Vehicle control method, device, controller, vehicle and storage medium

Abstract

This application relates to a vehicle control method, device, controller, vehicle, and storage medium. The method includes: acquiring cutting-in intention information of a second vehicle in an adjacent lane next to the lane occupied by a first vehicle; the cutting-in intention information is determined based on vehicle information monitored for the second vehicle, and includes cutting-in intention intensity and cutting-in probability; determining an intention intensity range based on a preset intention intensity threshold; determining the cutting-in intention type corresponding to the intention intensity range as the cutting-in intention type of the second vehicle if a preset type determination condition corresponding to the intention intensity range is met based on vehicle information; determining a control strategy for the first vehicle based on the cutting-in intention type, cutting-in intention intensity, cutting-in probability, and the vehicle category of the second vehicle; and controlling the first vehicle to drive according to the control strategy. This method can effectively balance the intensity of the cutting-in game with vehicle driving safety.

Description

Technical Field

[0001] This application relates to the field of vehicle control technology, and in particular to a vehicle control method, device, controller, vehicle, and storage medium. Background Technology

[0002] In intelligent driving and advanced driver assistance systems, anti-cutting control is a crucial element in improving driving safety and traffic efficiency. Related anti-cutting technologies typically determine whether there is an intention to cut in based on the movement status of adjacent vehicles (such as lateral speed and relative distance), and then implement corresponding strategies such as accelerating to prevent cutting or slowing down to yield based on safety considerations. However, real-world road scenarios are complex, and these anti-cutting technologies can easily become overly aggressive, leading to safety risks, or excessively yield, sacrificing traffic efficiency, making it difficult to achieve an effective balance between the intensity of the cut-in game and vehicle driving safety. Summary of the Invention

[0003] Based on this, this application addresses the aforementioned technical problems by providing a vehicle control method, device, controller, vehicle, and storage medium that can effectively balance the intensity of the lane-jumping game with vehicle driving safety.

[0004] In a first aspect, this application provides a vehicle control method, including:

[0005] The cut-in intention information of the second vehicle in the adjacent lane to the lane of the first vehicle is obtained. The cut-in intention information is determined based on the vehicle information monitored for the second vehicle. The cut-in intention information includes the cut-in intention intensity, which characterizes the strength of the second vehicle's subjective intention to cut in, and the cut-in probability, which characterizes the possibility of the second vehicle cutting in.

[0006] Based on the preset intent intensity threshold, determine the intent intensity range in which the intention to cut in line falls;

[0007] If the vehicle information determines that the preset type determination condition corresponding to the intent intensity range is met, the cut-in intent type corresponding to the intent intensity range is determined as the cut-in intent type of the second vehicle.

[0008] Based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category of the second vehicle, determine the control strategy for the first vehicle;

[0009] The first vehicle is controlled to move according to the control strategy.

[0010] In the aforementioned vehicle control method, a preset intent intensity threshold is used to determine the intent intensity range of the cutting intent information, thereby achieving a coarse classification of cutting intent types. Then, the vehicle information is type-verified according to the type judgment conditions corresponding to the intent intensity range in which the cutting intent intensity is located, resulting in an accurate cutting intent type. Combining the cutting intent type, cutting intent intensity, cutting probability, and the vehicle category of the second vehicle, the control strategy for the first vehicle is determined, and the first vehicle is controlled according to this control strategy. The cutting intent type can accurately reflect the nature of the cutting behavior of adjacent vehicles, and the vehicle category can distinguish the differentiated safety risks and traffic priorities of vehicles with different physical attributes. Thus, it can simultaneously take into account the nature, intensity, probability, and vehicle category of cutting behavior in complex cutting scenarios, flexibly adjust the corresponding control strategy, and make the control strategy match the real cutting scenario, thereby achieving an effective balance between the intensity of the cutting game and vehicle driving safety.

[0011] In an optional embodiment of the first aspect, when it is determined based on vehicle information that a preset type determination condition corresponding to the intent intensity interval is met, the type of cutting-in intent corresponding to the intent intensity interval is determined as the cutting-in intent type of the second vehicle. This includes: determining cutting-in intent classification parameters based on road information of the road where the first vehicle and the second vehicle are located, and vehicle information; wherein the cutting-in intent classification parameters include at least one of headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road status parameters; when it is determined based on the cutting-in intent classification parameters that a preset type determination condition corresponding to the intent intensity interval is met, the type of cutting-in intent corresponding to the intent intensity interval is determined as the cutting-in intent type of the second vehicle; the type determination condition includes logical rules for classifying and verifying the intent intensity interval in which the cutting-in intent intensity is located based on the cutting-in intent classification parameters.

[0012] In this optional embodiment, fine-grained verification of the lane-cutting intention type can be performed based on at least one of the following: vehicle headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road state parameters. This can improve the accuracy of lane-cutting intention type identification.

[0013] In an optional embodiment of the first aspect, determining the type of the second vehicle's intention to cut in line based on the intensity of the intention and vehicle information includes: determining the type of the second vehicle's intention to cut in line based on the intensity of the intention and vehicle information when the probability of cutting in line exceeds a preset probability threshold.

[0014] In this optional embodiment, when the probability of cutting in line exceeds a preset probability threshold, the processing to determine the type of cutting in line is performed. This can reduce unnecessary calculations in scenarios where the probability of cutting in line is extremely low, and reduce the consumption of computing resources while ensuring the timeliness and accuracy of the determination of the type of cutting in line.

[0015] In an optional embodiment of the first aspect, determining a control strategy for the first vehicle based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs includes: determining an intensity adjustment weight based on the type of cutting-in intention, and determining a cutting-in permission coefficient based on the vehicle category to which the second vehicle belongs; the cutting-in permission coefficient is used to characterize the priority of the first vehicle to yield to the second vehicle; determining the cutting-in game intensity based on the intensity adjustment weight, the cutting-in permission coefficient, the intensity of the cutting-in intention, and the probability of cutting-in; and determining a control strategy for the first vehicle based on the cutting-in game intensity.

[0016] In this optional embodiment, the intensity adjustment weight and the lane-jumping permission coefficient are determined based on the type of lane-jumping intention and the vehicle category to which the second vehicle belongs, respectively. The intensity of the lane-jumping game is determined by comprehensively considering the intensity adjustment weight, the lane-jumping permission coefficient, the intensity of the lane-jumping intention, and the lane-jumping probability. The control strategy is then determined based on the intensity of the lane-jumping game. This approach can comprehensively consider multi-dimensional information such as discrete lane-jumping intention types and vehicle categories, as well as continuous lane-jumping intention intensity and lane-jumping probability, to determine the control strategy, thus ensuring the pertinence and accuracy of the control strategy.

[0017] In an optional embodiment of the first aspect, determining the intensity of the lane-jumping game based on the intensity adjustment weight, the lane-jumping permission coefficient, the intensity of the lane-jumping intention, and the lane-jumping probability includes: adjusting the intensity of the lane-jumping intention through the intensity adjustment weight to obtain the adjusted lane-jumping intention intensity; determining the game weighting weight based on the driving scenarios of the first vehicle and the second vehicle; and fusing the adjusted lane-jumping intention intensity, the lane-jumping permission coefficient, and the lane-jumping probability according to the game weighting weight to obtain the lane-jumping game intensity.

[0018] In this optional embodiment, the intensity of the intention to cut in line is adjusted by adjusting the intensity weight, so that the intensity of the game of cutting in line is precisely matched with the actual threat level of the second vehicle, thereby improving the humanization and safety of the decision-making process. By determining the game weighting weight according to the driving scenario, the intensity of the game of cutting in line can be adaptively determined according to the driving scenario, which can take into account both traffic efficiency and driving safety, thereby ensuring the pertinence and reliability of the intensity of the game of cutting in line.

[0019] In an optional embodiment of the first aspect, determining the control strategy for the first vehicle based on the intensity of the lane-jumping game includes: determining the game intensity range where the lane-jumping game intensity is located based on a preset game intensity threshold; and determining the preset control strategy that conforms to safety constraints corresponding to the game intensity range as the control strategy for the first vehicle.

[0020] In this optional embodiment, the game intensity range of the lane-jumping game intensity is determined according to a preset game intensity threshold, and the preset control strategy corresponding to the game intensity range that meets safety constraints is determined as the control strategy for the first vehicle. This can ensure that the control strategy can match the real lane-jumping scenario and effectively balance the lane-jumping game intensity and vehicle driving safety.

[0021] In an optional embodiment of the first aspect, the vehicle control method further includes: when the cut-in intention type is a misoperation cut-in type, controlling the first vehicle to issue a cut-in misoperation reminder to the second vehicle; and when the second vehicle continues to cut in, returning to the step of determining the cut-in intention type of the second vehicle based on the cut-in intention intensity and vehicle information.

[0022] In this optional embodiment, when a erroneous lane-cutting type is identified, the first vehicle is controlled to issue a lane-cutting error warning. When the second vehicle continues to cut in, the lane-cutting intention type is redefined. This reduces unnecessary anti-lane-cutting control intervention and ensures the accuracy of lane-cutting intention type identification, thereby improving the accuracy of vehicle control in complex driving scenarios.

[0023] In an optional embodiment of the first aspect, the vehicle control method further includes: determining the control strategy as proactive yielding when the second vehicle belongs to a special vehicle category.

[0024] In this optional embodiment, when the second vehicle is a special vehicle, controlling the first vehicle to actively give way can ensure the passage efficiency of special vehicles.

[0025] In an optional embodiment of the first aspect, the vehicle control method further includes: predicting a cutting-in intention based on wheel-end motion information and cutting-in trend prediction information included in the vehicle information, to obtain cutting-in intention information, wherein the cutting-in trend prediction information is determined based on the historical driving information of the second vehicle.

[0026] In this optional embodiment, the intention to cut in line is predicted by combining the wheel end motion information and the cutting-in trend prediction information in the vehicle information, so as to obtain the cutting-in intention information of the second vehicle. This can accurately capture the cutting-in intention before the body movement of the second vehicle obviously shows the cutting-in action, which can improve the accuracy and timeliness of cutting-in intention recognition.

[0027] Secondly, this application also provides a vehicle control device, comprising:

[0028] The lane-cutting intention information acquisition module is used to acquire lane-cutting intention information of a second vehicle in an adjacent lane adjacent to the lane where the first vehicle is located. The lane-cutting intention information is determined based on vehicle information monitored for the second vehicle. The lane-cutting intention information includes lane-cutting intention intensity, which characterizes the strength of the second vehicle's subjective intention to cut in, and lane-cutting probability, which characterizes the possibility of the second vehicle cutting in.

[0029] The lane-cutting intent type determination module is used to determine the intent intensity range in which the lane-cutting intent intensity is located based on a preset intent intensity threshold; and if the preset type determination conditions corresponding to the intent intensity range are met based on vehicle information, the lane-cutting intent type corresponding to the intent intensity range is determined as the lane-cutting intent type of the second vehicle.

[0030] The control strategy determination module is used to determine the control strategy for the first vehicle based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs.

[0031] The control strategy execution module is used to control the first vehicle to move according to the control strategy.

[0032] Thirdly, this application also provides a controller, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the methods described above.

[0033] Fourthly, this application also provides a vehicle including at least one sensor and a controller provided in the third aspect, wherein the sensor is used to monitor a second vehicle to obtain vehicle information of the second vehicle.

[0034] Fifthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the above aspects.

[0035] Sixthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in any of the above aspects.

[0036] Regarding the beneficial effects of any of the technical solutions in the second to sixth aspects mentioned above, refer to the beneficial effects of the corresponding technical solutions in the first aspect; repeated examples will not be listed here. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram of an optional application environment for a vehicle control method in one embodiment.

[0039] Figure 2 This is a schematic diagram of an optional flow of a vehicle control method in one embodiment;

[0040] Figure 3 This is a schematic diagram of an optional process for determining a control strategy in one embodiment;

[0041] Figure 4 This is a schematic diagram of an optional process for determining the strength of a race-up game in one embodiment;

[0042] Figure 5 This is a schematic diagram of an optional structure of the vehicle control device in one embodiment;

[0043] Figure 6 This is a schematic diagram of an optional internal structure of the controller in one embodiment. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0045] The terms "first," "second," etc., used in this application may be used to describe various elements, but these elements are not limited by these terms. These terms are used only to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0046] Currently, in the relevant technologies of intelligent driving, when the intention of adjacent vehicles to cut in is detected, the process of preventing cutting in usually involves shortening the following distance between the vehicle and the vehicle in front. However, the control effect of this method is limited. In some situations, the vehicle attempting to cut in cannot receive an effective signal to refuse cutting in, resulting in poor anti-cutting effect.

