A motion state determination method and apparatus, and a storage medium

Abstract

The application discloses a motion state determination method and device and a storage medium, and the method comprises the following steps: performing real-time motion trajectory prediction on the traffic subjects around the ego vehicle to obtain the predicted trajectory data of the traffic subjects at the current time; determining the lane-changing direction and target lane of the traffic subjects according to the predicted trajectory data, and determining the target lane line type and target lane attribute of the target lane; performing weighted score calculation on the predicted trajectory data, target lane line type and target lane attribute according to the type of the traffic subjects to obtain the lane-changing intention score of the traffic subjects at the current time; and determining the motion state of the traffic subjects according to the lane-changing intention score and lane-changing direction. In the application, the influence of the predicted trajectory, type and traffic rules of the traffic subjects on the lane-changing intention is considered, the accuracy of the lane-changing intention of the traffic subjects is improved, and therefore the motion state of the traffic subjects can be accurately determined.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and in particular to a method, apparatus and storage medium for determining motion state. Background Technology

[0002] In complex urban roads, lane changes occur frequently. During autonomous driving, the vehicle's autonomous driving system needs to constantly monitor the traffic conditions around the vehicle, identify the lane-changing intentions of surrounding traffic subjects (including pedestrians and vehicles), determine the driving status of traffic subjects, and provide a basis for making reasonable driving decisions, thereby ensuring driving safety and efficient traffic flow.

[0003] Existing driving status recognition technologies typically acquire vehicle driving status parameters and predict other vehicles' lane-changing intentions based on these parameters. However, this method, which uses driving status parameters to predict lane-changing intentions, considers fewer influencing factors and has significant limitations, resulting in low accuracy in predicting driving status. Summary of the Invention

[0004] This invention provides a method, apparatus, and storage medium for determining motion state, in order to solve the problem that existing driving state recognition technologies consider too few influencing factors, resulting in low accuracy in predicting driving state recognition.

[0005] A method for determining motion state is provided, including:

[0006] Real-time prediction of the movement trajectories of traffic entities around the vehicle to obtain the predicted trajectory data of the traffic entities at the current moment;

[0007] Based on the predicted trajectory data, determine the lane-changing direction and target lane of the main traffic vehicle, and determine the target lane line type and target lane attributes.

[0008] Based on the type of traffic subject, a weighted score is calculated on the predicted trajectory data, the target lane line type, and the target lane attributes to obtain the lane-changing intention score of the traffic subject at the current moment.

[0009] The movement state of the traffic vehicle is determined based on the score of the lane-changing intention and the direction of the lane change.

[0010] Furthermore, based on the type of traffic subject, a weighted score is calculated for the predicted trajectory data, target lane line type, and target lane attributes to obtain the traffic subject's lane-changing intention score, including:

[0011] Based on the predicted trajectory data of the traffic vehicle, calculate the first intention score of the traffic vehicle to change lanes to the target lane;

[0012] The second intent score is determined based on the target lane line type, and the third intent score is determined based on the target lane attributes.

[0013] The corresponding weights for the first intention score, second intention score, and third intention score are determined based on the type of traffic subject.

[0014] Based on their respective weights, the scores for the first intention, the second intention, and the third intention are weighted and summed to obtain the lane-changing intention score.

[0015] Furthermore, based on the predicted trajectory data of the traffic vehicle, the first intention score of the traffic vehicle to change lanes to the target lane is calculated, including:

[0016] Based on the predicted trajectory data, the target time when the traffic subject moves to the target lane line is determined, and the target duration from the current time to the target time is determined. The target lane line is the lane line between the lane where the traffic subject is located and the target lane.

[0017] Determine the center line and width of the lane where the main traffic vehicle is located;

[0018] In the predicted trajectory data, determine the maximum deviation distance between the predicted trajectory of the traffic vehicle and the road centerline of the lane in which the traffic vehicle is located;

[0019] The first intent score is calculated based on the target duration, maximum deviation distance, and lane width of the lane where the main traffic vehicle is located.

[0020] Furthermore, based on the lane-changing intention score and lane-changing direction, the movement state of the traffic vehicle is determined, including:

[0021] The lane-changing intention of the traffic subject is determined based on the lane-changing intention score and lane-changing direction at consecutive preset time points.

[0022] Determine the relative position of the traffic vehicle's lane to the lane of the vehicle itself;

[0023] The movement state of the main traffic vehicle is determined based on the relative positional relationship and the intention to change lanes.

[0024] Furthermore, based on the lane-changing intention scores and lane-changing directions of the traffic vehicle at consecutive preset time points, the lane-changing intention of the traffic vehicle is determined, including:

[0025] Determine whether the lane-changing intention score of the traffic subject at each of the consecutive preset time points is greater than the preset score, and determine whether the lane-changing direction of the traffic subject at each of the consecutive preset time points is the same;

[0026] If the lane-changing intention score of a traffic subject is not greater than the preset score for a consecutive preset number of time periods, or if the lane-changing direction of the traffic subject is different for a consecutive preset number of time periods, then the lane-changing intention is determined to be lane keeping.

[0027] If the lane-changing intention score of a traffic entity is greater than the preset score for a consecutive preset number of time periods, and the lane-changing direction of the traffic entity is the same for a consecutive preset number of time periods, then the lane-changing intention is determined to be a lane-changing in the direction of lane changing.

[0028] Furthermore, based on the relative positional relationship and lane-changing intention, the movement state of the traffic vehicle is determined, including:

[0029] When the relative position relationship is the current lane of the vehicle, if the lane changing intentions are lane keeping, lane changing to the left and lane changing to the right, the movement states of the traffic subject are lane keeping and going straight, lane changing to the left and lane changing to the right, respectively.

[0030] When the relative position relationship is the left lane of the vehicle, if the lane changing intentions are lane keeping, changing lanes to the left and changing lanes to the right, the movement states of the traffic subject are respectively maintaining the straight state in the left lane, moving out of the left lane, and entering the current lane from the left.

[0031] When the relative position relationship is the right lane of the vehicle, if the lane changing intentions are lane keeping, changing lanes to the left, and changing lanes to the right, the movement states of the traffic vehicle are respectively maintaining the straight state of the right lane, entering the current lane from the right, and leaving the right lane.

[0032] Furthermore, the vehicle's surrounding traffic entities are predicted in real time to obtain predicted trajectory data of the traffic entities at the current moment, including:

[0033] The vehicle's environmental perception system collects driving perception data in real time, including the vehicle's status data, the status information of traffic entities around the vehicle, and the road information on which the vehicle is traveling.

[0034] The status information of the vehicle and the traffic entity is converted into the world coordinate system and filtered to obtain filtered status data.

[0035] The center line of the lane where the vehicle is located is determined based on the driving road information. The center line of the lane is used as a reference line to perform Frenet coordinate system transformation on the filtered state data to obtain preprocessed data.

[0036] The preprocessed data is input into the trajectory prediction model so that the trajectory prediction model can predict the movement trajectory of the traffic subject and obtain the predicted trajectory data of the traffic subject at the current time.

[0037] A motion state determination device is provided, comprising:

[0038] The prediction module is used to predict the movement trajectory of traffic entities around the vehicle in real time, so as to obtain the predicted trajectory data of the traffic entities at the current moment.

[0039] The first determining module is used to determine the lane changing direction and target lane of the traffic body based on the predicted trajectory data, and to determine the target lane line type and target lane attributes of the target lane.

