΔS) + w2 f(ΔV)+w2 f(Δa). Where w1, w2, and w3 are weighting coefficients, and f, g, and h are normalization functions (such as the Sigmoid function). The physical parameters can be mapped to the interval [0, 1] through the above normalization function. The greater the distance, the smaller the velocity difference (or a positive value), and the negative acceleration (deceleration), the higher the Fp score.

[0110] Optionally, the "yield willingness flag" can be parsed from the V2X message. The willingness friendliness factor Fw can be calculated based on the "yield willingness flag" displayed by the vehicle. If the flag is "true", a high score is assigned, and the "yield willingness flag" displayed by the vehicle can be used as the display signal (P1) of the target vehicle.

[0111] Optionally, if the target vehicle does not show any explicit signal, an implicit behavior P2 can be set to continuously monitor the target vehicle's acceleration 'a'. If a significant deceleration of the target vehicle is detected (a < negative threshold), and the distance between the target vehicle and the vehicle in front (the main vehicle) increases within a safe range, the system determines that the target vehicle has an implicit intention to yield and assigns P2 a moderate score. If there are no signs of yielding, the score is 0. The willingness to yield friendliness Fw can be determined using the above method. Where Fw = P1 + P2. The above friendliness score Y can be calculated as Y = α Fp+β Fw, where α and β are weighting coefficients, and α+β=1.

[0112] Optionally, sorting and filtering may include sorting all target vehicles in descending order of their Y-values. A score threshold is set to filter out vehicles with excessively low scores (i.e., unfriendly vehicles), forming a recommended candidate list.

[0113] Recommending the best target can include identifying the vehicle with the highest score as the primary recommended target. Human-computer interaction output can include HUD (Head-Up Display) interaction, such as highlighting a vehicle model that can cut in line in dark green, accompanied by voice and text pop-up prompts, with the message "Recommended to cut in from the green vehicle." Instrument panel / large screen interaction, for example, highlighting a vehicle model that can cut in line in dark green, accompanied by voice and text pop-up prompts, with the message "Recommended to cut in from the green vehicle."

[0114] Suppose the vehicle is traveling on a highway ramp and needs to merge onto the main road. The driver activates the right turn signal, and the vehicle's driver assistance system kicks in.

[0115] Optionally, data perception includes: the right-side radar of the main vehicle detects three vehicles behind it in the current lane of the main road: X1, X2, and X3. Simultaneously, the V2X module receives basic safety messages broadcast by these three vehicles.

[0116] Optionally, target extraction includes system confirmation that X1, X2, and X3 are target vehicles to be evaluated.

[0117] Optionally, the friendliness calculation includes the following: For X1, the distance is relatively far, the speed is similar to the host vehicle, but there is no yield flag in its V2X message, and the acceleration is 0. The calculated values ​​are: Fp=0.8, Fw=0, Y1=0.8α. For X2, the distance is moderate, but it is detected to be continuously decelerating (a=-1.5m / s²), and the "yield flag" in the above V2X message is "true". The calculated values ​​are: Fp=0.6 (conditions improved due to deceleration), Fw=0.9, Y2=0.6α+0.9β. Since β is usually given a high weight (e.g., 0.7), Y2 will be significantly higher than Y1. For X3, the distance is very close and it is accelerating towards the host vehicle, with no yield signal. The calculated value is: Y3 is very low and it is filtered out.

[0118] Optionally, the vehicles to be cut in line are sorted. If the sorting result is Y2>Y1, the vehicle will select X2 as the preferred recommended target. The HUD / instrument / large screen will highlight the X2 vehicle model in dark green, accompanied by a voice and text pop-up prompt, which reads "Recommended to cut in from the green vehicle".

[0119] Optionally, in scenarios where V2X functionality is unavailable or V2X communication fails, onboard sensors (vision, radar) can be used entirely to identify yielding intentions, thus serving as an alternative data source. For example, cameras can be used to identify the flashing behavior of vehicles behind (such as hazard lights or high beams at specific frequencies) as explicit yielding signals; radar data can be used to more accurately analyze their deceleration curves as implicit signals.

[0120] Optionally, the friendliness calculation model does not have to be limited to linear weighted sums. For example, fuzzy logic systems can be used to handle uncertainties such as "close distance but strong yielding", or trained machine learning models (such as neural networks) can be used to directly output friendliness scores end-to-end, thereby achieving the purpose of algorithm model replacement.

[0121] Optionally, the recommended results do not have to be limited to a single optimal target. A ranked list of vehicles (e.g., Top 2) can be provided for the driver to choose from based on the actual situation. Alternatively, instead of specifying a particular vehicle, an "optimal cut-in time window" can be suggested, which is derived from the predicted trajectories of the recommended vehicles.