[0047] Research has revealed that during the process of adjacent vehicles cutting in line, there are varying degrees of cutting in, and different vehicle types have different safety impacts on cutting-in control. If corresponding control strategies can be adopted by combining the vehicle type and the degree of cutting in, it is possible to achieve differentiated anti-cutting vehicle control based on the game strength of cutting in and the vehicle type, which is determined by the degree of cutting in. This enables anti-cutting control in different scenarios, helps to achieve safe and effective anti-cutting vehicle control, improves the effectiveness of anti-cutting control, and reduces the probability of safety accidents.

[0048] The vehicle control method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, a first vehicle is traveling forward in lane 1, and a second vehicle is traveling in lane 2 adjacent to lane 1. The second vehicle intends to cut in line, meaning it intends to enter lane 1 from lane 2. The controller can acquire the second vehicle's intention to cut in line. This intention is determined based on vehicle information monitored for the second vehicle and includes the intensity and probability of the intention. The controller can determine the intensity range of the intention based on a preset threshold. If the vehicle information confirms that the preset type of the intention range is met, the controller identifies the type of intention corresponding to that intensity range as the second vehicle's intention type. Combining the intention type, intensity, probability, and vehicle category of the second vehicle, the controller determines the control strategy for the first vehicle. The controller can then control the first vehicle to engage in a lane-cutting game according to this strategy.

[0049] The controller can be located in the first vehicle, such as a central domain controller or a distributed controller in the first vehicle; the controller can also be located outside the first vehicle, such as a remote server, a roadside unit (RSU), a mobile edge computing node, or a cloud server that can control the first vehicle.

[0050] In one exemplary embodiment, such as Figure 2 As shown, a vehicle control method is provided. Taking the application of this method to a controller in a first vehicle as an example, the method includes the following steps S201 to S204. Wherein:

[0051] Step S201: Obtain the cutting-in intention information of the second vehicle in the adjacent lane of the lane adjacent to the lane of the first vehicle. The cutting-in intention information is determined based on the vehicle information monitored for the second vehicle. The cutting-in intention information includes the cutting-in intention intensity, which characterizes the strength of the second vehicle's subjective intention to cut in, and the cutting-in probability, which characterizes the possibility of the second vehicle cutting in.

[0052] In this scenario, the first vehicle is the one that needs to be controlled in the lane-cutting situation, such as decelerating, maintaining its original speed, or making slight lateral movements. The second vehicle is a potential lane-cutter located in the adjacent lane (near lane) of the first vehicle. Lane-cutting intention information describes the possible lane-cutting behavior of the second vehicle. This information is obtained through lane-cutting intention analysis based on vehicle information monitored for the second vehicle. The information includes lane-cutting intention strength and lane-cutting probability. Lane-cutting intention strength characterizes the subjective willingness of the second vehicle to cut in. This subjective willingness strength can be predicted based on the second vehicle's vehicle information, meaning it represents a prediction result for the second vehicle. For example, the value of lane-cutting intention strength can range from 0 to 1. A value closer to 1 indicates a stronger predicted subjective willingness to cut in based on the second vehicle's vehicle information, suggesting a more urgent desire from the driver to cut in. The probability of a vehicle cutting in line is used to quantify the likelihood of a second vehicle actually cutting in line. It characterizes the degree of probability that the second vehicle will cut in line. The probability of cutting in line can be predicted based on the vehicle information of the second vehicle; that is, the probability of cutting in line is also a prediction result for the second vehicle. For example, the probability of cutting in line can range from 0 to 1. The closer the value is to 1, the higher the predicted probability of the second vehicle cutting in line based on the vehicle information, meaning that the second vehicle is considered more likely to cut in line.

[0053] Vehicle information includes various data obtained from monitoring the second vehicle, such as, but not limited to, image data of the second vehicle, kinematic information (longitudinal velocity, lateral velocity, lateral acceleration, and heading angle of the second vehicle), spatial position information (lane line distance of the second vehicle, lateral distance between the second vehicle and the vehicle in front), vehicle lighting information (such as vehicle lighting status and duration), wheel end motion information (front wheel steering angle, wheel end lateral offset rate), lane-cutting trend prediction information (wheel track deviation trend, probing characteristics, lane-cutting acceleration), or driving characteristic information (probing characteristics, lane-cutting behavior characteristics). Vehicle information can be obtained by sensors installed in the first vehicle (such as forward-view cameras, surround-view cameras, millimeter-wave radar, wheel speed sensors, etc.) monitoring the second vehicle.

[0054] Optionally, during the driving of the first vehicle, the controller can acquire the lane-cutting intention information of the second vehicle. The lane where the second vehicle is located is adjacent to the lane where the first vehicle is located. The lane-cutting intention information can include the lane-cutting probability and the lane-cutting intention intensity. The lane-cutting probability can characterize the likelihood of the second vehicle cutting in front of the first vehicle, and the lane-cutting intention intensity can characterize the strength of the second vehicle's subjective willingness to cut in front of the first vehicle. Both the lane-cutting probability and the lane-cutting intention intensity are prediction results for the second vehicle. In some embodiments, the controller can continuously monitor the second vehicle located in the adjacent lane through various sensors (such as cameras, radar, etc.) on the first vehicle to obtain the vehicle information of the second vehicle. Based on the vehicle information of the second vehicle, the lane-cutting intention information of the second vehicle can be determined. For example, the lane-cutting intention information determined by the controller based on the vehicle information of the second vehicle can be "lane-cutting probability 0.9, lane-cutting intention intensity 0.8", which means that the second vehicle is highly likely to cut in front of the first vehicle and has a strong subjective willingness to do so.

[0055] In some embodiments, the controller can perform lane-cutting intention analysis based on the vehicle information of the second vehicle. For example, the controller can input the vehicle information into a pre-trained lane-cutting intention prediction model. The lane-cutting intention prediction model can be trained based on pre-labeled sample data (lane-cutting intention information samples - vehicle information samples). The lane-cutting intention prediction model can be at least one of various models, including but not limited to a temporal prediction model based on a long short-term memory network, a Transformer model based on a multi-head attention mechanism, a regression model based on a gradient boosting decision tree, or a temporal convolutional network. The lane-cutting intention prediction model can perform lane-cutting intention analysis based on the input vehicle information to output the lane-cutting intention information of the second vehicle. The lane-cutting intention information can include the lane-cutting intention strength and lane-cutting probability of the second vehicle. In some embodiments, the controller can acquire vehicle information of vehicles cutting in various lane-cutting scenarios. The lane-cutting vehicle can be the vehicle that triggers lane-cutting in the lane-cutting scenario. The vehicle information can be at least one of the following: image data, kinematic information, spatial position information, headlight information, wheel-end motion information, lane-cutting trend prediction information, or driving feature information of the lane-cutting vehicle. Based on vehicle information of vehicles that cut in line, experts can manually annotate the information to label the cutting-in intentions of each vehicle. This intention information can include the intensity and probability of cutting in line. For example, multiple experts can score the strength of the driver's subjective intention to cut in line based on the vehicle information, with scores ranging from 0 to 1. The average score from multiple experts is used as the intensity label for the cutting-in intention of that vehicle. Alternatively, based on each vehicle information, it can be determined whether the vehicle ultimately cuts in line. For instance, monitoring whether more than 50% of the vehicle's body is completely inside the adjacent lane within a preset time (e.g., 3 seconds) can be used. If so, the probability of cutting in line for that vehicle is determined to be 1. Furthermore, the probability of cutting in line can be determined based on the percentage of the vehicle's body that is inside the adjacent lane. When the percentage exceeds 50%, the probability is determined to be 1, and the closer the percentage is to 50%, the closer the probability is to 1. This allows for the determination of the cutting-in probability for each vehicle. The controller can use individual vehicle information as vehicle information samples, and also use the lane-cutting intent strength labels and lane-cutting probabilities labeled for each vehicle information as vehicle information samples, thus constructing multiple pre-labeled sample data. The controller can then use the sample data to obtain a lane-cutting intent prediction model.For example, the controller can input sample data into an initialized LSTM (Long Short-Term Memory) or Transformer network for iterative training. For the probability of cutting in line, the cross-entropy loss function is used; for the intensity of the intention to cut in line, the mean squared error loss function is used. The training optimizer is Adam, and the learning rate is initially set to 0.001. Through backpropagation and gradient descent algorithms, the model weights are continuously updated until the weighted sum of the cross-entropy loss function and the mean squared error loss function converges. Then the training is completed, and a pre-trained model for predicting the intention to cut in line is obtained. This model can analyze the intention to cut in line based on the input vehicle information to output information on the intention to cut in line, including the intensity and probability of the intention to cut in line. The values ​​of the intensity and probability of the intention to cut in line can both be in the range of 0-1.

[0056] Step S202: Determine the intention intensity range in which the intention to cut in is located based on the preset intention intensity threshold.

[0057] The intent intensity threshold is used to quantify the intensity of consecutive lane-cutting intentions into non-overlapping intent intensity intervals. The specific value of the intent intensity threshold can be pre-determined using empirical data or machine learning. For example, if the preset intent intensity thresholds are 0.2, 0.4, and 0.8, these three values ​​divide the [0,1] interval into four intent intensity intervals: less than 0.2, greater than or equal to 0.2 and less than 0.4, greater than or equal to 0.4 and less than 0.8, and greater than or equal to 0.8. Each intent intensity interval can correspond to a type of lane-cutting intention. The intent intensity interval is a numerical range determined by the intent intensity threshold, and each intent intensity interval can correspond one-to-one with a preset type of lane-cutting intention. Based on the intent intensity interval, a coarse classification can be performed on the type of lane-cutting intention of the second vehicle.

[0058] Optionally, the controller can obtain a preset intent strength threshold and compare the cut-in intent strength with the intent strength threshold to determine the intent strength interval in which the intent strength threshold is located. The intent strength interval is divided by each intent strength threshold. For example, the intent strength threshold may include [0.15, 0.35, 0.75]. If the cut-in intent strength is ≥0.75, it falls into interval A ([0.75, 1]). Otherwise, if 0.75 > cut-in intent strength ≥0.35, it falls into interval B ([0.35, 0.75)). Otherwise, if 0.35 > cut-in intent strength ≥0.15, it falls into interval C ([0.15, 0.35)). Otherwise, it falls into interval D ([0, 0.15)).

[0059] Step S203: If the vehicle information determines that the preset type determination condition corresponding to the intent intensity range is met, the cut-in intent type corresponding to the intent intensity range is determined as the cut-in intent type of the second vehicle.

[0060] The "cutting intention type" is used to finely classify the cutting-in behavior of the second vehicle to characterize the nature of the cutting-in behavior. For example, the cutting-in intention type can include three types: active cutting-in, passive cutting-in, and accidental cutting-in. Each cutting-in intention type can correspond to a corresponding quantifiable type determination condition, which can be obtained in advance through calibration. The cutting-in intention type of the second vehicle can be obtained by combining the cutting-in intention intensity and vehicle information, based on the preset type determination condition. The type determination condition is a set of preset logical rules for each intention intensity range, used to classify and verify the cutting-in intention type of the second vehicle. The type determination condition can be determined based on the cutting-in intention classification parameters corresponding to various cutting-in intention types. The cutting-in intention classification parameters can include, but are not limited to, at least one of the following: headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road state parameters. For different cutting-in intention classification parameters, corresponding logical rules can be configured to obtain the type determination condition. In some embodiments, the controller can use the cutting-in intention intensity as a coarse classification basis, and then use various cutting-in intention classification parameters of vehicle information for fine-grained verification to determine the cutting-in intention type of the second vehicle.

[0061] For example, the controller can determine the preset type determination conditions corresponding to the intention intensity range. For instance, it can query and determine the preset type determination conditions corresponding to the lane-cutting intention type based on the type identifier represented by the intention intensity range. The type determination conditions can include various determination conditions for vehicle information related to the lane-cutting intention type. For example, for intention intensity range C, the corresponding type determination conditions can include: whether the turn signal state T is equal to 1, and whether the vehicle's lateral speed is between 2 and 3. up to 5 Is the lateral acceleration between 0.5 and 0.5? (meters per second squared) to 2 Whether the angle of entry is between 5° and 15°, and whether the gap encroachment acceleration is between -0.01°. (meters per millisecond squared) to -0.005 At least one of the various determination conditions, such as between.