[0040] The calculation module is used to perform weighted scoring calculations on the predicted trajectory data, target lane line type, and target lane attributes based on the type of traffic subject, in order to obtain the lane-changing intention score of the traffic subject at the current moment.

[0041] The second determination module is used to determine the movement state of the traffic vehicle based on the lane-changing intention score and lane-changing direction.

[0042] A motion state determination device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the motion state determination method described above.

[0043] A readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the motion state determination method described above.

[0044] In one scheme provided by the aforementioned motion state determination method, device, and storage medium, the motion trajectory of traffic entities surrounding the vehicle is predicted in real time to obtain the predicted trajectory data of the traffic entities at the current moment. Then, based on the predicted trajectory data, the lane-changing direction and target lane of the traffic entity are determined, as well as the target lane line type and target lane attribute. Next, based on the type of traffic entity, a weighted score is calculated on the predicted trajectory data, target lane line type, and target lane attribute to obtain the lane-changing intention score of the traffic entity at the current moment. Finally, the motion state of the traffic entity is determined based on the lane-changing intention score and lane-changing direction. In this invention, the motion trajectory of the traffic entity is predicted in real time to determine the lane-changing target and direction. Then, a weighted score is calculated based on the target lane, its target lane line type, and target lane attribute to obtain the lane-changing intention score. Finally, the driving state is determined by combining the lane-changing intention score and lane-changing direction. This method considers the influence of factors such as the predicted trajectory, type, and traffic rules of the traffic entity on the lane-changing intention, improving the accuracy of the traffic entity's lane-changing intention and thus enabling accurate judgment of the traffic entity's motion state. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of a motion state determination system according to an embodiment of the present invention;

[0047] Figure 2 This is a flowchart illustrating a motion state determination method according to an embodiment of the present invention;

[0048] Figure 3 yes Figure 2 A schematic diagram of the implementation process of step S30;

[0049] Figure 4 This is a schematic diagram of a motion state determination device according to an embodiment of the present invention;

[0050] Figure 5 This is another structural schematic diagram of the motion state determination device in one embodiment of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] The motion state determination method provided in this embodiment of the invention can be applied to, for example... Figure 1 The motion state determination system shown includes an environmental perception system and a motion state determination device installed on the vehicle. The motion state determination device communicates with the environmental perception system via a bus. Through the vehicle's secure environmental perception system, the motion state determination device collects relevant information about the vehicle and other traffic entities, enabling subsequent prediction of the traffic entities' trajectories based on this information.

[0053] In this embodiment, the movement trajectories of traffic entities surrounding the vehicle are predicted in real time to obtain the predicted trajectory data of the traffic entities at the current moment. Then, based on the predicted trajectory data, the lane-changing direction and target lane of the traffic entity are determined, as well as the target lane line type and target lane attributes. Next, a weighted score is calculated based on the type of traffic entity to obtain the lane-changing intention score of the traffic entity at the current moment. Finally, the movement state of the traffic entity is determined based on the lane-changing intention score and the lane-changing direction. By predicting the movement trajectory of traffic entities in real time to determine the lane-changing target and direction, and then calculating a weighted score based on the target lane, its target lane line type, and target lane attributes to obtain the lane-changing intention score, the driving state is determined by combining the lane-changing intention score and the lane-changing direction. This approach considers the influence of factors such as the movement trajectory, type, and traffic rules of the traffic entity on the lane-changing intention, improving the accuracy of the traffic entity's lane-changing intention. This allows for accurate judgment of the movement state of the traffic entity, thus providing an important decision-making basis for the autonomous driving of the vehicle.

[0054] In this embodiment, the motion state determination system includes an environmental perception system and a motion state determination device installed on the vehicle. This is only an illustrative example. In other embodiments, the motion state determination system may also include other devices, which will not be described in detail here.

[0055] In one embodiment, such as Figure 2 As shown, a method for determining motion state is provided, which is then applied to... Figure 1 Taking the motion state determination device in the example, the following steps are included:

[0056] S10: Real-time prediction of the movement trajectories of traffic entities around the vehicle to obtain the predicted trajectory data of the traffic entities at the current moment.

[0057] It's important to understand that open roads involve a large number of traffic participants, including autonomously mobile traffic entities and static traffic facilities. Traffic entities (intelligent traffic entities) refer to people with intelligence and autonomous mobility, or vehicles driven by people. Traffic entities possess autonomous mobility and, while in motion, travel according to traffic rules and under defined conditions. During driving, drivers must not only pay attention to the current vehicle situation around their vehicle but also anticipate the situation of pedestrians around their vehicle. They need to determine whether the movement of surrounding traffic entities will affect the normal driving of their vehicle, thus assisting in taking relevant driving actions in advance.

[0058] Therefore, during the vehicle's operation, the motion state determination device collects real-time driving perception data from the vehicle's environmental perception system. This driving perception data includes the vehicle's state data, the state information of surrounding traffic entities, and the vehicle's driving road information. The vehicle's state information includes, but is not limited to, its position, attitude, speed, and acceleration; the traffic entity's state information includes, but is not limited to, its tracking ID, type, position, attitude, speed, and acceleration; and the driving road information includes, but is not limited to, the road type within a certain distance before and after the vehicle's current road segment, as well as information such as lane centerlines, lane lines, road edge lines, target lane line type, and target lane attributes for all lanes.

[0059] After acquiring the vehicle's state data, the state information of other traffic entities, and the road information, the vehicle's and other traffic entities' state information are transformed into a world coordinate system to obtain unified coordinate comprehensive data for subsequent data fusion processing. Then, based on the road information, the lane centerline of the vehicle's lane is determined. Using the lane centerline as a reference line and the vehicle's coordinate position as the origin of the coordinate system, the filtered state data undergoes a Frenet coordinate system transformation to obtain preprocessed data of the traffic entities relative to the vehicle. This preprocessed data is then input into a trajectory prediction model to predict the trajectory of the traffic entities within a preset time period (e.g., 10 seconds), resulting in predicted trajectory data of the traffic entities relative to the vehicle's position at the current moment. This prediction rule data includes the predicted trajectory of the traffic entities within the preset future time period, as well as lane information around the trajectory and information on the position, attitude, and acceleration of other traffic entities.

[0060] In this embodiment, the vehicle's state data, the state information of the traffic entities around the vehicle, and the vehicle's driving road information are converted into preprocessed data. Then, the trajectory of the traffic entities is predicted according to the trajectory prediction model. This not only considers the state information of the vehicle and the traffic entities, but also the road information, which improves the accuracy of the prediction results.

[0061] S20: Based on the predicted trajectory data, determine the lane-changing direction and target lane of the main traffic vehicle, and determine the target lane line type and target lane attributes.

[0062] After obtaining the predicted trajectory data of the traffic vehicle at the current moment, the lane-changing direction of the current traffic vehicle is predicted based on the predicted trajectory data. The direction in which the predicted trajectory extends is taken as the lane-changing direction. Then, the nearest lane in the lane-changing direction is taken as the target lane, thereby determining the target lane line type and target lane attributes of the target vehicle. The target lane line type refers to the type of lane line on the side of the lane-changing direction, that is, the type of lane line between the target vehicle and the lane where the traffic vehicle is located. The target lane attributes refer to the road surface attributes of the target lane, that is, the road surface attributes of the lane on the side of the lane-changing direction.