[0122] 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 a part of the embodiments of this application, and not all of them; the numbers in these embodiments are only illustrative and are not intended to be specific limitations. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application and are not specifically limited here.

[0123] Figure 2 This is a flowchart of an optional method for determining vehicles cutting in line according to an embodiment of this application. Figure 2 As shown, the method includes the following steps.

[0124] Step S202: Sensing data fusion and tracking.

[0125] Through step S202 of this application, technologies such as Kalman filter tracking, multi-target association, and point cloud fusion localization can be employed to accurately and in real-time grasp information about the vehicle's surrounding environment, including key parameters such as the position, speed, and acceleration of nearby vehicles, providing a solid data foundation for subsequent decision-making. This achieves high-precision perception of the dynamic traffic environment and ensures the reliability and continuity of the data.

[0126] Step S204, target vehicle behavior prediction.

[0127] Through step S204 of this application, the future behavior of surrounding vehicles can be predicted using methods such as LSTM driving behavior model and acceleration trend prediction, especially the dynamic changes that can affect lane-cutting operations, such as deceleration, acceleration, lane keeping, etc. Thus, it is possible to gain insight into the behavioral trends of vehicles before they make obvious actions, providing a forward-looking basis for lane-cutting feasibility assessment.

[0128] Step S206, Feasibility assessment model for adding a gap.

[0129] Through step S206 of this application, combined with the real-time perception data and prediction results obtained in steps S202 and S204, the lane-jumping friendliness scoring system of this application comprehensively evaluates each target vehicle to determine whether the target vehicle is suitable as a lane-jumping target, as well as the feasibility and safety of the lane-jumping operation. This achieves a comprehensive consideration of the possibility of lane-jumping operations, improving the accuracy and reliability of the system's recommendations.

[0130] Step S208, HMI interactive recommendation.

[0131] Through step S208 of this application, the system presents the evaluation results to the driver in an intuitive and easy-to-understand form through various human-computer interaction methods such as the dashboard, large screen, HUD head-up display, and voice prompts, guiding the driver to adopt better lane-cutting strategies.

[0132] In this embodiment, technologies such as Kalman filter tracking, multi-target association, and point cloud fusion localization are employed to accurately and in real-time grasp information about the vehicle's surrounding environment, including key parameters such as the position, speed, and acceleration of nearby vehicles. LSTM driving behavior models and acceleration trend prediction methods are used to predict the future behavior of surrounding vehicles, particularly dynamic changes that can affect lane-cutting operations, such as deceleration, acceleration, and lane keeping. Combining the real-time perception data and prediction results obtained in steps S202 and S204, each target vehicle is comprehensively evaluated to determine its suitability as a lane-cutting target. The evaluation results are presented to the driver in an intuitive and easy-to-understand manner through various human-computer interaction methods such as the dashboard, large screen, HUD head-up display, and voice prompts, guiding the driver to adopt better lane-cutting strategies. This achieves the technical effect of improving the accuracy of identifying lane-cutting vehicles, thereby solving the technical problem of low accuracy in identifying lane-cutting vehicles.

[0133] According to an embodiment of this application, a device for determining vehicles cutting in line is also provided. It should be noted that the vehicle lane entry device of this embodiment can be used to execute the method for determining vehicles cutting in line in this application embodiment.

[0134] Figure 3 This is a schematic diagram of an optional vehicle-cutting determination device according to an embodiment of this application. Figure 3 As shown, the device for determining vehicles cutting in line may include: an acquisition module 302, a prediction module 304, and a determination module 306.

[0135] The acquisition module 302 is used to acquire multiple perception data collected by multiple sensors in the vehicle, wherein the perception data is used to characterize the initial driving state of at least one vehicle waiting to cut in during the current driving process, and the at least one vehicle waiting to cut in is located behind the vehicle.

[0136] The prediction module 304 is used to predict multiple perception data to obtain the target driving data of the vehicle waiting to cut in line during its future driving process. The target driving data is used to characterize the target driving state of the vehicle waiting to cut in line during its future driving process.

[0137] The determination module 306 is used to determine the target lane-jumping vehicle from at least one lane-jumping vehicle based on the target driving data, wherein the success rate of a vehicle cutting into the lane where the target lane-jumping vehicle is located is greater than the success rate of other lane-jumping vehicles among the at least one lane-jumping vehicle excluding the target lane-jumping vehicle.

[0138] In this embodiment, the acquisition module 302 acquires multiple perception data collected by multiple sensors in the vehicle, the prediction module 304 predicts the multiple perception data to obtain the target driving data of the vehicle to be cut in the future driving process, and the determination module 306 determines the target vehicle to cut in from at least one vehicle to be cut in based on the target driving data. This achieves the technical effect of improving the accuracy of determining the vehicle to cut in, and thus solves the technical problem of low accuracy in determining the vehicle to cut in.