[0062] The controller can perform type verification on the vehicle information of the second vehicle based on the type determination criteria corresponding to the intent intensity interval to determine whether the second vehicle belongs to the type of cutting-in intent represented by the intent intensity interval. In some embodiments, the controller can determine the cutting-in intent classification parameters based on the vehicle information of the second vehicle, and compare the cutting-in intent classification parameters with the type determination criteria corresponding to the intent intensity interval to determine whether the cutting-in intent classification parameters meet the type determination criteria. If the type determination criteria are met, the controller can determine that the cutting-in intent type of the second vehicle is the cutting-in intent type corresponding to the intent intensity interval. In some embodiments, if it is determined based on the vehicle information that the type determination criteria are not met, the controller can return to reacquire the cutting-in intent information of the second vehicle, and then determine the cutting-in intent type of the second vehicle based on the reacquired cutting-in intent information.

[0063] Step S204: Determine the control strategy for the first vehicle based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs.

[0064] The vehicle category can be determined by classifying the second vehicle according to dimensions such as physical form, motion characteristics, and legal status. Different vehicle categories can correspond to different anti-jamming game strategies. The classification method of vehicle categories can be predetermined according to actual needs, such as classifying them into two-wheeled vehicles, three-wheeled vehicles, passenger cars, large vehicles, and special vehicles. The vehicle category of the second vehicle can be determined by category recognition based on the image data of the second vehicle. The control strategy can be used to control the first vehicle in the anti-jamming game. For example, the control strategy can include driving instructions for the first vehicle, which can be used to control the longitudinal movement of the first vehicle, such as accelerating, decelerating, or maintaining speed; it can also control the lateral movement of the first vehicle, such as applying a slight steering action to convey a rejection signal; it can also include human-machine interaction measures, such as flashing lights or honking the horn as a warning. The control strategy can be obtained by combining the type of jamming intention, the intensity of jamming intention, the probability of jamming, and the vehicle category, so as to execute a matching anti-jamming game strategy for complex jamming scenarios.

[0065] Optionally, the controller can determine the vehicle category of the second vehicle, which can be determined based on the image data of the second vehicle through category classification processing. The controller can combine the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to determine the control strategy for the first vehicle. For example, the controller can obtain a preset strategy mapping table, which can record the correspondence between different parameter combinations and corresponding control strategies. The parameter combinations can be obtained based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category. The controller can query the strategy mapping table based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category of the second vehicle to determine the control strategy for the first vehicle. In some embodiments, the controller can also use a pre-trained decision tree or artificial neural network model to map based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to obtain the corresponding control strategy. The decision tree or artificial neural network model can be trained based on pre-labeled sample data (samples of cutting-in intention type - samples of cutting-in intention intensity - samples of cutting-in probability - samples of vehicle category - samples of control strategy).

[0066] Step S205: Control the first vehicle to drive according to the control strategy.

[0067] For example, the controller can determine a driving instruction for the first vehicle based on a control strategy, and control the first vehicle to drive according to the driving instruction, thereby controlling the first vehicle to engage in anti-cutting game against the second vehicle. For instance, if the controller determines the control strategy as "execute longitudinal soft rejection, i.e., smooth deceleration to narrow the following window" based on the cutting-in intention type being active cutting-in, cutting-in probability being 0.7, cutting-in intention intensity being 0.6, and the vehicle category being a passenger car, then the controller can generate "target acceleration of -1.5". The controller sends a driving instruction of "(meters per second), duration of 2 seconds" to the first vehicle to control the first vehicle to drive according to the instruction. In some embodiments, while controlling the first vehicle to drive according to the control strategy, the controller can also continuously monitor the second vehicle, thereby dynamically controlling the first vehicle based on the real-time cutting-in intention information of the second vehicle.

[0068] In the aforementioned vehicle control method, a preset intent intensity threshold is used to determine the intent intensity range of the cutting intent information, thereby achieving a coarse classification of cutting intent types. Then, the vehicle information is type-verified according to the type judgment conditions corresponding to the intent intensity range in which the cutting intent intensity is located, resulting in an accurate cutting intent type. Combining the cutting intent type, cutting intent intensity, cutting probability, and the vehicle category of the second vehicle, the control strategy for the first vehicle is determined, and the first vehicle is controlled according to this control strategy. The cutting intent type can accurately reflect the nature of the cutting behavior of adjacent vehicles, and the vehicle category can distinguish the differentiated safety risks and traffic priorities of vehicles with different physical attributes. Thus, it can simultaneously take into account the nature, intensity, probability, and vehicle category of cutting behavior in complex cutting scenarios, flexibly adjust the corresponding control strategy, and make the control strategy match the real cutting scenario, thereby achieving an effective balance between the intensity of the cutting game and vehicle driving safety.

[0069] In an exemplary embodiment, when it is determined based on vehicle information that the preset type determination conditions corresponding to the intent intensity range are met, the type of cutting-in intent corresponding to the intent intensity range is determined as the cutting-in intent type of the second vehicle. This includes: determining cutting-in intent classification parameters based on road information of the road where the first and second vehicles are located, and vehicle information; wherein the cutting-in intent classification parameters include at least one of headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road status parameters; when it is determined based on the cutting-in intent classification parameters that the preset type determination conditions corresponding to the intent intensity range are met, the type of cutting-in intent corresponding to the intent intensity range is determined as the cutting-in intent type of the second vehicle; the type determination conditions include logical rules for classifying and verifying the intent intensity range in which the cutting-in intent intensity is located based on the cutting-in intent classification parameters.

[0070] The road information includes the status data of the roads where the first and second vehicles are located. This road information can be used to help determine whether the second vehicle's lane-cutting behavior was forced by road conditions, such as lane merging, construction, or an accident blocking the lane. Road information may include, but is not limited to, lane-merging warnings provided by the navigation system, road construction events, lane reduction information, and at least one of the visual features identified by forward-looking or surround-view cameras, such as lane line disappearances, construction cones, and accident vehicles. The controller can obtain road information by fusing navigation data with visual recognition results. For example, if the navigation indicates a lane merging 2 kilometers ahead and the camera confirms that lane markings are gradually decreasing, the controller determines that the road information is a lane-merging scenario.

[0071] The lane-cutting intent classification parameters can include multi-dimensional features used to determine the lane-cutting intent type of the second vehicle. These parameters can be extracted from vehicle and road information and may include one or more of the following: headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road state parameters. After determining the intent intensity range, the controller can use the lane-cutting intent classification parameters to verify whether the type determination conditions corresponding to that intent intensity range are met, thereby verifying whether the second vehicle belongs to the lane-cutting intent type corresponding to that intent intensity range.

[0072] Among them, the headlight status parameters are indicators used to reflect whether the second vehicle actively transmits lane change or cut-off signals through its headlights. For example, headlight status parameters may include the turn signal activation indicator T (with a value of 0 or 1, where 1 indicates that the turn signal is on) and the continuous flashing time of the turn signal. (This indicates the duration of continuous flashing of the turn signal). In some embodiments, the controller can perform image recognition on the headlight area of ​​the second vehicle using the forward-facing camera of the first vehicle, detect the on / off state and flashing cycle of the second vehicle's turn signals, and thus determine T and The system obtains the vehicle light status parameters. These parameters can capture the intentions of neighboring vehicles in the classification of lane-cutting intentions. For example, failure to use a turn signal usually corresponds to malicious lane-cutting or misoperation, while prolonged flashing indicates a stable intention. For instance, the controller determines through multi-frame image analysis that the left turn signal of the second vehicle is constantly on (T=1), and has been on for 700 milliseconds since its initial activation. =700ms), indicating that the second vehicle's lane change intention is clear and continuous. Alternatively, if the controller does not detect any illumination of the second vehicle's turn signal (T=0) and there is no change within the last 2 seconds, it indicates that the second vehicle did not actively transmit a lane change signal.

[0073] Spatial parameters are used to describe the relative positional relationships between the second vehicle and the first vehicle, lane lines, and road clearances. Spatial parameters may include, but are not limited to, the vertical distance from the center of the second vehicle to the lane line. The lateral distance from the center of the second vehicle to the safety clearance between the first vehicle and the vehicle in front. The distance between the second vehicle and the vehicle in front and distance to the car behind and the number of vehicles within 20 meters of the second vehicle. At least one of various parameters. Spatial index parameters can be measured using visual ranging, millimeter-wave radar, or lidar. These parameters are used in the lane-cutting intent classification to determine whether the second vehicle has the space to cut in and whether cutting in is necessary. For example, the controller obtains spatial index parameters through lidar point cloud computing. It is 1.2 meters. It is 0.2 meters, and A single vehicle indicates a narrow gap, with the second vehicle very close to the lane and sparse surrounding traffic, suggesting the second vehicle may have intentionally cut in. Similarly, the controller obtains distance via camera measurement. The distance is 0.5 meters (adjacent car closely following the car in front). It is 0.8 meters (the car behind is close), and The number of vehicles is 8, indicating that the second vehicle was squeezed in the dense traffic from both the front and rear, suggesting that the second vehicle may have been forced to cut in.

[0074] Kinematic parameters are used to describe the motion characteristics of the second vehicle. Kinematic parameters may include, but are not limited to, longitudinal velocity. lateral velocity lateral acceleration The angle between the vehicle's heading angle and the lane centerline and angle of entry At least one of various parameters. The controller can calculate kinematic parameters by performing position-time difference calculations on visual multi-frame tracking or radar measurements. These kinematic parameters reflect the aggressiveness and maneuverability of the second vehicle's cutting-in action. For example, a larger lateral velocity and acceleration of the second vehicle indicate a more rapid cut-in; a larger angle indicates a more pronounced deviation trend. In some embodiments, the controller calculates the kinematic parameters based on the lateral displacement changes of the second vehicle in five consecutive frames of images. =4.5 Then, the lateral velocity difference is obtained. =1.5 Simultaneously, by fitting the vehicle body contour, =12°, these kinematic parameters indicate that the second vehicle is veering into the lane where the first vehicle is located at a higher speed.

[0075] Driving characteristic parameters are used to describe the driving behavior and trends exhibited by the second vehicle before or during cutting in. Driving characteristic parameters may include, but are not limited to, the real-time steering angle of the front wheels. Front wheel steering angular velocity Lateral offset rate at wheel ends Wheel track offset trend , probing features (Derivative of longitudinal acceleration), gap encroachment acceleration ( The controller can capture the position and steering changes of the wheels in the second vehicle using a camera, and combine this with multi-frame motion to determine the driving characteristic parameters of the second vehicle. For example, by continuously detecting the profile of the left front wheel of the second vehicle, the controller can detect that its steering angle gradually increases from 0 degrees to 3 degrees. =3°), and the steering angular velocity The lateral offset rate at the wheel end is 1.5° / s. It is 0.002 This indicates that the wheels of the second vehicle have begun to drift towards the lane where the first vehicle is located, but the vehicle body has not yet moved significantly.

[0076] Road state parameters describe the road conditions of the road where the first and second vehicles are located. These parameters can describe various road conditions such as normal roads, lane merging scenarios (reduced lanes, disappearance of adjacent lanes), lane construction, or accidents blocking lanes. The controller can obtain road event warnings (such as lane merging and construction areas) through the in-vehicle navigation system and use forward-looking or surround-view cameras to identify auxiliary features such as lane line disappearances, construction cones, and accident vehicles. The controller then merges these features to determine the road state parameters. Based on these parameters, it can distinguish whether the second vehicle's cutting in is forced by road conditions; for example, cutting in during a lane merging scenario is highly likely to be a passive act.

[0077] The type determination criteria include logical rules for classifying and verifying the intensity of the cutting-in intention based on the cutting-in intention classification parameters within the desired intensity range. Different logical rules can be configured for different cutting-in intention classification parameters. For example, for the malicious cutting-in type corresponding to intensity range A, the type determination criteria may include the headlight status parameter T=0, kinematic index parameters, etc. >5 Spatial index parameters Multiple logical rules, such as <1.5m, can be used to classify and verify the intensity of the intention to cut in line based on the intensity range of the intention.