[0063] The types of lane markings include at least no-lane-changing (such as solid lines) and allowed-lane-changing (dashed lines), and the road surface attributes of lanes include lanes in the same direction, lanes in opposite directions, emergency lanes, lanes that are not allowed to be driven, and lanes for non-motorized vehicles.

[0064] S30: Based on the type of traffic subject, a weighted score is calculated on the predicted trajectory data, the target lane line type, and the target lane attributes to obtain the lane-changing intention score of the traffic subject at the current moment.

[0065] To determine the lane-changing direction and target lane for a traffic vehicle, and to identify the target lane's lane type and attributes, it's necessary to score and weight the predicted trajectory data, target lane type, and target lane attributes based on the traffic vehicle's type. This weighted sum is then used to obtain the traffic vehicle's lane-changing intention score at the current moment, which is used to determine the lane-changing intention. Specifically, the probability of a traffic vehicle changing lanes to the target lane is scored based on the predicted trajectory data, and the target lane type and attributes are also scored separately. These scores are then weighted and summed to obtain the traffic vehicle's lane-changing intention score.

[0066] In this embodiment, when determining lane-changing intentions, the influence of traffic rules on lane-changing intentions is considered by using data such as the type of traffic subject, the type of target lane line, and the attributes of the target lane, thereby improving the accuracy of lane-changing intention determination.

[0067] S40: Determine the motion state of the traffic vehicle based on the lane-changing intention score and lane-changing direction.

[0068] After determining the lane-changing intention score of a traffic subject at the current moment, the movement state of the traffic subject can be determined based on the lane-changing intention score and the lane-changing direction. For example, if the lane-changing intention score of a traffic subject at the current moment is greater than a certain score threshold, it indicates that the traffic subject has the intention to change lanes. Then, the driving state of the traffic subject is determined based on the lane-changing direction, that is, whether the traffic subject is moving towards the vehicle, moving away from the vehicle, or maintaining its original driving state. The determination result is then sent to the vehicle's autonomous driving decision-making device to improve the accuracy of the vehicle's safe driving.

[0069] In this embodiment, the movement trajectories of traffic entities surrounding the vehicle are predicted in real time to obtain the predicted trajectory data of the traffic entities at the current moment. Then, based on the predicted trajectory data, the lane-changing direction and target lane of the traffic entity are determined, as well as the target lane line type and target lane attributes. Then, based on the type of traffic entity, a weighted score is calculated on the predicted trajectory data, target lane line type, and target lane attributes to obtain the lane-changing intention score of the traffic entity at the current moment. Finally, the movement state of the traffic entity is determined based on the lane-changing intention score and lane-changing direction. By predicting the movement trajectory of traffic entities in real time to determine the lane-changing target and direction, and then calculating a weighted score based on the target lane, its target lane line type, and target lane attributes to obtain the lane-changing intention score, the driving state is determined by combining the lane-changing intention score and the lane-changing direction. This approach considers the influence of factors such as the prediction and type of traffic entities and traffic rules on the lane-changing intention, improving the accuracy of the traffic entity's lane-changing intention. This allows for accurate judgment of the movement state of traffic entities, thus providing an important decision-making basis for the autonomous driving of the vehicle.

[0070] In one embodiment, step S10, which involves real-time prediction of the motion trajectories of traffic entities surrounding the vehicle to obtain the predicted trajectory data of the traffic entities at the current moment, specifically includes the following steps:

[0071] S11: Through the vehicle's environmental perception system, the vehicle's driving perception data is collected in real time. The driving perception data includes the vehicle's status data, the status information of the traffic entities around the vehicle, and the vehicle's driving road information.

[0072] During vehicle operation, the motion state determination device collects real-time driving perception data from the vehicle's environmental perception system. This driving perception data includes the vehicle's state data, the state information of surrounding traffic entities, and the vehicle's driving road information. The vehicle's state information includes, but is not limited to, its position, attitude, speed, and acceleration; the traffic entity's state information includes, but is not limited to, its tracking ID, type, position, attitude, speed, and acceleration; and the driving road information includes, but is not limited to, the road type within a certain distance before and after the vehicle's current road segment, as well as information such as lane center lines, lane markings, road edge lines, target lane line type, and target lane attributes for all lanes.

[0073] The driving road information can be map navigation information obtained by the map navigation module to improve information acquisition speed, reduce information processing volume, and thus reduce system load. In other embodiments, to improve the real-time performance and accuracy of the driving road information, the driving road information can also be road information collected by the vehicle in real time.

[0074] S112 converts the status information of the vehicle and the traffic entity into the world coordinate system and performs filtering to obtain filtered status data.

[0075] After acquiring the vehicle's status data, the traffic entity's status information, and the driving road information, the vehicle's status information, the traffic entity's status information, and the driving road information are transformed into the world coordinate system, so that the three are in the same dimension, in order to obtain comprehensive data after unified coordinates, so as to carry out subsequent data fusion processing.

[0076] Then, the integrated data pairs after unifying coordinates are filtered to obtain filtered state data. The filtering process can employ the Kalman filter algorithm or its variations; other filtering algorithms may also be used in other embodiments, which will not be elaborated here. In this embodiment, the extended Kalman filter algorithm is used to filter the coordinate positions, attitudes, speeds, accelerations, and other data of the vehicle and traffic entities in the integrated data. This reduces data noise caused by vehicle vibration, perception errors, etc., to obtain the optimal estimate of the motion state of the vehicle and traffic entities, making the data smoother and more reliable.

[0077] The world coordinate system can be the Universal Transverse Mercator (UTM) coordinate system to determine the accuracy of data transformation.

[0078] S13: Determine the center line of the lane where the vehicle is located based on the driving road information. Using the center line as a reference line, perform Frenet coordinate system transformation on the filtered state data to obtain preprocessed data.

[0079] After obtaining the filtered data, the lane centerline of the vehicle's lane is determined based on the road information. Using the lane centerline as a reference line and the vehicle's coordinate position as the origin of the coordinate system, the filtered state data is transformed into the Frenet coordinate system. The coordinate position, attitude, velocity, and acceleration of the vehicle and other traffic objects in the Frenet coordinate system, as well as the distance between the traffic objects and the lane centerline of their respective lanes, are calculated to obtain preprocessed data of the traffic objects relative to the vehicle. Transforming the filtered state data into the Frenet coordinate system allows the lane centerline of the vehicle's lane to be used as an anchor reference line for predicting the trajectories of surrounding traffic objects, adding road constraints to trajectory prediction and improving its accuracy.

[0080] After obtaining the preprocessed data of the traffic subject relative to the vehicle at the current moment, the preprocessed data of the current moment is saved in the historical trajectory dataset. In this historical trajectory dataset, the preprocessed data of the same traffic subject is saved for a maximum of a preset time period (e.g., 10 seconds). After the preset time period, the data is saved in a rolling manner, that is, the earliest frame of preprocessed data of the traffic subject is deleted and the latest preprocessed data is added to the last frame, so as to reduce the amount of data storage and thus reduce the load.

[0081] S14: Input the preprocessed data into the trajectory prediction model so that the trajectory prediction model can predict the movement trajectory of the traffic subject and obtain the predicted trajectory data of the traffic subject at the current time.