[0139] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to execute the method for determining vehicles cutting in line according to the embodiments of this application.

[0140] According to another aspect of the embodiments of this application, a processor is also provided for running a program, wherein the program is executed by the processor to perform the method for determining vehicles cutting in line according to the embodiments of this application.

[0141] According to another aspect of the embodiments of this application, a program product is also provided, the program product including computer instructions, wherein when the computer instructions are executed by a processor, they implement the method for determining vehicles cutting in line according to the embodiments of this application.

[0142] According to another aspect of the embodiments of this application, a vehicle is also provided, which can be used to perform the method for determining vehicles cutting in line according to the embodiments of this application.

[0143] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0144] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0145] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0146] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0147] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0148] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to related technologies, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0149] The above are merely preferred embodiments of this application. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for determining vehicles that cut in line, characterized in that, include: During the current driving process of the vehicle, multiple perception data collected by multiple sensors in the vehicle are acquired, wherein the perception data is used to characterize the initial driving state of at least one vehicle waiting to cut in during the current driving process, and at least one of the vehicles waiting to cut in is located behind the vehicle. Predicting multiple sets of perception data yields target driving data for the vehicle to be cut off during its future driving process, wherein the target driving data is used to characterize the target driving state of the vehicle to be cut off during the future driving process. Based on the target driving data, a target lane-jumping vehicle is determined from at least one of the vehicles waiting to cut in, wherein the success rate of the vehicle cutting into the lane where the target lane-jumping vehicle is located is greater than the success rate of the other vehicles waiting to cut in from the at least one group of vehicles excluding the target lane-jumping vehicle.

2. The method according to claim 1, characterized in that, The step of determining the target vehicle to cut in line from at least one of the vehicles waiting to cut in line, based on the target driving data, includes: Based on the target driving data, determine the relative distance between the vehicle to be cut in and the vehicle during the future driving process; In response to the relative distance being greater than a safe distance threshold, the vehicle that is about to cut in is identified as the initial vehicle that is about to cut in. Determine the cut-in friendliness of the initial vehicle to be cut in, wherein the cut-in friendliness is related to the cut-in success rate; Using the aforementioned "line-cutting friendliness" rating, the target line-cutting vehicle is selected from at least one of the initial vehicles waiting to cut in line.

3. The method according to claim 2, characterized in that, Determining the "line-cutting friendliness" of the initial vehicle to be cut in line includes: The target driving data of the initial vehicle waiting to cut in line is converted to obtain the relative speed between the initial vehicle waiting to cut in line and the vehicle. Based on the relative distance, the relative speed, and the target driving data, a first friendliness factor is obtained corresponding to the initial vehicle waiting to cut in line, wherein the first friendliness factor is associated with the target movement state of the initial vehicle waiting to cut in line. The friendliness of the queue addition is determined based on the first friendliness factor.

4. The method according to claim 3, characterized in that, Determining the friendliness of cutting in line based on the first friendliness factor includes: In response to the detection of the driving intention data of the initial vehicle waiting to cut in line, a willingness to give way is determined that matches the driving intention data; Based on the yielding intention, the target driving data, and the relative distance, a second friendliness factor is determined for the initial vehicle waiting to cut in, wherein the second friendliness factor is used to characterize the yielding intention of the initial vehicle waiting to cut in. The first friendliness factor and the second friendliness factor are weighted to obtain the friendliness score.

5. The method according to claim 4, characterized in that, Determining the friendliness of cutting in line based on the first friendliness factor includes: In response to the absence of detected driving intention data, a second friendliness factor corresponding to the initial vehicle waiting to cut in line is determined based on the target driving data of the initial vehicle waiting to cut in line and the relative distance. The first friendliness factor and the second friendliness factor are weighted to obtain the friendliness score.

6. The method according to claim 2, characterized in that, The step of using the "line-cutting friendliness" rating to select the target line-cutting vehicle from at least one initial group of vehicles vying to cut in line includes: Using at least one of the initial vehicles waiting to cut in line, and at least one of the initial vehicles waiting to cut in line, and a friendliness threshold, the initial vehicles waiting to cut in line are screened. The vehicle with the highest "friendliness" among the selected initial vehicles waiting to cut in line is identified as the target vehicle for cutting in line.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: The display module in the vehicle is controlled to display the target vehicle that is cutting in line, according to the display requirements, and to output a prompt message, wherein the prompt message is used to notify the vehicle of the target vehicle that is about to cut in line.

8. A vehicle, characterized in that, Used to perform the method according to any one of claims 1 to 7.

9. A processor, characterized in that, The processor is used to run a program, wherein the program is executed by the processor to perform the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that, Includes computer instructions that, when executed by a processor, implement the method described in any one of claims 1 to 7.

Images