[0078] For example, the controller can acquire vehicle information of the second vehicle, as well as road information of the road where the first and second vehicles are located. The vehicle information of the second vehicle can be obtained through monitoring of the second vehicle using various sensors installed in the first vehicle (such as cameras, radar, and inertial measurement units); the road information can be determined based on road event data provided by the first vehicle's onboard navigation system and the road infrastructure status identified by the vision system. The controller can extract lane-cutting intent classification parameters from the vehicle information and road information. For example, the controller can obtain the turn signal activation sign and flashing time of the second vehicle from the forward-facing camera image through target detection and tracking, thereby determining the vehicle light status parameters; the controller can obtain the distance data between the second vehicle and lane lines, gaps, and vehicles in front and behind from the ranging results of millimeter-wave radar and vision fusion, thereby determining spatial index parameters; the controller can calculate the lateral velocity, acceleration, and heading angle of the second vehicle through continuous multi-frame position difference, thereby determining kinematic index parameters; the controller can obtain the front wheel steering angle, steering angular velocity, and wheel end offset rate of the second vehicle through wheel detection and motion analysis algorithms, and calculate the probing characteristics and gap encroachment acceleration of the second vehicle through the difference of longitudinal acceleration and the second derivative of gap distance, thereby determining driving characteristic parameters; the controller can obtain road state parameters through the fusion results of navigation events and visual recognition. The lane-cutting intention classification parameters may include one or more of the following: vehicle light status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road state parameters. In some embodiments, the controller may determine at least one lane-cutting intention classification parameter based on the actual sensor configuration or computing resources.

[0079] The controller can compare and verify the lane-cutting intention classification parameters with the type determination conditions corresponding to the intention intensity range. When the lane-cutting intention classification parameters determine that the type determination conditions are met, if all or part of the lane-cutting intention classification parameters meet the threshold requirements set in the type determination conditions, the controller can determine the lane-cutting intention type corresponding to the intention intensity range as the lane-cutting intention type of the second vehicle. In some embodiments, when the lane-cutting intention classification parameters determine that the type determination conditions are not met, the controller can return to the process of obtaining lane-cutting intention information. For example, for the malicious lane-cutting type corresponding to intention intensity range A, its type determination conditions may include the headlight state parameter T=0, kinematic index parameters, etc. >5 Spatial index parameters Multiple logical rules, such as <1.5m, apply if the controller determines that the headlight status parameter in the lane-cutting intention classification parameters is off (T=0) and the lateral speed... It is 6.2 lateral acceleration It is 2.5 Horizontal distance If the value is 1.1m, then the cut-in intention classification parameter completely matches the threshold range of the malicious cut-in condition, and the controller can determine the cut-in intention type of the second vehicle as malicious cut-in.

[0080] In this exemplary embodiment, fine-grained verification of the lane-cutting intention type is performed based on at least one of the following: vehicle headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road state parameters. This can improve the accuracy of lane-cutting intention type identification.

[0081] In an exemplary embodiment, determining the type of the second vehicle's intention to cut in line based on the intensity of the intention and vehicle information includes: when the probability of cutting in line exceeds a preset probability threshold, determining the type of the second vehicle's intention to cut in line based on the intensity of the intention and vehicle information.

[0082] The preset probability threshold can be used to determine whether to trigger the classification processing of the intention to cut in line. The controller only initiates the classification processing of the cutting intention type based on the intensity of the intention and vehicle information when the probability of cutting in line exceeds the threshold. The value of the probability threshold can be determined through real-vehicle calibration, simulation optimization, or online adaptive learning, and can be, for example, 0.25, 0.28, or 0.32. In some embodiments, the probability threshold can be dynamically determined based on the driving scenario (city, highway, weather, etc.) of the first and second vehicles. For example, a threshold mapping table can be pre-built, which can record the corresponding calibrated probability thresholds for different driving scenarios. The controller can determine the corresponding probability threshold from the threshold mapping table based on the driving scenario (city, highway, weather, etc.) of the first and second vehicles to determine whether to trigger the classification processing of the cutting intention type.

[0083] For example, the controller can determine a preset probability threshold and compare the cutting-in probability in the cutting-in intention information of the second vehicle with the probability threshold. If the cutting-in probability is greater than the probability threshold, the controller can perform processing to determine the cutting-in intention type of the second vehicle based on the cutting-in intention strength and vehicle information. In some embodiments, when the cutting-in probability does not exceed the probability threshold, the controller can maintain the current state, for example, continue to observe or not perform any classification processing of cutting-in intention types.

[0084] In this exemplary embodiment, when the probability of cutting in line exceeds a preset probability threshold, the processing to determine the type of cutting in line can reduce unnecessary calculations in scenarios where the probability of cutting in line is extremely low. This can reduce the consumption of computing resources while ensuring the timeliness and accuracy of the determination of the type of cutting in line.

[0085] In one exemplary embodiment, such as Figure 3As shown, based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category of the second vehicle, a control strategy is determined for the first vehicle, including:

[0086] Step S301: Determine the intensity adjustment weight based on the type of cutting-in intention, and determine the cutting-in permission coefficient based on the vehicle category to which the second vehicle belongs; the cutting-in permission coefficient is used to characterize the priority of the first vehicle to yield to the second vehicle;

[0087] Step S302: Determine the intensity of the game of cutting in line based on the intensity adjustment weight, the cutting in line permission coefficient, the intensity of the cutting in line intention, and the cutting in line probability;

[0088] Step S303: Determine the control strategy for the first vehicle based on the strength of the lane-jumping game.

[0089] The intensity adjustment weight is used to correct the intensity of the cutting-in intention. For example, it can amplify, maintain, or weaken the intensity of the cutting-in intention directly determined based on the vehicle information of the second vehicle, according to the nature of the cutting-in behavior corresponding to the type of cutting-in intention. The correspondence between cutting-in intention types and intensity adjustment weights can be pre-configured. For example, a mapping table can be used to store the correspondence between different cutting-in intention types and different intensity adjustment weights. By querying this mapping table, the intensity adjustment weight corresponding to the cutting-in intention type can be determined. For example, malicious cutting-in corresponds to an intensity adjustment weight greater than 1 to amplify its threat; active cutting-in corresponds to an intensity adjustment weight of 1; passive cutting-in corresponds to an intensity adjustment weight less than 1 to weaken its threat; and accidental cutting-in corresponds to an intensity adjustment weight much less than 1 to minimize its impact.

[0090] The lane-cutting permission coefficient represents the priority of the first vehicle yielding to the second vehicle. A higher coefficient indicates a lower priority, meaning the first vehicle is less likely to yield. Conversely, a lower coefficient indicates a higher priority, meaning the first vehicle is more likely to yield. The coefficient quantifies the yielding priority of different vehicle categories in lane-cutting scenarios. Different vehicle categories correspond to different coefficients, allowing for targeted anti-lane-cutting strategies against second vehicles belonging to different categories. For example, the coefficient can range from 0 to 3. 0 indicates the highest priority for yielding, meaning the second vehicle has the highest priority and the first vehicle must yield; 3 indicates the lowest priority, meaning the second vehicle has the lowest priority and the first vehicle can engage in the strongest possible lane-cutting strategy. The correspondence between different vehicle categories and different lane-cutting permission coefficients can be pre-configured. For example, for vulnerable traffic participants such as two-wheeled vehicles, the lane-cutting permission coefficient can be set to 0, indicating that the first vehicle should give way unconditionally; for passenger cars, the lane-cutting permission coefficient can be set to 1, indicating that the first vehicle can engage in standard anti-lane-cutting game; for large vehicles, considering their blind spots and braking distance, the lane-cutting permission coefficient can be set to a lower value, indicating that the first vehicle needs to give way in a timely manner to reduce confrontation and ensure driving safety.

[0091] The intensity of the lane-jumping game characterizes the degree of resistance or yielding that a first vehicle should take in response to a lane-jumping event triggered by a second vehicle. The intensity of the lane-jumping game can be determined comprehensively based on multi-dimensional information including intensity adjustment weights, lane-jumping permission coefficients, lane-jumping intent intensity, and lane-jumping probability. The intensity of the lane-jumping game can include continuous values, normalized to between 0 and 1. A higher intensity value indicates that the first vehicle needs to take stronger countermeasures to defend its right-of-way; a lower intensity value indicates that the first vehicle should be more inclined to yield or maintain the status quo. Different lane-jumping game intensities can correspond to different control strategies. The correspondence between lane-jumping game intensity and control strategy can be pre-configured according to actual needs, and the controller can also generate corresponding control strategies in real time based on the lane-jumping game intensity.

[0092] For example, the controller can determine the intensity adjustment weight and the lane-jumping permission coefficient based on the type of lane-jumping intent and the vehicle category to which the second vehicle belongs, respectively. For instance, the controller can determine the intensity adjustment weight based on the type of lane-jumping intent and the lane-jumping permission coefficient based on the vehicle category to which the second vehicle belongs by looking up a mapping table. In some embodiments, the controller can also calculate the corresponding intensity adjustment weight in real time based on the type of lane-jumping intent using a preset mapping function. For example, the controller can determine the degree of malice of the second vehicle's lane-jumping based on the type of lane-jumping intent and calculate an intensity adjustment weight greater than or less than 1 based on this degree of malice. The controller can also use a pre-trained neural network model, taking the type of lane-jumping intent as input, and directly outputting the corresponding intensity adjustment weight. This neural network model can be pre-trained using pre-labeled sample data (samples of lane-jumping intent types - samples of intensity adjustment weights). In some embodiments, the lane-jumping permission coefficient can also be calculated using a formula bound to the vehicle category. The lane-jumping permission coefficient reflects the degree of concession that the first vehicle should give to vehicles of that category.

[0093] The controller can calculate the intensity of the lane-jumping game by comprehensively considering the intensity adjustment weights, the lane-jumping permission coefficient, the lane-jumping intention intensity, and the lane-jumping probability. For example, the controller can use a pre-trained model, such as at least one of various models including decision trees, fuzzy inference systems, or artificial neural networks, and input these four parameters into the model so that the model maps the intensity adjustment weights, the lane-jumping permission coefficient, the lane-jumping intention intensity, and the lane-jumping probability to a lane-jumping game intensity between 0 and 1. This model can be pre-trained using pre-labeled sample data (intensity adjustment weight samples - lane-jumping permission coefficient samples - lane-jumping intention intensity samples - lane-jumping probability samples - lane-jumping game intensity samples).

[0094] After determining the intensity of the cut-off game, the controller determines a specific and executable control strategy based on the intensity. In some embodiments, the controller can employ a hierarchical mapping method, presetting multiple game intensity thresholds to categorize consecutive cut-off game intensities into discrete game levels. Each game level can correspond to a pre-defined control strategy, which may include a set of lateral and longitudinal control behaviors, such as maintaining the original speed in a weak game, decelerating longitudinally in a medium game, and decelerating longitudinally while adding lateral micro-motion in a strong game. In some embodiments, the controller can also use a continuous mapping function to directly map the cut-off game intensity to specific control parameters. For example, the magnitude of longitudinal deceleration may be proportional to the cut-off game intensity, or the lateral displacement amplitude may be non-linearly proportional to the cut-off game intensity. The controller can obtain corresponding control strategies based on various control parameters.

[0095] In this exemplary embodiment, the intensity adjustment weight and the lane-jumping permission coefficient are determined based on the lane-jumping intention type and the vehicle category to which the second vehicle belongs, respectively. The lane-jumping game intensity is determined by comprehensively considering the intensity adjustment weight, the lane-jumping permission coefficient, the lane-jumping intention intensity, and the lane-jumping probability. The control strategy is then determined based on the lane-jumping game intensity. This approach can comprehensively consider multi-dimensional information such as discrete lane-jumping intention types and vehicle categories, as well as continuous lane-jumping intention intensity and lane-jumping probability, to determine the control strategy, ensuring its pertinence and accuracy.

[0096] In one exemplary embodiment, such as Figure 4 As shown, the strength of the lane-jumping game is determined based on the strength adjustment weight, the lane-jumping permission coefficient, the strength of the lane-jumping intention, and the lane-jumping probability, including:

[0097] Step S401: Adjust the intensity of the intention to cut in line by adjusting the intensity adjustment weight to obtain the adjusted intensity of the intention to cut in line;

[0098] Step S402: Determine the game weighting based on the driving scenarios of the first and second vehicles;

[0099] Step S403: According to the game weighting, the adjusted intention to cut in line intensity, the cutting in line authority coefficient, and the cutting in line probability are fused to obtain the cutting in line game intensity.