[0082] After obtaining the preprocessed data of the traffic subject relative to the vehicle at the current moment, the preprocessed data is input into the trajectory prediction model. This allows the trajectory prediction model to predict the movement trajectory of the traffic subject within a preset time period (e.g., 10 seconds), resulting in predicted trajectory data of the traffic subject's position relative to the vehicle at the current moment. This prediction data includes the predicted trajectory of the traffic subject within the preset future time period, as well as information such as lane information around the trajectory, the position, attitude, acceleration of other traffic subjects, and the distance of the traffic subject from the centerline of its lane.

[0083] The trajectory prediction model is a pre-trained deep learning network model. The deep learning network for the trajectory prediction model can employ Long Short-Term Memory (LSTM) networks or Social Graph Networks, among others. These networks can be applied to motion trajectory prediction models and achieve high prediction accuracy. In this embodiment, a social graph network model is used to train the trajectory prediction model. Based on the filtered pre-processed data, a directed social graph is constructed with vehicles and traffic entities as nodes. This graph is input into the social graph network, and a temporal stochastic method is used to sequentially learn the uncertainties during social interactions, thereby predicting the motion trajectories of all nodes within a preset time period, thus improving the accuracy of the predicted trajectory. In this embodiment, the preset time period is 10 seconds, meaning the trajectory prediction model predicts the motion trajectory of the traffic entity over the next 10 seconds to obtain the predicted trajectory data of the traffic entity at the current moment.

[0084] In this embodiment, the vehicle's environmental perception system collects real-time driving perception data, including the vehicle's state data, the state information of surrounding traffic entities, and the vehicle's driving road information. The vehicle's state information and the traffic entity's state information are converted to a world coordinate system and filtered to obtain filtered state data. The lane centerline of the vehicle's lane is determined based on the driving road information. Using the lane centerline as a reference line, the filtered state data is transformed to a Frenet coordinate system to obtain preprocessed data. This preprocessed data is then input into a trajectory prediction model to predict the movement trajectory of traffic entities, resulting in the predicted trajectory data of the traffic entities at the current moment. This clarifies the specific process of real-time trajectory prediction of traffic entities surrounding the vehicle to obtain the predicted trajectory data of the traffic entities at the current moment. The conversion and filtering of the vehicle's state data, the state information of surrounding traffic entities, and the vehicle's driving road information to obtain preprocessed data improves data accuracy. Furthermore, the trajectory prediction model predicts the movement trajectory of traffic entities, considering not only the state information of the vehicle and traffic entities but also road information, further improving the accuracy of the prediction results.

[0085] In one embodiment, before converting the vehicle's state information and the state information of traffic entities to the world coordinate system, it is necessary to select several traffic entities closest to the vehicle in each lane based on the lane where the traffic entity is located and its distance from the vehicle. Other traffic entities with minimal impact on the vehicle are then eliminated to obtain target traffic entities. This allows for subsequent processing of the target traffic entity's state data and the vehicle's state data, thereby reducing data processing, system complexity, and improving operational efficiency. Specifically, the vehicle's position is used as a transverse segment. In each lane in front of the vehicle, n (e.g., 3) traffic entities closest to the vehicle are selected. If there are fewer than n traffic entities in a lane, all traffic entities in that lane are selected. In each lane behind the vehicle, m traffic entities closest to the vehicle are selected. Since the driving state behind the vehicle has minimal impact on the vehicle, m is less than n. If there are no traffic entities in a lane, they are ignored. Simultaneously, all traffic entities on different planes from the road where the vehicle is located, or those with insurmountable areas (such as guardrails or medians) between them, are eliminated.

[0086] In this embodiment, before converting the state information of the vehicle and the state information of the traffic subject to the world coordinate system, the traffic subject is screened so that the state data of the target traffic subject and the state data of the vehicle can be processed in the future, thereby reducing data processing, reducing system complexity, and improving operating efficiency.

[0087] In one embodiment, such as Figure 3 As shown, in step S30, a weighted score is calculated based on the type of traffic subject, considering the predicted trajectory data, the target lane line type, and the target lane attributes, to obtain the lane-changing intention score of the traffic subject. This specifically includes the following steps:

[0088] S31: Calculate the first intention score of the traffic vehicle to change lanes to the target lane based on the predicted trajectory data of the traffic vehicle.

[0089] After obtaining the predicted trajectory data, the probability of the traffic subject changing lanes to the target lane is calculated based on the predicted trajectory data of the traffic subject at the current time, thereby obtaining the first intention score.

[0090] Among them, the closer the predicted trajectory of a traffic subject is to the lane line on one side of the traffic subject, that is, the shorter the distance between the predicted trajectory and the lane line on one side of the traffic subject, the greater the probability that the traffic subject will change lanes to the target lane, and the higher the first intention score.

[0091] S32: Determine the second intent score based on the target lane line type, and determine the third intent score based on the target lane attributes.

[0092] After determining the target lane line type and target lane attributes, a second intention score is determined based on the target lane line type, and a third intention score is determined based on the target lane attributes.

[0093] The rules for calculating the second intent score include, but are not limited to, the following:

[0094] (1) When the lane line type (target lane line type) in the direction of lane change is a solid line or other lane line type that prohibits lane change, the second intention score is the first preset value. The first preset value is negative, which indicates the penalty for illegal lane change intention. For example, when the target lane line type is a solid line, the second intention score can be -10.

[0095] (2) When the lane line type in the lane changing direction is a dashed line or other lane line type that allows lane changing, the second intention score is the second preset value. The second preset value is a non-negative value, for example, the second preset value can be 0.

[0096] The rules for calculating the third intent score include, but are not limited to, the following:

[0097] 1. When the road surface attribute (target lane attribute of the target lane) on the side of the lane change direction is the same direction lane, the third intention score is the third preset value. The third preset value is a non-negative value, for example, the third preset value can be 0.

[0098] 2. When the road surface attribute on the side of the lane change direction is the oncoming lane, the third intention score is the fourth preset value. The fourth preset value is negative, which indicates a penalty for dangerous lane change intentions. For example, if the target lane attribute is the oncoming lane, the third intention score can be -10.

[0099] 3. When the road surface attribute on the side of the lane change direction is the emergency lane, the third intention score is the fifth preset value. The fifth preset value is negative, which indicates the penalty for illegal lane change intention. For example, if the target lane attribute is the emergency lane, the third intention score can be -10.

[0100] 4. When the road surface attribute on the side of the lane change direction is a non-motorized vehicle lane and the traffic subject is a motor vehicle, the third intention score is the sixth preset value. The sixth preset value is negative, which indicates the penalty for illegal lane change intention. The sixth preset value is greater than the first preset value, the fourth preset value and the fifth preset value. For example, when the target lane attribute is a non-motorized vehicle lane and the traffic subject is a motor vehicle, the third intention score can be -5.

[0101] 5. When the road surface on the side of the lane change direction is a non-drivable area, the third intention score is the seventh preset value. The seventh preset value is negative, which indicates a penalty for incorrect lane change intention. The seventh preset value can be the minimum value among the preset values ​​mentioned above, and the third intention score can be -20.

[0102] S33: Determine the corresponding weights of the first intention score, second intention score, and third intention score based on the type of traffic subject.

[0103] After determining the scores for the first, second, and third intentions, it is necessary to determine the corresponding weights for these scores based on the type of traffic subject. Traffic rules impose different levels of restrictions on different types of traffic subjects; therefore, different types of traffic subjects require different combinations of weights.

[0104] The weighting rules for the first intention score, second intention score, and third intention score include, but are not limited to, the following:

[0105] 1) When the type of traffic subject is a car, the weights of the first intention score, the second intention score and the third intention score are equal, that is, the weight combination ratio of the first intention score, the second intention score and the third intention score is 1:1:1.

[0106] 2) When the type of traffic subject is a non-motorized vehicle, electric bicycle, motorcycle, etc., since the traffic rules have a weaker binding force on these types of traffic subjects, the weights corresponding to the second intention score and the third intention score are set to lower weights. For example, the weight combination ratio of the first intention score, the second intention score and the third intention score is 1:0.5:0.5.

[0107] 3) When the type of traffic subject is a pedestrian, since the traffic rules have a very weak binding force on pedestrians, the weights corresponding to the second intention score and the third intention score are set to very low weights, much lower than the weight of the first intention score. For example, the weight combination ratio of the first intention score, the second intention score and the third intention score is 1:0.5:0.5.

[0108] S34: Based on the corresponding weights, the scores of the first intention, the second intention, and the third intention are weighted and summed to obtain the lane-changing intention score.

[0109] After determining the first intention score, the second intention score, and the third intention score, the first intention score, the second intention score, and the third intention score are weighted and summed according to their respective weights to obtain the lane-changing intention score.

[0110] The lane-changing intention score can be calculated using the following formula:

[0111] S t =w1*S1+w2*S2+w3*S3;

[0112] Where S is the lane-changing intention score of the traffic subject at time t; S1 is the first intention score, S2 is the second intention score, and S3 is the third intention score.

[0113] In this embodiment, based on the predicted trajectory data of the traffic subject, a first intention score for changing lanes to the target lane is calculated, a second intention score is determined based on the target lane line type, and a third intention score is determined based on the target lane attributes. Then, the corresponding weights of the first, second, and third intention scores are determined according to the type of the traffic subject. Subsequently, the first, second, and third intention scores are weighted and summed according to their corresponding weights to obtain the lane-changing intention score. This clarifies the specific process of calculating a weighted score based on the predicted trajectory data, target lane line type, and target lane attributes to obtain the lane-changing intention score of the traffic subject. By determining and scoring the probability of lane changing based on the predicted trajectory, and scoring based on the lane line type and lane attributes of the lane change, the lower the probability of a lane change that is not permitted by traffic rules, the lower the score. This weighted calculation ensures the accuracy of the intention score and provides a basis for subsequently determining the driving status of the traffic subject based on the lane-changing intention score and the lane-changing direction.

[0114] In one embodiment, step S31, which involves calculating the first intention score of the traffic vehicle to change lanes to the target lane based on the predicted trajectory data of the traffic vehicle, specifically includes the following steps:

[0115] S311: Based on the predicted trajectory data, predict the target time when the main traffic vehicle moves to the target lane line, and determine the target duration from the current time to the target time.

[0116] After obtaining the predicted trajectory data, based on the predicted trajectory extension of the traffic vehicle in the predicted trajectory data, the target time when the traffic vehicle moves to the target lane line is predicted, and the target duration from the current time to the target time is determined. The target lane line is the lane line between the lane where the traffic vehicle is located and the target lane.

[0117] S312: Determine the center line and width of the lane where the main traffic vehicle is located.

[0118] At the same time, based on the lane information in the predicted trajectory data, the lane centerline and lane width of the lane where the main traffic vehicle is located are determined.

[0119] S313: In the predicted trajectory data, determine the maximum deviation distance between the predicted trajectory of the traffic vehicle and the road centerline of the lane in which the traffic vehicle is located.

[0120] At the same time, the maximum deviation distance between the predicted trajectory of the traffic vehicle and the road centerline of the lane in which the traffic vehicle is located is determined in the predicted trajectory data.

[0121] S314: Calculate the first intention score based on the maximum deviation distance, the lane centerline of the lane where the main traffic vehicle is located, and the lane width.

[0122] Finally, the first intent score is calculated based on the target duration, maximum deviation distance, and lane width of the lane where the main traffic vehicle is located.

[0123] The first intent score can be calculated as follows:

[0124] S1 = (10 - TC) + (DM - LW);

[0125] Where S1 is the first intention score, TC is the target duration, DM is the maximum deviation distance, and LW is the lane width of the lane where the traffic subject is located.

[0126] In this embodiment, based on the predicted trajectory data, the target time for the traffic subject to move to the target lane line is predicted, and the target duration from the current time to the target time is determined. The target lane line is the lane line between the lane where the traffic subject is located and the target lane. The lane centerline and lane width of the lane where the traffic subject is located are determined. In the predicted trajectory data, the maximum deviation distance between the predicted trajectory of the traffic subject and the road centerline of the lane where the traffic subject is located is determined. Based on the target duration, the maximum deviation distance, and the lane width of the lane where the traffic subject is located, the first intention score is calculated. This clarifies the specific process of calculating the first intention score of the traffic subject changing lanes to the target lane based on the predicted trajectory data of the traffic subject. The first intention score is calculated based on the duration between the current time and the target time for the traffic subject to move to the target lane line, the maximum deviation distance between the predicted trajectory and the lane centerline, and the lane width, making the first intention score more accurate and the prediction of the possibility of the traffic subject changing lanes more accurate.

[0127] In one embodiment, step S40, which determines the motion state of the traffic vehicle based on the lane-changing intention score and lane-changing direction, specifically includes the following steps:

[0128] S41: Determine the lane-changing intention of the traffic subject based on the lane-changing intention score and lane-changing direction at consecutive preset time points.

[0129] After acquiring the predicted trajectory data of the traffic subject at the current moment and calculating the predicted lane-change intention score based on this data, the predicted lane-change intention score and lane-change direction at the current moment are stored as lane-change intention data in the historical lane-change intention dataset. In the historical lane-change intention dataset, a maximum of a preset number of frames (one frame per moment) of lane-change intention data are stored for the same traffic subject. When the preset number of frames is exceeded, the lane-change intention data for that traffic subject is rolled over; that is, the earliest frame of lane-change intention data for that traffic subject in the historical lane-change intention dataset is deleted, and the most recent frame (the latest moment) is added to the corresponding lane-change intention data for that traffic subject as the last frame. To reduce system load and data storage, the preset number of frames can be 3, meaning the historical lane-change intention dataset stores the lane-change intention score and lane-change direction of the most recent 3 consecutive frames for each traffic subject.

[0130] Then, in the historical lane change intention dataset, the lane change intention data of consecutive preset frames (consecutive preset time periods) are determined, that is, the lane change intention score and lane change direction of consecutive preset frames are determined. Finally, based on the lane change intention score and lane change direction of the traffic subject at consecutive preset time periods, the lane change intention of the traffic subject is determined.

[0131] Specifically, the lane-changing intention of the traffic entity is determined based on its lane-changing intention score and lane-changing direction over a consecutive preset number of time periods. This involves determining whether the sum or average of the lane-changing intention scores over these preset number of time periods exceeds a certain score threshold, and whether the lane-changing directions are the same over these preset number of time periods. If the sum or average of the lane-changing intention scores over these preset number of time periods exceeds a certain threshold, and the lane-changing directions are the same over these preset number of time periods, it indicates that the traffic entity's lane-changing probability is high, and its lane-changing intention is to change lanes in that direction. Other determination methods exist in other embodiments, which will not be elaborated here.

[0132] S42: Determine the relative position of the lane where the main traffic vehicle is located relative to the lane where the vehicle is located.