[0100] The driving scenario can be used to describe the current road environment and traffic conditions of the first and second vehicles. The driving scenario can be divided based on at least one of the following factors: vehicle speed, road type, and traffic density. For example, the driving scenario can include urban congestion scenarios (such as urban roads or congested sections where the first vehicle's speed is less than or equal to 60 km / h) and expressway scenarios (such as highways or urban expressways where the first vehicle's speed is greater than or equal to 60 km / h). In some embodiments, the controller can acquire scene information such as the first vehicle's speed signal, road type markings on the navigation map, and lane environment identified by the camera, and determine the current driving scenario based on the scene information according to a preset speed threshold or road type label.

[0101] The game-theoretic weighting can include weighting coefficients for the adjusted intensity of the intention to cut in line, the coefficient of permission to cut in line, and the probability of cutting in line, used to adjust the proportion of each factor in the intensity of the game-theoretic competition for cutting in line. The game-theoretic weighting can correspond to the driving scenarios of the first and second vehicles; different driving scenarios can correspond to different game-theoretic weightings. For example, in an urban scenario, the game-theoretic weighting can include the weight corresponding to the coefficient of permission to cut in line. Adjusted weights corresponding to the intensity of the intention to cut in line Weights corresponding to the probability of cutting in line In high-speed scenarios, game weighting can include... , , The game-theoretic weighting can be calibrated based on real vehicle data or determined through machine learning optimization.

[0102] Optionally, the controller can adjust the intensity of the cutting-in intention based on the intensity adjustment weight. For example, the intensity of the cutting-in intention can be multiplied by the intensity adjustment weight to obtain the adjusted intensity of the cutting-in intention. The intensity adjustment weight can correct the intensity of the cutting-in intention, such as amplifying malicious cutting-in intentions and weakening passive and accidental cutting-in intentions, making the impact of different cutting-in intention types on the intensity of the cutting-in intention more reasonable. The controller can determine the current driving scenario of the first and second vehicles. For example, the controller can determine the driving scenario based on scenario information such as the state of the first vehicle and environmental information, and the controller can determine the corresponding game-theoretic weighting weight based on this driving scenario. The controller can then weight and fuse the adjusted cutting-in intention intensity, cutting-in permission coefficient, and cutting-in probability according to the game-theoretic weighting weight to obtain the cutting-in game intensity.

[0103] In this exemplary embodiment, the intensity of the intention to cut in line is adjusted by adjusting the intensity weight, so that the intensity of the game of cutting in line is precisely matched with the actual threat level of the second vehicle, thereby improving the humanization and safety of the decision-making process. By determining the game weighting weight according to the driving scenario, the intensity of the game of cutting in line can be adaptively determined according to the driving scenario, which can take into account both traffic efficiency and driving safety, thereby ensuring the pertinence and reliability of the intensity of the game of cutting in line.

[0104] In an exemplary embodiment, determining the control strategy for the first vehicle based on the intensity of the lane-jumping game includes: determining the game intensity range where the lane-jumping game intensity is located based on a preset game intensity threshold; and determining the preset control strategy that conforms to safety constraints corresponding to the game intensity range as the control strategy for the first vehicle.

[0105] The game intensity threshold is used to quantify the intensity of consecutive game-changing moves into non-overlapping game intensity intervals, thereby mapping them to different control strategies. The specific value of the game intensity threshold can be pre-determined using empirical data or machine learning. The game intensity interval is a numerical range determined by the game intensity threshold, and each interval can correspond one-to-one with a preset control strategy. For example, the game intensity threshold can include three values: 0.15, 0.35, and 0.85. The intensity of the game of cutting in line is divided into four intensity intervals: [0, 0.15), [0.15, 0.35), [0.35, 0.85), and [0.85, 1.0]. When the controller is positioned within the game intensity range of [0.15, 0.35), it determines the control strategy corresponding to this range as the control strategy for the first vehicle. The preset control strategies for each game intensity range can comply with safety constraints, including mandatory safety boundary conditions that the first vehicle must adhere to when performing game actions. For example, safety constraints may include, but are not limited to, at least one of the following constraints: longitudinal distance constraint (the distance between the first vehicle and the vehicle in front must not be lower than the safe following distance), speed constraint (the speed of the first vehicle must not exceed the cruising speed), and lateral distance constraint (the lateral distance between the first vehicle and adjacent vehicles or lane boundaries must not be lower than the preset safe interval). The safety constraints of the control strategy can be sensed in real time by sensors or preset by the driver.

[0106] Optionally, the controller can acquire a preset game strength threshold and compare the intensity of the lane-jumping game with the threshold to determine the game strength range within which the lane-jumping game intensity falls. After determining the game strength range, the controller can query a pre-stored strategy library for a control strategy that matches the game strength range and complies with safety constraints, and use this control strategy as the control strategy for the first vehicle. For example, based on the game strength threshold, the controller determines the game strength range for the lane-jumping game as range A, and the preset control strategy for range A that complies with safety constraints is "weak game: maintain the original speed, do not actively create space, and interact normally according to traffic rules." The controller can then use this control strategy as the control strategy for the first vehicle and control its movement accordingly, ensuring the vehicle maintains its current speed.

[0107] In this exemplary embodiment, the game intensity range of the lane-jumping game intensity is determined according to a preset game intensity threshold, and the preset control strategy corresponding to the game intensity range and conforming to safety constraints is determined as the control strategy for the first vehicle. This can ensure that the control strategy can match the real lane-jumping scenario and effectively balance the lane-jumping game intensity and vehicle driving safety.

[0108] In an exemplary embodiment, the vehicle control method further includes: when the cut-in intention type is a misoperation cut-in type, controlling the first vehicle to issue a cut-in misoperation reminder to the second vehicle; and when the second vehicle continues to cut in, returning to the step of determining the cut-in intention type of the second vehicle based on the cut-in intention intensity and vehicle information.

[0109] Among these, the "misoperation-based lane-cutting" type indicates that the second vehicle did not intentionally change lanes or seize space, but rather briefly encroached on the lane occupied by the first vehicle due to operational error, distraction, or a brief deviation. In this case, the threat posed by the second vehicle cutting in is minimal. The lane-cutting misoperation alert can include a warning interactive operation targeting the second vehicle's misoperation, prompting it to correct its lane departure promptly, while ensuring the first vehicle does not engage in any confrontational actions, thus guaranteeing driving safety. In some embodiments, the alert method for lane-cutting misoperation can be adaptively set according to environment, regulations, and vehicle configuration, such as being based on at least one of light signals or sound signals.

[0110] Optionally, when the controller determines that the second vehicle's intention to cut in is a mis-cutting maneuver, it can issue a mis-cutting warning to the second vehicle to signal its awareness of vehicle deviation, prompting it to check its driving operation and correct its course as soon as possible. In some embodiments, the controller can generate a warning instruction, which may include at least one of the following parameters: warning method, frequency, and duration. These parameters can be pre-calibrated. The controller can send this warning instruction to the first vehicle, controlling it to issue a mis-cutting warning to the second vehicle according to the instruction. For example, it can control the first vehicle to flash its headlights a preset number of times and, depending on environmental conditions, to sound its horn briefly. While the first vehicle is issuing the mis-cutting warning, the controller can maintain the first vehicle's current following distance and refrain from performing any countermeasures such as reducing the following gap or making lateral micro-movements.

[0111] After the first vehicle issues a warning to the second vehicle for its erroneous cutting-in maneuver, the controller can continuously monitor the second vehicle to determine whether it has corrected its mistake. If the second vehicle corrects its maneuver, the controller may not apply anti-cutting control to the first vehicle, allowing it to continue driving normally. If the second vehicle fails to correct its maneuver, i.e., continues to cut in, the controller can return to the process of determining the second vehicle's cutting-in intention type based on the strength of the cutting-in intention and vehicle information, thereby re-determining the second vehicle's cutting-in intention type.

[0112] In some embodiments, after the first vehicle issues a lane-cutting misoperation warning, the controller can initiate an observation period. During this period, the controller can continuously monitor the second vehicle's information, such as its vehicle position sequence and lateral speed, to determine whether the second vehicle is continuing to cut in. For example, if the second vehicle continues to encroach on the lane occupied by the first vehicle after the lane-cutting misoperation warning is issued, without showing any signs of straightening or deceleration, it is considered that the second vehicle is continuously cutting in. In this case, the controller believes that the initial misoperation lane-cutting determination may be inaccurate and needs to be reassessed. The controller can return to the process of determining the lane-cutting intent type to re-determine the lane-cutting intent type of the second vehicle. The new lane-cutting intent type may become active or malicious lane-cutting, and the controller can determine the control strategy for the first vehicle based on the new lane-cutting intent type. Furthermore, if the second vehicle straightens or decelerates during the observation period, indicating that the second vehicle is not continuously cutting in, the controller determines that the lane-cutting was indeed a misoperation, and the controller can control the first vehicle to drive normally.

[0113] In this exemplary embodiment, when a erroneous lane-cutting type is identified, the first vehicle is controlled to issue a lane-cutting error warning. When the second vehicle continues to cut in, the lane-cutting intention type is redefined. This reduces unnecessary anti-lane-cutting control intervention and ensures the accuracy of lane-cutting intention type identification, thereby improving the accuracy of vehicle control in complex driving scenarios.

[0114] In one exemplary embodiment, the vehicle control method further includes: if the second vehicle belongs to a special vehicle category, determining the control strategy as active yielding.

[0115] Special vehicles can include those that are legally entitled to priority passage due to performing emergency tasks, and which other vehicles are legally obligated to yield to. These include, but are not limited to, fire trucks, ambulances, and police cars. Proactive yielding is a control strategy to ensure priority passage for special vehicles; that is, the first vehicle does not engage in a race-to-the-line game against the special vehicle, but rather unconditionally yields.

[0116] For example, when the second vehicle is identified as a special vehicle, i.e., the second vehicle is a special vehicle, the controller can directly determine the control strategy as active yielding, so as to control the first vehicle to actively yield to the second vehicle, so as to ensure the passage efficiency of special vehicles.

[0117] In this exemplary embodiment, when the second vehicle is a special vehicle, controlling the first vehicle to actively give way can ensure the passage efficiency of special vehicles.

[0118] In an exemplary embodiment, the vehicle control method further includes: predicting a cutting-in intention based on wheel-end motion information and cutting-in trend prediction information included in the vehicle information, thereby obtaining cutting-in intention information, wherein the cutting-in trend prediction information is determined based on the historical driving information of the second vehicle.

[0119] Among them, wheel-end motion information is used to characterize the motion state of the second vehicle's wheel level, such as including but not limited to the second vehicle's front wheel steering angle (the angle of the front wheel relative to the vehicle body, which may be a precursor to cutting in when the direction is towards the lane), front wheel steering angular velocity (the rate of change of the front wheel steering angle, the larger the rate, the more urgent the intention to cut in), and wheel-end lateral offset rate (the instantaneous approach rate of the wheel edge to the lane line, which can be determined by the wheel edge position of multiple consecutive frames).

[0120] The lane-cutting trend prediction information can be used to assess the tendency of a second vehicle to intrude into the lane occupied by a first vehicle. This information may include, but is not limited to, wheel track offset trends (by fitting a predicted wheel track line to the left and right front wheel position coordinates over multiple consecutive frames, calculating the intrusion distance between this predicted line and the lane line of the first vehicle; a larger distance indicates a clearer intention to cut in), probing characteristics (the rate of change of the second vehicle's longitudinal acceleration, reflecting the urgency of the second vehicle's slight acceleration or braking to test the gap before cutting in; a larger absolute value indicates a more pronounced probing), and gap-occupying acceleration (the velocity gradient of the second vehicle moving towards the gap between the first vehicle and the vehicle in front, reflecting the intensity of its willingness to seize the gap). The lane-cutting trend prediction information can be determined based on the second vehicle's historical driving information, which records the second vehicle's motion state. This historical driving information may include at least one of various continuously monitored data such as the second vehicle's position, speed, acceleration, heading angle, and wheel status.