[0133] After determining the lane-changing intention of the traffic vehicle, the relative position of the traffic vehicle's lane to the lane of the vehicle itself is determined. The relative position of the traffic vehicle's lane to the lane of the vehicle itself includes: the current lane (both are in the same lane), the left lane (the traffic vehicle is in the left lane of the vehicle), the right lane (the traffic vehicle is in the right lane of the vehicle), the left outer lane (the traffic vehicle is in the left lane of the vehicle), the right outer lane (the traffic vehicle is in the right lane of the vehicle), and other lanes.

[0134] S43: Determine the movement state of the main traffic vehicle based on the relative positional relationship and lane-changing intention.

[0135] After determining the relative position of the traffic vehicle's lane to the lane of the vehicle and the traffic vehicle's lane-changing intention, the movement state of the traffic vehicle relative to the vehicle is determined based on the relative position and lane-changing intention.

[0136] In this embodiment, the lane-changing intention of the traffic subject is determined based on the lane-changing intention score and lane-changing direction at a consecutive preset number of time points. The relative positional relationship between the lane where the traffic subject is located and the lane where the vehicle is located is also determined. Then, based on the relative positional relationship and lane-changing intention, the motion state of the traffic subject is determined. This refines the steps of determining the motion state of the traffic subject based on the lane-changing intention score and lane-changing direction. First, determining the lane-changing intention based on historical lane-changing intention scores and directions at multiple consecutive time points is more valuable and improves the accuracy of the lane-changing intention. Then, based on the relative positional relationship between the traffic subject and the vehicle and the lane-changing intention, the motion state of the traffic subject relative to the vehicle is accurately described. This provides accurate data support for the decision-making device, reduces the judgment and calculation workload of the decision-making device, and improves the decision-making speed.

[0137] In one embodiment, step S41, which involves determining the lane-changing intention of the traffic vehicle based on its lane-changing intention score and lane-changing direction at consecutive preset time points, specifically includes the following steps:

[0138] S411: Determine whether the lane-changing intention score of the traffic subject at each of the consecutive preset time points is greater than the preset score, and determine whether the lane-changing direction of the traffic subject at each of the consecutive preset time points is the same.

[0139] S412: If the lane-changing intention score of the traffic subject at consecutive preset time points is not greater than the preset score, or the lane-changing direction of the traffic subject at consecutive preset time points is not the same, then the lane-changing intention is determined to be lane keeping.

[0140] If it is determined that the lane-changing intention score of the traffic subject at consecutive preset time points is not greater than the preset score, and there is a possibility that the lane-changing intention score is less than or equal to the preset score, it indicates that the traffic subject is unlikely to change lanes, and the lane-changing intention of the traffic subject can be considered as lane keeping, that is, staying in the current lane; or, if the lane-changing direction of the traffic subject is different at consecutive preset time points, it indicates that the lane-changing direction has changed, which may be a driving skill problem, and the lane-changing intention of the traffic subject is considered as lane keeping.

[0141] S413: If the lane-changing intention score of a traffic subject is greater than the preset score for a consecutive preset number of time periods, and the lane-changing direction of the traffic subject is the same for a consecutive preset number of time periods, then the lane-changing intention is determined to be a lane-changing in the lane-changing direction.

[0142] If the lane-changing intention score of a traffic entity is greater than the preset score for a consecutive preset number of time periods, and the lane-changing direction of the traffic entity is the same for a consecutive preset number of time periods, it indicates that the traffic entity is more likely to change lanes, and the lane-changing intention is determined to be to change lanes in the direction of lane changing. If the lane-changing direction of the traffic entity is to the left for a consecutive preset number of time periods, the lane-changing intention is to change lanes to the left. If the lane-changing direction of the traffic entity is to the right for a consecutive preset number of time periods, the lane-changing intention is to change lanes to the right.

[0143] For example, the preset score can be 5, and the preset number of consecutive time points can be the three most recent time points (3 frames). If the lane change intention scores of the three consecutive frames all exceed the preset score, and the lane change direction obtained from the predicted trajectory is to the left, then the lane change intention of the traffic subject is predicted to be to the left; if the lane change intention scores of the three consecutive frames all exceed the preset score, and the lane change direction obtained from the predicted trajectory is to the right, then the lane change intention of the traffic subject is predicted to be to the right; otherwise, the lane change intention of the traffic subject is predicted to be lane keeping.

[0144] In this embodiment, it is determined whether the lane-changing intention score of the traffic subject at consecutive preset time points is less than a preset score, and whether the lane-changing direction of the traffic subject at consecutive preset time points is the same. If the lane-changing intention score of the traffic subject at consecutive preset time points is not greater than the preset score, or the lane-changing direction of the traffic subject at consecutive preset time points is not the same, then the lane-changing intention is determined to be lane keeping. If the lane-changing intention score of the traffic subject at consecutive preset time points is greater than the preset score, and the lane-changing direction of the traffic subject at consecutive preset time points is the same, then the lane-changing intention is determined to be changing lanes in the lane-changing direction. This clarifies the specific process of determining the lane-changing intention of the traffic subject based on the lane-changing intention score and lane-changing direction of the traffic subject at consecutive preset time points. By referring to multiple historical lane-changing intention scores and lane-changing directions for judgment, the specific lane-changing intention is determined only when the lane-changing intention scores and lane-changing directions of multiple historical lane-changing intentions meet the conditions, thus improving the accuracy of subsequent judgment results.

[0145] In one embodiment, step S43, which involves determining the movement state of the traffic vehicle based on the relative positional relationship and lane-changing intention, specifically includes the following steps:

[0146] S431: When the relative position relationship is the current lane of the vehicle, if the lane changing intentions are lane keeping, lane changing to the left and lane changing to the right, the movement states of the traffic subject are lane keeping straight, lane changing to the left, and lane changing to the right, respectively.

[0147] S432: When the relative position relationship is the left lane of the vehicle, if the lane changing intentions are lane keeping, changing lanes to the left and changing lanes to the right, the movement states of the traffic vehicle are respectively maintaining the straight state of the left lane, moving out of the left lane, and entering the current lane from the left.

[0148] S433: When the relative position relationship is the right lane of the vehicle, if the lane changing intentions are lane keeping, changing lanes to the left and changing lanes to the right, the movement states of the traffic vehicle are respectively maintaining the right lane and going straight, entering the current lane from the right, and leaving the right lane.

[0149] Based on the relative positional relationship between the lane where the main traffic vehicle is located and the lane where the vehicle is located, and the intention of the main traffic vehicle to change lanes, the motion state (driving state) of the main traffic vehicle relative to the vehicle is determined as shown in Table 1. Table 1 is the driving state table of the main traffic vehicle, and the specific contents are as follows:

[0150] Table 1

[0151]

[0152] (1) When the lane where the traffic subject is located is the lane where the vehicle is located: if the lane change intention is to maintain the lane, the movement state of the traffic subject is determined to be maintaining the current lane straight-through state; if the lane change intention is to change lanes to the left, the movement state of the traffic subject is determined to be the current lane left-to-left state; if the lane change intention is to change lanes to the right, the movement state of the traffic subject is determined to be the current lane right-to-right state.

[0153] (2) When the lane where the traffic subject is located is the left lane of the lane where the vehicle is located: if the lane change intention is to maintain the lane, the movement state of the traffic subject is determined to be maintaining the straight state of the left lane; if the lane change intention is to change lanes to the left, the movement state of the traffic subject is determined to be driving out of the left lane; if the lane change intention is to change lanes to the right, the movement state of the traffic subject is determined to be inserting into the current lane from the left.