[0121] For example, the vehicle information may include wheel-end motion information and lane-cutting trend prediction information of the second vehicle. The controller can predict the lane-cutting intention based on the wheel-end motion information and the lane-cutting trend prediction information to obtain the lane-cutting intention information of the second vehicle. In some embodiments, the controller can predict the lane-cutting intention based on the wheel-end motion information and the lane-cutting trend prediction information of the second vehicle using a pre-trained lane-cutting intention prediction model to obtain the lane-cutting intention information of the second vehicle. The lane-cutting intention prediction model may include pre-labeled logical rules or a trained neural network model, which can be trained based on pre-labeled sample data (wheel-end motion information samples - lane-cutting trend prediction information samples - lane-cutting intention information samples). For example, the lane-cutting intention prediction model may include at least one of various neural network models such as Long Short-Term Memory (LSTM), Temporal Convolutional Network (TCN), Transformer model based on self-attention mechanism, and CNN-LSTM (Convolutional Neural Network-Long Short-Term Memory) hybrid model.

[0122] In some embodiments, wheel-end motion information may include the front wheel steering angle, front wheel steering angular velocity, and wheel-end lateral offset rate of the second vehicle; lane-cutting trend prediction information may include the wheel track offset trend, abruptness, and clearance encroachment rate of the second vehicle. Furthermore, lane-cutting intention prediction can be performed separately based on wheel-end motion information and lane-cutting trend prediction information, and the respective lane-cutting intention prediction results can be combined to obtain lane-cutting intention information. For example, the controller can input wheel-end motion information into a first lane-cutting intention prediction model to obtain a lane-cutting intention prediction result SA corresponding to the wheel-end motion information. The first lane-cutting intention prediction model can be trained based on pre-labeled sample data (wheel-end motion information samples - lane-cutting intention prediction result samples). The controller can input lane-cutting trend prediction information into a second lane-cutting intention prediction model to obtain a lane-cutting intention prediction result SB corresponding to the lane-cutting trend prediction information. The controller can combine the lane-cutting intention prediction results SA and SB to obtain the lane-cutting intention information of the second vehicle, which may include the lane-cutting intention intensity and lane-cutting probability of the second vehicle.

[0123] In some embodiments, the vehicle information of the second vehicle may further include headlight information, kinematic information, and spatial position information. Headlight information may include the turn signal activation indicator of the second vehicle and the duration of turn signal flashing; kinematic information may include the longitudinal velocity, lateral velocity, lateral acceleration of the second vehicle, and the angle between the vehicle heading angle of the second vehicle and the centerline of the adjacent lane; spatial position information may include the vertical distance between the center of the second vehicle and the target lane line of the adjacent lane, and the lateral distance between the second vehicle and the target clearance; the target clearance is the safe clearance between the first vehicle and the vehicle in front of the first vehicle.

[0124] The controller can also predict lane-cutting intentions based on vehicle headlight information, kinematic information, and spatial position information, respectively, obtaining lane-cutting intention prediction results SC, SD, and SE. In some embodiments, for each of the vehicle headlight information, kinematic information, and spatial position information, a corresponding lane-cutting intention prediction model can be pre-trained, such as by training on pre-labeled sample data (vehicle information samples - lane-cutting intention prediction result samples), so that lane-cutting intention prediction can be achieved through the corresponding lane-cutting intention prediction model. The controller can combine the lane-cutting intention prediction results SA, SB, SC, SD, and SE to obtain the lane-cutting intention information of the second vehicle.

[0125] For example, the controller can combine the cut-in intention prediction results SA, SB, SC, SD, and SE to determine the cut-in probability of the second vehicle; and based on the cut-in probability of the second vehicle, the front wheel steering angular velocity of the second vehicle included in the wheel end motion information, the gap encroachment rate of the second vehicle included in the cut-in trend prediction information, and the lateral velocity of the second vehicle, determine the cut-in intention intensity of the second vehicle, thereby obtaining the cut-in intention information of the second vehicle.

[0126] In this exemplary embodiment, the intention to cut in line is predicted by combining the wheel end motion information and the cutting-in trend prediction information in the vehicle information, so as to obtain the cutting-in intention information of the second vehicle. This can accurately capture the cutting-in intention before the body movement of the second vehicle obviously shows the cutting-in action, which can improve the accuracy and timeliness of cutting-in intention recognition.

[0127] This application also provides an application scenario in which the above-described vehicle control method is applied. Specifically, the vehicle control method is applied in this scenario as follows:

[0128] In lane-cutting scenarios, accurately understanding the behavioral intentions and physical attributes of adjacent vehicles, and formulating differentiated response strategies accordingly, is key to achieving safe, efficient, and humane lane-cutting prevention. However, related advanced driver assistance technologies face the following technical challenges at the decision-making level:

[0129] 1. The classification of cut-off intentions is incomplete, and the processing logic is rather coarse.

[0130] Current anti-cutting solutions only address the binary issue of "whether or not someone cuts in," failing to differentiate the nature of such behavior. However, actual road cutting can manifest in drastically different ways, such as malicious cutting, active cutting, passive cutting, and accidental cutting. These different types of cutting pose entirely different threats to the vehicle, require entirely different response strategies, and necessitate entirely different interaction methods. Current technologies employ a "one-size-fits-all" approach, applying the same acceleration to prevent or deceleration to all cutting-in attempts.

[0131] 2. Ignoring the types of vehicles cutting in line poses a significant safety risk.

[0132] The existing solutions lack precise identification and targeted processing of the physical categories of vehicles cutting in line. Different types of vehicles (such as two-wheeled vehicles, three-wheeled vehicles, passenger cars, large vehicles, and special vehicles) have drastically different motion characteristics, risk features, and legal status, but the relevant technologies often treat them all as "obstacles," which poses serious safety hazards.

[0133] 3. The game strategy is too simplistic and lacks dynamic fusion decision-making.

[0134] The relevant technologies do not involve comprehensive decision-making based on multi-dimensional information such as the intensity of the intention to cut in line, the probability of cutting in line, and vehicle type.

[0135] Based on this, the vehicle control method provided in this application determines the type of cutting-in intention based on the intensity of cutting-in intention in the cutting-in intention information and the vehicle information of the second vehicle. It then determines the control strategy of the first vehicle by combining the type of cutting-in intention, the intensity of cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs. The method then controls the first vehicle to drive according to the control strategy. The type of cutting-in intention can reflect the nature of the cutting-in behavior of neighboring vehicles, and the vehicle category can distinguish the differentiated safety risks and traffic priorities of vehicles with different physical attributes. Thus, it can simultaneously take into account the nature, intensity, probability, and vehicle category of cutting-in behavior in complex cutting-in scenarios, flexibly adjust the corresponding control strategy, and make the control strategy match the real cutting-in scenario, thereby achieving an effective balance between the intensity of the cutting-in game and the safety of vehicle driving.

[0136] In this application scenario, the vehicle control method provided in this application can determine the type of cutting-in intention of the adjacent vehicle (the second vehicle) through a pre-trained cutting-in intention classification model. The input to the cutting-in intention classification model may include:

[0137] 1. Information on the cutting-in intentions of vehicles in adjacent lanes that are adjacent to the lane occupied by the vehicle (the first vehicle);

[0138] Information on intention to cut in line includes the probability of cutting in line. and the intensity of the intention to cut in line The intention to cut in line can be determined based on the vehicle information detected for neighboring vehicles. For example, a pre-trained early prediction model for the intention to cut in line can be used to predict the intention to cut in line based on the vehicle information of neighboring vehicles, so as to obtain the intention to cut in line information of neighboring vehicles.

[0139] 2. Classification parameters for the intention to cut in line;

[0140] The parameters for classifying lane-cutting intent can be determined based on vehicle information of neighboring vehicles and road information of the surrounding road. This vehicle and road information can be collected by various sensors installed on the vehicle (such as onboard forward-facing cameras and surround-view cameras). The parameters for classifying lane-cutting intent can include vehicle headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, and road condition parameters. Among these:

[0141] Vehicle headlight status parameters are used to capture lane change / cut-off signals actively transmitted by adjacent vehicles, improving prediction accuracy and reducing misjudgments. These parameters can be collected by sensors and determined based on visual detection (a forward-facing camera identifies the on / off state of adjacent vehicles' turn signals). Vehicle headlight status parameters may include:

[0142] Turn signal activation indicator: T∈{0,1}, where T=1 indicates that the adjacent vehicle has activated the turn signal for the corresponding lane (left turn / right turn, matching the direction of cutting in), and T=0 indicates that the turn signal is not activated;

[0143] Turn signal flashing duration The unit is ms, which refers to the duration of continuous flashing of the turn signal, reflecting the clarity of the adjacent vehicle's intention to change lanes / cut in (the longer the flashing time, the clearer the intention).

[0144] Space index parameters are used to determine whether adjacent vehicles have the space to cut in, and may include:

[0145] Lateral distance between adjacent vehicles and the vehicle in front. This refers to the lateral distance from the center of the adjacent vehicle to the "safe gap between your vehicle and the vehicle in front." The smaller the distance, the closer the adjacent vehicle is to the point where it can cut in. Based on sensor-collected data, the lateral distance between adjacent vehicles and the gap between the vehicle and the vehicle in front can be determined by visual detection.

[0146] Traffic flow environment parameters around neighboring vehicles Unit: vehicles, refers to the number of vehicles within 20m in front of and behind the adjacent vehicle, reflecting whether the adjacent vehicle is forced to cut in. Based on the data collected by sensors, visual detection algorithms can be used to count and identify targets around neighboring vehicles;

[0147] Distance between adjacent car and the car in front Distance between adjacent and following vehicles Unit: m, to determine whether a neighboring vehicle is forced to cut in due to pressure from vehicles in front and behind; and It can determine the distance between a neighboring vehicle and the vehicle directly in front of it by visually detecting the distance based on the data collected by the sensors;

[0148] Kinematic parameters are used to determine the cutting-in motion of adjacent vehicles, and may include:

[0149] Lateral speed of adjacent vehicle (Along the perpendicular direction of the lane) reflects the rate at which adjacent vehicles deviate from the lane of the vehicle; It can be determined by tracking and calculating based on multi-frame data collected by sensors;

[0150] Lateral acceleration of adjacent vehicles (Rate of change of lateral velocity) reflects the urgency of the adjacent vehicle's deflection action; Based on The time difference was calculated.

[0151] Adjacent vehicle lane change angle The unit is °, which refers to the angle between the vehicle body of the adjacent vehicle and the center line of the lane when the adjacent vehicle cuts into the lane, reflecting the aggressiveness of the cutting-in action; The focus is on the moment when the adjacent vehicle cuts into the lane of the vehicle, which expresses the execution phase of cutting in and directly reflects the aggressiveness of the cutting-in action; Based on the data collected by sensors, the angle between the axis of the adjacent vehicle and the center line of the lane of the vehicle can be determined by visual detection.

[0152] Driving characteristic parameters are used to describe the driving behavior and trends exhibited by the second vehicle before or during cutting in, and may include:

[0153] Neighboring vehicle probing characteristics (unit: (), represents jerk, and the formula is as follows:

[0154]

[0155] in, The longitudinal acceleration of the adjacent vehicle reflects the driver's subtle probing actions (slight throttle / brake test gap) before cutting in. The larger the value of |, the more obvious the neighboring car's attempt to cut in is, and the higher the probability of cutting in.

[0156] Gap encroachment acceleration The formula is as follows:

[0157]

[0158] Gap encroachment acceleration This refers to the speed gradient of a neighboring vehicle moving towards the "vehicle-to-front gap," directly reflecting the intensity of the neighboring vehicle's intention to "occupy the gap." The larger the absolute value (negative, indicating a decrease in distance), the more urgent the intention to cut in line.

[0159] Behavioral characteristic parameters: After intruding into the lane of a vehicle, the vehicle quickly returns to its original lane (determined by vehicle position tracking over multiple consecutive frames), without any continuous encroachment gaps. This may include the time taken for a neighboring vehicle to quickly return to its original lane.

[0160] Road state parameters are used to describe the road conditions of the vehicle and neighboring vehicles, and may include:

[0161] Road facilities and traffic status parameters , which are discrete values ​​(0 = normal road, 1 = lane merging scenario (e.g., lane 3 becomes 2, adjacent lane disappears), 2 = adjacent lane is under construction / accident). In some embodiments, it can be determined by vehicle navigation information + forward / surround view camera recognition. For example, the navigation information can provide road warnings such as lane merging and construction, and the camera can help identify features such as lane line disappearance, construction cones, and accident vehicles. The two are fused to determine the type of lane-cutting intention. For example, lane-cutting caused by lane merging or adjacent lane obstruction is more likely to be passive lane-cutting.