[0154] (3) When the lane where the traffic subject is located is the right lane of the lane where the vehicle is located: if the lane change intention is to maintain the lane, the movement state of the traffic subject is determined to be maintaining the straight state of the right lane; if the lane change intention is to change lanes to the left, the movement state of the traffic subject is determined to be entering the current lane from the right; if the lane change intention is to change lanes to the right, the movement state of the traffic subject is determined to be leaving the right lane.

[0155] (4) When the lane where the traffic subject is located is the outer lane to the left of the lane where the vehicle is located: if the intention to change lanes is to change lanes to the right, the movement state of the traffic subject is determined to be the state of inserting into the left lane from the outside. If the intention to change lanes is to maintain the lane or change lanes to the left, the movement state of the traffic subject is determined to be the irrelevant state.

[0156] (5) When the lane where the traffic subject is located is the outer lane to the right of the lane where the vehicle is located: if the intention to change lanes is to change lanes to the left, the movement state of the traffic subject is determined to be the state of inserting into the right lane from the outside. If the intention to change lanes is to maintain the lane or change lanes to the right, the movement state of the traffic subject is determined to be the irrelevant state.

[0157] (6) When the lane where the traffic subject is located is not one of the above lanes, the movement state of the traffic subject is determined to be irrelevant.

[0158] In this embodiment, when the relative position relationship is the current lane of the vehicle, if the lane-changing intentions are lane keeping, changing lanes to the left, and changing lanes to the right, the movement states of the traffic subject are respectively: maintaining the current lane and going straight, changing lanes to the left, and changing lanes to the right. When the relative position relationship is the left lane of the vehicle, if the lane-changing intentions are lane keeping, changing lanes to the left, and changing lanes to the right, the movement states of the traffic subject are respectively: maintaining the left lane and going straight, leaving the left lane, and entering the current lane from the left. When the relative position relationship is the right lane of the vehicle, if the lane-changing intentions are lane keeping, changing lanes to the left, and changing lanes to the right, the movement states of the traffic subject are respectively: maintaining the right lane and going straight, entering the current lane from the right, and leaving the right lane. This clarifies the specific process of determining the movement state of the traffic subject based on the relative position relationship between the traffic subject and the vehicle and the lane-changing intentions, providing an accurate basis for autonomous driving decisions.

[0159] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0160] In one embodiment, a motion state determination device is provided, which corresponds one-to-one with the motion state determination method in the above embodiments. For example... Figure 4 As shown, the motion state determination device includes a prediction module 401, a first determination module 402, a calculation module 403, and a second determination module 404. Detailed descriptions of each functional module are as follows:

[0161] The prediction module 401 is used to predict the motion trajectory of traffic entities around the vehicle in real time, so as to obtain the predicted trajectory data of the traffic entities at the current moment.

[0162] The first determining module 402 is used to determine the lane changing direction and target lane of the traffic body lane changing based on the predicted trajectory data, and to determine the target lane line type and target lane attributes of the target lane.

[0163] The calculation module 403 is used to perform weighted scoring calculation on the predicted trajectory data, target lane line type and target lane attributes according to the type of traffic subject, so as to obtain the lane changing intention score of the traffic subject at the current moment.

[0164] The second determining module 404 is used to determine the motion state of the traffic subject based on the lane-changing intention score and lane-changing direction.

[0165] Furthermore, the calculation module 403 is specifically used for:

[0166] Based on the predicted trajectory data of the traffic vehicle, calculate the first intention score of the traffic vehicle to change lanes to the target lane;

[0167] The second intent score is determined based on the target lane line type, and the third intent score is determined based on the target lane attributes.

[0168] The corresponding weights for the first intention score, second intention score, and third intention score are determined based on the type of traffic subject.

[0169] Based on their respective weights, the scores for the first intention, the second intention, and the third intention are weighted and summed to obtain the lane-changing intention score.

[0170] Furthermore, the computing module 403 is specifically used for:

[0171] Based on the predicted trajectory data, the target time when the traffic subject moves to the target lane line is predicted, and the target duration from the current time to the target time is determined. The target lane line is the lane line between the lane where the traffic subject is located and the target lane.

[0172] Determine the center line and width of the lane where the main traffic vehicle is located;

[0173] In the predicted trajectory data, determine the maximum deviation distance between the predicted trajectory of the traffic vehicle and the road centerline of the lane in which the traffic vehicle is located;

[0174] The first intent score is calculated based on the target duration, maximum deviation distance, and lane width of the lane where the main traffic vehicle is located.

[0175] Furthermore, the second determining module 404 is specifically used for:

[0176] The lane-changing intention of the traffic subject is determined based on the lane-changing intention score and lane-changing direction at consecutive preset time points.

[0177] Determine the relative position of the traffic vehicle's lane to the lane of the vehicle itself;

[0178] The movement state of the main traffic vehicle is determined based on the relative positional relationship and the intention to change lanes.

[0179] Furthermore, the second determining module 404 is specifically used for:

[0180] Determine whether the lane-changing intention score of the traffic subject at each of the consecutive preset time points is greater than the preset score, and determine whether the lane-changing direction of the traffic subject at each of the consecutive preset time points is the same;

[0181] If the lane-changing intention score of a traffic subject is not greater than the preset score for a consecutive preset number of time periods, or if the lane-changing direction of the traffic subject is different for a consecutive preset number of time periods, then the lane-changing intention is determined to be lane keeping.

[0182] If the lane-changing intention score of a traffic entity is greater than the preset score for a consecutive preset number of time periods, and the lane-changing direction of the traffic entity is the same for a consecutive preset number of time periods, then the lane-changing intention is determined to be a lane-changing in the direction of lane changing.

[0183] Furthermore, the second determining module 404 is specifically used for:

[0184] When the relative position relationship is the current lane of the vehicle, if the lane changing intentions are lane keeping, lane changing to the left and lane changing to the right, the movement states of the traffic subject are lane keeping and going straight, lane changing to the left and lane changing to the right, respectively.

[0185] When the relative position relationship is the left lane of the vehicle, if the lane changing intentions are lane keeping, changing lanes to the left and changing lanes to the right, the movement states of the traffic subject are respectively maintaining the straight state in the left lane, moving out of the left lane, and entering the current lane from the left.

[0186] When the relative position relationship is the right lane of the vehicle, if the lane changing intentions are lane keeping, changing lanes to the left, and changing lanes to the right, the movement states of the traffic vehicle are respectively maintaining the straight state of the right lane, entering the current lane from the right, and leaving the right lane.

[0187] Furthermore, the prediction module 401 is specifically used for:

[0188] The vehicle's environmental perception system collects driving perception data in real time, including the vehicle's status data, the status information of traffic entities around the vehicle, and the road information on which the vehicle is traveling.

[0189] The status information of the vehicle and the traffic entity is converted into the world coordinate system and filtered to obtain filtered status data.

[0190] The center line of the lane where the vehicle is located is determined based on the driving road information. The center line of the lane is used as a reference line to perform Frenet coordinate system transformation on the filtered state data to obtain preprocessed data.

[0191] The preprocessed data is input into the trajectory prediction model so that the trajectory prediction model can predict the movement trajectory of the traffic subject and obtain the predicted trajectory data of the traffic subject at the current time.

[0192] Specific limitations regarding the motion state determination device can be found in the limitations of the motion state determination method described above, and will not be repeated here. Each module in the aforementioned motion state determination device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0193] In one embodiment, a motion state determination device is provided, which may be an in-vehicle terminal. The motion state determination device includes a processor, a memory, and a network interface connected via a system bus. The processor provides computational and control capabilities. The memory of the computer device includes a storage medium and internal memory. The storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the storage medium. The network interface of the motion state determination device is used to communicate with an environmental perception system via a network connection. When the computer program is executed by the processor, it implements a motion state determination method.

[0194] In one embodiment, such as Figure 5 As shown, a motion state determination device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the motion state determination method described above.

[0195] In one embodiment, a readable storage medium is provided on which a computer program is stored, which, when executed by a processor, implements the steps of the motion state determination method described above.

[0196] 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 can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Furthermore, any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory.

[0197] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0198] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for determining motion state, characterized in that, include: Real-time prediction of the movement trajectories of traffic entities around the vehicle to obtain the predicted trajectory data of the traffic entities at the current moment; Based on the predicted trajectory data, the lane-changing direction and target lane of the main traffic vehicle are determined, and the target lane line type and target lane attributes of the target lane are determined. Based on the type of the traffic subject, a weighted score is calculated on the predicted trajectory data, the target lane line type, and the target lane attribute to obtain the lane-changing intention score of the traffic subject at the current moment. The motion state of the traffic vehicle is determined based on the lane-changing intention score and the lane-changing direction; The step of calculating a weighted score based on the predicted trajectory data, the target lane line type, and the target lane attribute according to the type of the traffic subject to obtain the lane-changing intention score of the traffic subject includes: Based on the predicted trajectory data of the traffic vehicle, calculate the first intention score of the traffic vehicle to change lanes to the target lane; A second intent score is determined based on the target lane line type, and a third intent score is determined based on the target lane attributes; The corresponding weights of the first intention score, the second intention score, and the third intention score are determined based on the type of the traffic subject. Based on the corresponding weights, the first intention score, the second intention score, and the third intention score are weighted and summed to obtain the lane-changing intention score.

2. The motion state determination method as described in claim 1, characterized in that, The step of calculating the first intention score of the traffic vehicle to change lanes to the target lane based on the predicted trajectory data of the traffic vehicle includes: Based on the predicted trajectory data, the target time when the traffic vehicle moves to the target lane line is predicted, and the target duration from the current time to the target time is determined. The target lane line is the lane line between the lane where the traffic vehicle is located and the target lane. Determine the lane centerline and lane width of the lane where the main traffic vehicle is located; In the predicted trajectory data, the maximum deviation distance between the predicted trajectory of the traffic vehicle and the road centerline of the lane in which the traffic vehicle is located is determined; The first intent score is calculated based on the target duration, the maximum deviation distance, and the lane width of the lane where the traffic subject is located.

3. The motion state determination method as described in claim 1, characterized in that, Determining the motion state of the traffic vehicle based on the lane-changing intention score and the lane-changing direction includes: The lane-changing intention of the traffic subject is determined based on the lane-changing intention score and lane-changing direction at a consecutive preset number of time points. Determine the relative positional relationship between the lane where the traffic vehicle is located and the lane where the vehicle is located; The movement state of the traffic vehicle is determined based on the relative positional relationship and the lane-changing intention.

4. The motion state determination method as described in claim 3, characterized in that, Determining the lane-changing intention of the traffic vehicle based on its lane-changing intention score and lane-changing direction at consecutive preset time points includes: Determine whether the lane-changing intention score of the traffic subject at consecutive preset time points is greater than a preset score, and determine whether the lane-changing direction of the traffic subject at consecutive preset time points is the same; If the lane-changing intention score of the traffic subject is not greater than the preset score for a consecutive preset number of time periods, or if the lane-changing direction of the traffic subject is not the same for a consecutive preset number of time periods, then the lane-changing intention is determined to be lane keeping. If the lane-changing intention score of the traffic subject is greater than the preset score for a consecutive preset number of time periods, and the lane-changing direction of the traffic subject is the same for a consecutive preset number of time periods, then the lane-changing intention is determined to be a lane-changing in the lane-changing direction.

5. The motion state determination method as described in claim 3, characterized in that, Determining the movement state of the traffic vehicle based on the relative positional relationship and the lane-changing intention includes: When the relative position relationship is the current lane of the vehicle, if the lane changing intentions are lane keeping, lane changing to the left, and lane changing to the right, then the movement states of the traffic subject are lane keeping straight in the current lane, lane changing to the left in the current lane, and lane changing to the right in the current lane, respectively. When the relative position relationship is the left lane of the vehicle, if the lane changing intentions are lane keeping, lane changing to the left and lane changing to the right, then the movement states of the traffic subject are respectively maintaining the straight state in the left lane, driving out of the left lane, and inserting into the current lane from the left. When the relative position relationship is the right lane of the vehicle, if the lane changing intentions are lane keeping, lane changing to the left, and lane changing to the right, then the movement states of the traffic subject are respectively maintaining the straight state in the right lane, entering the current lane from the right, and leaving the right lane.

6. The motion state determination method according to any one of claims 1-5, characterized in that, The real-time prediction of the motion trajectories of traffic entities surrounding the vehicle to obtain the predicted trajectory data of the traffic entities at the current moment includes: The vehicle's environmental perception system collects driving perception data in real time, including the vehicle's status data, the status information of traffic entities around the vehicle, and the vehicle's driving road information. The status information of the vehicle and the status information of the traffic entity are converted into the world coordinate system and filtered to obtain filtered status data. Based on the driving road information, the center line of the lane where the vehicle is located is determined. Using the center line of the lane as a reference line, the filtered state data is transformed into Frenet coordinate system to obtain preprocessed data. The preprocessed data is input into the trajectory prediction model so that the trajectory prediction model can predict the movement trajectory of the traffic subject and obtain the predicted trajectory data of the traffic subject at the current time.

7. A motion state determination device, characterized in that, include: The prediction module is used to predict the movement trajectory of traffic entities around the vehicle in real time, so as to obtain the predicted trajectory data of the traffic entities at the current moment. The first determining module is used to determine the lane changing direction and target lane of the traffic vehicle lane changing based on the predicted trajectory data, and to determine the target lane line type and target lane attribute of the target lane. The calculation module is used to perform weighted scoring calculation on the predicted trajectory data, the target lane line type and the target lane attribute according to the type of the traffic subject, so as to obtain the lane-changing intention score of the traffic subject at the current time. The second determining module is used to determine the motion state of the traffic subject based on the lane-changing intention score and the lane-changing direction; The step of calculating a weighted score based on the predicted trajectory data, the target lane line type, and the target lane attribute according to the type of the traffic subject to obtain the lane-changing intention score of the traffic subject includes: Based on the predicted trajectory data of the traffic vehicle, calculate the first intention score of the traffic vehicle to change lanes to the target lane; A second intent score is determined based on the target lane line type, and a third intent score is determined based on the target lane attributes; The corresponding weights of the first intention score, the second intention score, and the third intention score are determined based on the type of the traffic subject. Based on the corresponding weights, the first intention score, the second intention score, and the third intention score are weighted and summed to obtain the lane-changing intention score.

8. A motion state determination device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the motion state determination method as described in any one of claims 1 to 6.

9. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the motion state determination method as described in any one of claims 1 to 6.

Images