[0162] The vehicle's controller can classify cutting-in intentions based on the intensity of the intention and vehicle information to determine the type of cutting-in intention from neighboring vehicles. The parameters for classifying various cutting-in intention types can be calibrated using real-vehicle data or optimized through extensive training. For example, the classification of cutting-in intention types can be as follows:

[0163] Category 1: Malicious lane cutting

[0164] This refers to a situation where a neighboring vehicle, without a reasonable reason or to signal a lane change, actively and quickly cuts into the gap between the vehicle in front, disregarding the driver's own driving status. This behavior is clearly aggressive and coercive, and easily leads to a minor collision. The criteria for this classification may include:

[0165] Strength of the intention to cut in line ≥0.8, and meets any of the following type determination conditions:

[0166] 1. Headlight status parameters: T=0 (corresponding turn signal not activated);

[0167] 2. Kinematic parameters: >5 (High lateral speed) >2 (Large lateral acceleration) >15° (large entry angle, aggressive entry);

[0168] 3. Spatial index parameters: <1.5m (narrow gap, forced entry) ≤2 (Sparse traffic in the surrounding area, no compelling conditions) >3m (The adjacent vehicle has enough space in front of it, so there is no need to cut in).

[0169] 4. Driving characteristic parameters: | |<0.5 (No obvious probing actions, directly and forcefully cutting in) < (The gap encroachment acceleration is fast, and the gap is quickly seized.)

[0170] Category 2: Actively cutting in line

[0171] This refers to a situation where the adjacent vehicle clearly intends to change lanes or cut in front, actively signaling the intention to do so, finding a reasonable gap between itself and the vehicle in front, without any obvious forceful cutting in, and the cutting-in action is relatively smooth; if the gap is insufficient, it will abandon the attempt to cut in. The criteria for this classification may include simultaneously meeting the following conditions:

[0172] 1. 0.4≤ <0.8;

[0173] 2. Headlight status parameters: T=1 (Turn on the corresponding turn signal). ≥500ms (sufficient flashing time to effectively convey lane change intention);

[0174] 3. Spatial index parameters: 1.5m≤ ≤3m (the gap is reasonable and conditions for insertion are met). =3 to 5 (moderate traffic flow in the surrounding area, with demand for lane changing) ≤2m (The adjacent vehicle has insufficient space in front of itself, so it is necessary to cut in).

[0175] And any one of them satisfies:

[0176] 1. Kinematic parameters: 2 ≤ ≤5 (Moderate lateral speed), 0.5 ≤ ≤2 (Moderate lateral acceleration), 5°≤ ≤15° (moderate entry angle, distinct from those reflecting the trend of vehicle body offset) );

[0177] 2. Driving characteristic parameters: 0.5 ≤|J|≤1.5 (There are obvious micro-probing movements, and the size of the probing intervals) ≤ ≤ (The gap encroachment acceleration is moderate, and the approach is slow.)

[0178] Category 3: Passive cutting in line

[0179] The adjacent vehicle had no intention of cutting in, but was forced to enter the lane due to being cut off by the vehicle in front, squeezed by the vehicle behind, merging, or sudden construction / accident. The cutting-in action was smooth and without aggressive behavior, but the vehicle was forced to change lanes. The criteria for this classification may include simultaneously meeting the following conditions:

[0180] 1. <0.4;

[0181] 2. Spatial index parameters: ≥3m (sufficient clearance, no need for forced entry) ≥6 (Dense traffic in the surrounding area, creating a compelling situation) <1m (squeezed by the car in front) or <1m (squeezed by the car behind), or =1 (lane merging scenario) =2 (adjacent lane construction / accident);

[0182] 3. Headlight status parameters: T=1 (corresponding turn signal activated), and ≥800ms (long flashing time, fully conveying the lane change intention);

[0183] And any one of them satisfies:

[0184] 1. Kinematic parameters: <2 (Slow lateral speed) <0.5 (Low lateral acceleration) <5° (small entry angle, 2. gentle entry, distinct from those reflecting the trend of vehicle body offset) );

[0185] 2. Driving characteristic parameters: |J| > 1.5 (Frequent minor probing attempts forced the search for space to retreat) > (The gap encroachment acceleration is slow, and it is passively approached.)

[0186] Category 4: Accidental lane cutting (briefly entering the lane from the outside)

[0187] This refers to a situation where a neighboring vehicle has no intention of cutting in, but briefly encroaches into the lane due to driver error (such as accidental steering wheel touch, distraction causing slight vehicle deviation, or driving over the lane lines). The encroachment is not sustained, and the driver quickly corrects the mistake, causing minimal impact on the other vehicle's driving and requiring no complex response strategy. The criteria for this classification may include:

[0188] <0.2 and simultaneously satisfy:

[0189] 1. Kinematic parameters: <1 (Extremely slow lateral speed) <0.2 (Extremely small lateral acceleration) <3° (very small cut-in angle), and the duration of intrusion into the vehicle lane is <1s;

[0190] 2. Spatial index parameters: ≥4m (sufficient clearance, no need to add filler) ≤3 (Sparse traffic in the surrounding area, no compelling conditions) =0 (Normal road, no special scenarios such as lane merging or construction);

[0191] 3. Driving characteristic parameters: After intruding into the vehicle's lane, it quickly returns to its original lane (determined by vehicle position tracking across multiple consecutive frames), with no sustained intrusion intervals; |J| < 0.3 (No obvious micro-probing action; the deviation was due to a sudden erroneous operation.) > (No tendency to actively encroach on gaps).

[0192] In some embodiments, the probability of cutting in line can be... When the value is ≥0.3 (indicating a possible or actual intention to cut in line), the cut-in intention classification model is activated to determine the type of cut-in intention of the adjacent vehicle, thereby reducing unnecessary calculations. In some embodiments, when it is determined that the type determination conditions corresponding to the various cut-in intention types above are met, the cut-in intention type corresponding to the met type determination conditions can be determined as the cut-in intention type of the adjacent vehicle; if the type determination conditions are not met, it can be determined as "suspected cut-in," and monitoring of adjacent vehicles can continue, and cut-in intention classification parameters can be continuously obtained for re-determination.

[0193] In this application scenario, the vehicle's controller can determine a control strategy for the vehicle based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category of the neighboring vehicles, and then control the vehicle's driving according to the control strategy. In some embodiments, vehicle category recognition can be performed for vehicles cutting in, such as by acquiring vehicle images through a forward-facing camera and / or a surround-view camera, and detecting and classifying vehicles cutting in. The vehicle category classification of vehicles cutting in can be shown in Table 1 below.

[0194] Table 1

[0195]

[0196] To quantify the impact of different vehicle categories on the player's game strategy, a lane-cutting permission coefficient can be assigned to each vehicle category. The value ranges from [0,2], with a larger value indicating a higher allowed game intensity. The cut-in permission coefficient settings for various vehicle categories are shown in Table 2 below.

[0197] Table 2

[0198]

[0199] The cut-in permission coefficients for various vehicle categories can be calibrated based on real vehicle data, or optimized and determined after extensive training.

[0200] In some embodiments, the confidence level of vehicle category identification can also be determined. If the confidence level is greater than or equal to a threshold (e.g., <0.5), the identification result of the vehicle category is considered reliable, and can be confirmed according to the logic in Table 2. If the confidence level for class identification is below the threshold (e.g., <0.5), the identification result for that vehicle class is considered potentially unreliable and can be directly discarded. The default value can be 0.5, and the default value can be obtained by pre-calibration.

[0201] In some embodiments, the vehicle controller can determine a control strategy for its own vehicle based on a game-theoretic response strategy generation model, considering the type of cutting-in intention, the strength of the cutting-in intention, the cutting-in probability, and the vehicle category of the neighboring vehicles. The input to the game-theoretic response strategy generation model may include the cutting-in probability. Intensity of the intention to cut in line Types of intent to cut in line (discrete values ​​{malicious, active, passive, misoperation}) and the coefficient of permission to cut in line. The game response strategy generation model can generate the intensity of the game by using a weighted formula. .

[0202] In some embodiments, the cut-off intention type is a discrete value that can participate in the decision-making process as an intensity-adjusted weight, as shown in Table 3 below.

[0203] Table 3

[0204]

[0205] Among them, intensity adjustment weight The value can be calibrated based on real vehicle data, or optimized and determined after extensive training.

[0206] The vehicle's controller can adjust the intensity of the cutting-in intention by adjusting the intensity weight. Essentially, it first knows "someone is about to cut in, how fast and how strong," and then, based on the type of cutting-in intention—whether it's malicious or passive—determines whether the intensity should be amplified, maintained, or weakened. The adjustment formula is as follows:

[0207]

[0208] in, ∈[0,1].

[0209] The vehicle's controller can adjust the intensity of the cut-off intention based on game-theoretic weighting. Interception permission coefficient and the probability of cutting in line By merging the results, we can obtain the strength of the game of adding a lane. The formula can be as follows:

[0210]

[0211] in, , , To assign weights to the game, satisfying The game's weighted average can be calibrated using real-vehicle data. (Intensity of the lane-cutting game) The value range can be [0,1].

[0212] In some embodiments, , , It can be dynamically adjusted according to the driving scenario, such as being divided based on urban / highway scenarios: urban congestion scenario (vehicle speed less than or equal to 60 kph): , , (Emphasizing the intention to cut in line); Highway / expressway scenario (speed greater than or equal to 60 kph): , , (Emphasis on category security). In some embodiments, regardless of , , How should the value be taken, if And if identified as a special vehicle, then mandatory .

[0213] In some embodiments, the strength of the bidding game can be considered. The value of divides the intensity of the game of scrambling into four levels, corresponding to different control strategies, as shown in Table 4 below.

[0214] Table 4

[0215]

[0216] The game intensity thresholds for various game intensity ranges can be calibrated based on real-vehicle data or optimized through extensive training. In some embodiments, a hysteresis interval can be set near the game intensity threshold to reduce frequent changes in the control strategy. For example, for the game intensity of cutting in line near the game intensity threshold... (Strength of the game of adding a lane) (The difference between the game strength threshold and the preset threshold is less than the preset threshold), only the game strength of adding a lane. The strength of the race-jumping game is only updated when the numerical change exceeds a certain threshold and the duration exceeds a certain threshold. The corresponding game intensity range.

[0217] In some embodiments, when the intention to cut in line is determined to be a erroneous cut in line type, regardless of the strength of the cut in line game. The higher or lower the value, the more likely it is to trigger a flashing light / honking alarm:

[0218] Warning method: Flash high beams twice (0.3 seconds apart), and if the environment and regulations permit, honk the horn once (<0.5 seconds).

[0219] During the reminder period: Maintain the current following distance and do not perform any aggressive maneuvers (do not close the following window or make any slight lateral movements);

[0220] Observation period: If the other vehicle returns to its original trajectory or slows down within 3 seconds of being alerted, it is considered a misoperation; if the other vehicle continues to intrude, it is considered a genuine misoperation. In this case, the process re-enters the queue-cutting intent category (potentially changing to active / malicious queue-cutting), and the intent is recalculated based on the new intent type. ;

[0221] After the warning: If the other vehicle returns to its original trajectory or slows down, it is determined to be a misoperation, and normal following resumes; if it is determined to be a genuine operation, the vehicle re-enters the process of classifying the cutting-in intention type (which may be changed to active / malicious cutting-in), and the calculation is recalculated according to the new cutting-in intention type. .

[0222] In some embodiments, the lateral and longitudinal control of the control strategy determined for the vehicle must meet preset safety constraints, which may include, but are not limited to:

[0223] Minimum longitudinal distance: must not be lower than the safe following distance set by the driver;

[0224] Vehicle speed: Must not exceed the cruise speed set by the driver during intelligent driving;

[0225] Minimum lateral distance: must not be less than the safe lateral clearance corresponding to the vehicle category.

[0226] The vehicle control method provided in this application classifies cutting-in intentions into four categories: malicious cutting-in, active cutting-in, passive cutting-in, and misoperated cutting-in. It also sets clear quantitative definitions for each type of cutting-in intention (based on multi-dimensional parameters such as vehicle light status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road state parameters). This refined classification enables accurate understanding of the true intentions of neighboring vehicles, thereby allowing for targeted strengthening of the game against malicious cutting-in, rule-based interaction against active cutting-in, appropriate yielding to passive cutting-in, and priority reminders for misoperated cutting-in.

[0227] The vehicle control method provided in this application also employs differentiated game theory mapping based on vehicle categories, balancing safety and game theory effectiveness. Related technologies uniformly treat different categories of vehicles as "obstacles," ignoring the vulnerability of two-wheeled vehicles, the blind spot risks of large vehicles, and the legal priority of special vehicles, posing serious safety hazards. This application categorizes vehicles that cut in line into five major categories: two-wheeled vehicles, three-wheeled vehicles, passenger cars, large vehicles, and special vehicles, and assigns a cutting-in authority coefficient to each category. (0~2) quantifies the degree of influence of the vehicle's game strategy. This mapping mechanism deeply binds the game strategy with the physical attributes of the vehicle that cuts in line, so that while pursuing traffic efficiency, the bottom line of safety can be firmly maintained.

[0228] The vehicle control method provided in this application can integrate multi-dimensional information to weightedly determine the control strategy, achieving dynamic game intensity calculation. It can organically unify discrete lane-cutting intention types, continuous lane-cutting intention strengths, and vehicle category safety levels, making the game intensity calculation more scientific and precise. Furthermore, by constructing a four-level game intensity system, it achieves a progressive game response from weak to strong. Related technologies adopt a binary "trigger-response" model, either exhibiting strong confrontation or excessive concession, lacking an intermediate state. This application, based on lane-cutting game intensity... The control strategies are divided into four levels: no-game, weak-game, medium-game, and strong-game, and the corresponding control behaviors for each level are clearly defined.

[0229] Related technologies lack the ability to recognize accidental lane-cutting, often treating it as normal lane-cutting, leading to frequent false triggers and abrupt operations. This application, however, specifically addresses accidental lane-cutting, improving the user-friendliness of human-machine interaction. The vehicle control method provided in this application can achieve vehicle control in lane-cutting scenarios without relying on V2X (Vehicle to Everything), making it feasible for mass production. The parameters involved can be calibrated and learned, possessing continuous evolution capabilities.

[0230] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0231] Based on the same inventive concept, this application also provides a vehicle control device for implementing the vehicle control method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more vehicle control device embodiments provided below can be found in the limitations of the vehicle control method described above, and will not be repeated here.

[0232] In one exemplary embodiment, such as Figure 5 As shown, a vehicle control device 500 is provided, including: a lane-cutting intention information acquisition module 501, a lane-cutting intention type determination module 502, a control strategy determination module 503, and a control strategy execution module 504, wherein:

[0233] The lane-cutting intention information acquisition module 501 is used to acquire lane-cutting intention information of a second vehicle in an adjacent lane adjacent to the lane where the first vehicle is located. The lane-cutting intention information is determined based on vehicle information monitored for the second vehicle. The lane-cutting intention information includes lane-cutting intention intensity, which characterizes the strength of the second vehicle's subjective intention to cut in, and lane-cutting probability, which characterizes the possibility of the second vehicle cutting in.

[0234] The lane-cutting intent type determination module 502 is used to determine the intent intensity range in which the lane-cutting intent intensity is located based on a preset intent intensity threshold; and when it is determined based on vehicle information that the preset type determination condition corresponding to the intent intensity range is met, the lane-cutting intent type corresponding to the intent intensity range is determined as the lane-cutting intent type of the second vehicle.

[0235] The control strategy determination module 503 is used to determine the control strategy for the first vehicle based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs.

[0236] The control strategy execution module 504 is used to control the first vehicle to drive according to the control strategy.

[0237] In some embodiments, the lane-cutting intention type determination module 502 is further configured to determine lane-cutting intention classification parameters based on road information of the road where the first vehicle and the second vehicle are located, and vehicle information; wherein, the lane-cutting intention classification parameters include at least one of headlight status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road status parameters; if the lane-cutting intention classification parameters determine that the preset type determination conditions corresponding to the intention intensity range are met, the lane-cutting intention type corresponding to the intention intensity range is determined as the lane-cutting intention type of the second vehicle; the type determination conditions include logical rules for classifying and verifying the intention intensity range in which the lane-cutting intention intensity is located based on the lane-cutting intention classification parameters.

[0238] In some embodiments, the lane-cutting intent type determination module 502 is further configured to determine the lane-cutting intent type of the second vehicle based on the lane-cutting intent strength and vehicle information when the lane-cutting probability exceeds a preset probability threshold.

[0239] In some embodiments, the control strategy determination module 503 is further configured to determine the intensity adjustment weight based on the type of cutting-in intention, and determine the cutting-in permission coefficient based on the vehicle category to which the second vehicle belongs; the cutting-in permission coefficient is used to characterize the priority of the first vehicle to yield to the second vehicle; the cutting-in game intensity is determined according to the intensity adjustment weight, the cutting-in permission coefficient, the cutting-in intention intensity and the cutting-in probability; and the control strategy for the first vehicle is determined according to the cutting-in game intensity.

[0240] In some embodiments, the control strategy determination module 503 is further configured to adjust the intensity of the intention to cut in line by adjusting the intensity adjustment weight to obtain the adjusted intensity of the intention to cut in line; determine the game weighting weight according to the driving scenarios of the first vehicle and the second vehicle; and fuse the adjusted intensity of the intention to cut in line, the cutting in line permission coefficient and the cutting in line probability according to the game weighting weight to obtain the game intensity of cutting in line.

[0241] In some embodiments, the control strategy determination module 503 is further configured to determine the game intensity range in which the game intensity of cutting in line is located based on a preset game intensity threshold; and to determine the preset control strategy that conforms to safety constraints corresponding to the game intensity range as the control strategy for the first vehicle.

[0242] In some embodiments, the vehicle control device 500 further includes a misoperation handling module, which controls the first vehicle to issue a misoperation warning to the second vehicle when the cut-in intention type is a misoperation cut-in type; and returns to the step of determining the cut-in intention type of the second vehicle based on the cut-in intention intensity and vehicle information when the second vehicle continues to cut in.

[0243] In some embodiments, the control strategy determination module 503 is further configured to determine the control strategy as proactive yielding when the second vehicle belongs to a special vehicle category.

[0244] In some embodiments, the vehicle control device 500 further includes a lane-cutting intention information determination module, which is used to predict lane-cutting intention based on wheel end motion information and lane-cutting trend prediction information included in the vehicle information, and obtain lane-cutting intention information. The lane-cutting trend prediction information is determined based on the historical driving information of the second vehicle.

[0245] Each module in the aforementioned vehicle control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of the controller in hardware form or independent of it, or stored in the memory of the controller in software form, so that the processor can call and execute the corresponding operations of each module.

[0246] In one exemplary embodiment, a controller is provided, the internal structure of which can be shown in the following diagram. Figure 6 As shown, the controller includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores various data involved in the vehicle control method. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a vehicle control method.

[0247] Those skilled in the art will understand that Figure 6 The structure shown is a block diagram of a partial structure related to the solution of this application, and does not constitute a limitation on the controller applied thereto by the solution of this application. The specific controller may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0248] In one exemplary embodiment, a controller is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0249] In one exemplary embodiment, a vehicle is provided, including at least one sensor and the controller described above, wherein the sensor is used to monitor a second vehicle to obtain vehicle information of the second vehicle.

[0250] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0251] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0252] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0253] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program mentioned can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0254] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0255] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A vehicle control method characterized by, The method includes: The cut-in intention information of a second vehicle in an adjacent lane adjacent to the lane occupied by the first vehicle is obtained. The cut-in intention information is determined based on vehicle information monitored for the second vehicle. The cut-in intention information includes cut-in intention intensity, which characterizes the strength of the subjective intention of the second vehicle to cut in, and cut-in probability, which characterizes the possibility of the second vehicle cutting in. Based on a preset intent intensity threshold, the intent intensity range in which the intention to cut in is located is determined; If, based on the vehicle information, it is determined that the preset type determination condition corresponding to the intent intensity range is met, the cut-in intent type corresponding to the intent intensity range is determined as the cut-in intent type of the second vehicle. Based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs, a control strategy is determined for the first vehicle; The first vehicle is controlled to drive according to the control strategy described above.

2. The method of claim 1, wherein, If, based on the vehicle information, a preset type determination condition corresponding to the intent intensity range is met, the cut-in intent type corresponding to the intent intensity range is determined as the cut-in intent type of the second vehicle, including: Based on the road information of the roads where the first vehicle and the second vehicle are located, and the vehicle information, the lane-cutting intention classification parameters are determined; wherein, the lane-cutting intention classification parameters include at least one of the following: vehicle light status parameters, spatial index parameters, kinematic index parameters, driving characteristic parameters, or road status parameters; If the pre-set type determination condition corresponding to the intent intensity range is met based on the lane-cutting intent classification parameters, the lane-cutting intent type corresponding to the intent intensity range is determined as the lane-cutting intent type of the second vehicle; the type determination condition includes logical rules for classifying and verifying the intent intensity range in which the lane-cutting intent intensity is located based on the lane-cutting intent classification parameters.

3. The method of claim 1, wherein, Based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs, a control strategy is determined for the first vehicle, including: The intensity adjustment weight is determined based on the type of cutting-in intention, and the cutting-in permission coefficient is determined based on the vehicle category to which the second vehicle belongs; the cutting-in permission coefficient is used to characterize the priority of the first vehicle to yield to the second vehicle; The intensity of the game of vying for space is determined based on the intensity adjustment weight, the permission coefficient for vying for space, the intensity of the intention to vying for space, and the probability of vying for space. The control strategy for the first vehicle is determined based on the strength of the lane-jumping game.

4. The method of claim 3, wherein, The intensity of the game of cutting in line is determined based on the intensity adjustment weight, the cutting-in permission coefficient, the intensity of the cutting-in intention, and the cutting-in probability, including: The intensity of the intention to cut in line is adjusted by the intensity adjustment weight to obtain the adjusted intensity of the intention to cut in line; The game weighting is determined based on the driving scenarios of the first vehicle and the second vehicle. According to the game weighting, the adjusted intention to cut in line intensity, the cutting in line authority coefficient, and the cutting in line probability are fused to obtain the cutting in line game intensity.

5. The method of claim 3, wherein, Determining the control strategy for the first vehicle based on the strength of the lane-jumping game includes: Based on a preset game strength threshold, the game strength range in which the blocking game strength falls is determined; The control strategy corresponding to the game intensity range and conforming to the preset safety constraints is determined as the control strategy for the first vehicle.

6. The method according to any one of claims 1 to 5, characterized in that, The method further includes at least one of the following: If the cut-in intention type is a misoperation cut-in type, the first vehicle is controlled to issue a cut-in misoperation reminder to the second vehicle; if the second vehicle continues to cut in, the process returns to the step of determining the cut-in intention type of the second vehicle based on the cut-in intention intensity and the vehicle information. If the second vehicle belongs to a special vehicle category, the control strategy is determined to be proactive yielding; Based on the wheel end motion information and cutting-in trend prediction information included in the vehicle information, cutting-in intention is predicted to obtain the cutting-in intention information. The cutting-in trend prediction information is determined based on the historical driving information of the second vehicle.

7. A vehicle control device characterized by comprising: The device includes: The lane-cutting intention information acquisition module is used to acquire lane-cutting intention information of a second vehicle in an adjacent lane adjacent to the lane where the first vehicle is located. The lane-cutting intention information is determined based on vehicle information monitored for the second vehicle. The lane-cutting intention information includes lane-cutting intention intensity, which characterizes the strength of the second vehicle's subjective intention to cut in, and lane-cutting probability, which characterizes the possibility of the second vehicle cutting in. The lane-cutting intent type determination module is used to determine the intent intensity range in which the lane-cutting intent intensity is located based on a preset intent intensity threshold; and when it is determined based on the vehicle information that the preset type determination condition corresponding to the intent intensity range is met, the lane-cutting intent type corresponding to the intent intensity range is determined as the lane-cutting intent type of the second vehicle. The control strategy determination module is used to determine a control strategy for the first vehicle based on the type of cutting-in intention, the intensity of the cutting-in intention, the probability of cutting-in, and the vehicle category to which the second vehicle belongs. The control strategy execution module is used to control the first vehicle to drive according to the control strategy.

8. A controller comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A vehicle characterized by comprising: It includes at least one sensor and a controller as described in claim 8, the sensor being used to monitor a second vehicle to obtain vehicle information of the second vehicle.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images