Simulation trajectory data acquisition method, device and equipment

Abstract

The invention provides a simulation trajectory data acquisition method, device and equipment, and the method comprises the steps: obtaining the actual trajectory data of an actual vehicle in the driving process of the actual vehicle; at a data acquisition moment, establishing a simulation vehicle corresponding to the actual vehicle, and taking the fixed motion parameter as a target motion parameter; starting from the simulation starting moment, controlling the simulation vehicle to run in the traffic simulation model based on the target motion parameters; in the driving process of the simulation vehicle, obtaining simulation track data of the simulation vehicle; predicting predicted trajectory data of the simulation vehicle in the prediction time period at each first update moment; and determining a first motion parameter based on the predicted trajectory data of the prediction time period and the actual trajectory data of the prediction time period, and updating the first motion parameter to a target motion parameter of the simulation vehicle. Through the technical scheme of the invention, the complete track data of the actual vehicle can be obtained.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method, device and equipment for acquiring simulation trajectory data. Background Art

[0002] During the driving process of the vehicle, the trajectory data of the vehicle can be obtained, and the trajectory data can include multiple trajectory points. For each trajectory point, the trajectory point can include a time point and a position (such as a latitude and longitude position or a position in a world coordinate system), indicating that the vehicle is at that position at that time point. Obviously, based on the trajectory data of the vehicle, the positioning of the vehicle can be realized, and functions such as vehicle detection and traffic flow control can be realized.

[0003] In order to obtain the vehicle's trajectory data, it is necessary to deploy a large number of positioning devices (such as cameras, etc.) on roads (such as highways). When the vehicle passes through these positioning devices, the positioning devices can obtain the vehicle's trajectory points (time points and locations) and send the trajectory points to the management device (such as a server, etc.). The management device summarizes the vehicle's trajectory data based on the trajectory points sent by multiple positioning devices.

[0004] However, due to limited deployment conditions and insufficient equipment density, there will inevitably be detection blind spots on roads (there is no positioning equipment inside the detection blind spots), resulting in the inability to obtain the trajectory data of the vehicle in the detection blind spots. The lack of trajectory data will affect subsequent vehicle detection, traffic control and other functions. Summary of the invention

[0005] The present application provides a method for acquiring simulation trajectory data, the method comprising:

[0006] During the driving process of the actual vehicle, obtaining actual trajectory data of the actual vehicle;

[0007] At the configured data collection time, a simulated vehicle corresponding to the actual vehicle is established, and fixed motion parameters are used as target motion parameters; a simulation start time is determined based on the data collection time and the acquired delay duration, and starting from the simulation start time, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameters; wherein the simulated position of the simulated vehicle at the simulation start time corresponds to the actual position of the actual vehicle at the data collection time;

[0008] During the driving process of the simulated vehicle, obtaining simulated trajectory data of the simulated vehicle;

[0009] When the actual vehicle is in the sensing area in front of the detection blind area, obtaining a plurality of first update moments;

[0010] At each first update time, predict the predicted trajectory data of the simulated vehicle in the prediction period, the predicted trajectory data is the trajectory data of the simulated vehicle when it travels in the traffic simulation model based on the target motion parameters, the end time of the prediction period is determined based on the first update time, and the start time of the prediction period is determined based on the first update time and the delay time;

[0011] A first motion parameter is determined based on the predicted trajectory data of the predicted period and the actual trajectory data of the predicted period, and the first motion parameter is updated to a target motion parameter of the simulated vehicle.

[0012] The present application provides a simulation trajectory data acquisition device, the device comprising:

[0013] An actual data processing module, used for acquiring actual track data of the actual vehicle during the driving process of the actual vehicle, and storing the actual track data of the actual vehicle;

[0014] A simulation data processing module is used to establish a simulated vehicle corresponding to the actual vehicle at a configured data collection time, and use fixed motion parameters as target motion parameters; determine a simulation start time based on the data collection time and the acquired delay time, and control the simulated vehicle to travel in a traffic simulation model based on the target motion parameters starting from the simulation start time; wherein the simulated position of the simulated vehicle at the simulation start time corresponds to the actual position of the actual vehicle at the data collection time; during the travel of the simulated vehicle, obtain simulated trajectory data of the simulated vehicle, and store the simulated trajectory data of the simulated vehicle;

[0015] A parameter data processing module is used to obtain multiple first update moments when the actual vehicle is in the perception area in front of the detection blind spot; at each first update moment, predict the predicted trajectory data of the simulated vehicle in the prediction period, the predicted trajectory data is the trajectory data of the simulated vehicle when it is traveling in the traffic simulation model based on the target motion parameters, the end time of the prediction period is determined based on the first update moment, and the start time of the prediction period is determined based on the first update moment and the delay duration; determine the first motion parameter based on the predicted trajectory data of the prediction period and the actual trajectory data of the prediction period, and update the first motion parameter to the target motion parameter of the simulated vehicle.

[0016] The present application provides an electronic device, comprising: a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions that can be executed by the processor; the processor is used to execute the machine-executable instructions to implement the simulation trajectory data acquisition method disclosed in the above example.

[0017] The present application provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the simulation trajectory data acquisition method disclosed in the above example is implemented.

[0018] The present application provides a machine-readable storage medium, which stores machine-executable instructions that can be executed by a processor; wherein the processor is used to execute the machine-executable instructions to implement the simulation trajectory data acquisition method disclosed in the above example.

[0019] As can be seen from the above technical solutions, in the embodiment of the present application, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameters. During the driving process of the simulated vehicle, the simulated trajectory data of the simulated vehicle is obtained, such as the simulated trajectory data of the simulated vehicle in the detection blind spot, and then the trajectory data of the actual vehicle is completed based on the simulated trajectory data of the simulated vehicle in the detection blind spot, and the complete trajectory data of the actual vehicle is obtained, thereby realizing functions such as vehicle detection and traffic control. The driving behavior model parameters (i.e., target motion parameters) of the simulated vehicle can be dynamically adjusted to obtain accurate and reliable target motion parameters, improve the trajectory completion accuracy, and realize the smooth connection between the simulated trajectory and the real trajectory. Based on the high operating efficiency and data caching mechanism of the traffic simulation model, the real-time output of the result is improved, and the running trajectory of the actual vehicle involving the detection blind spot can be simulated and completed with a delay of seconds. By using the driving behavior model parameters with a higher degree of abstraction as the behavioral characteristics of the actual vehicle, considering the interactive motion between multiple actual vehicles, making full use of the trajectory data in the perception area before and after the detection blind spot, and considering the behavioral differences between the actual vehicle individuals, the accuracy of the trajectory completion result of the detection blind spot can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flowchart of a method for acquiring simulation trajectory data in one embodiment of the present application;

[0021] Figure 2 is a schematic diagram of an actual vehicle driving process in one embodiment of the present application;

[0022] Figure 3 It is a schematic diagram of the process of obtaining simulation trajectory data in one embodiment of the present application;

[0023] Figure 4 is a schematic diagram of a dynamic adjustment process of target motion parameters in one embodiment of the present application;

[0024] Figure 5 is a schematic diagram of a dynamic adjustment process of target motion parameters in one embodiment of the present application;

[0025] Fig. 6AIt is an overall logic diagram of trajectory completion in one embodiment of the present application;

[0026] Figure 6B It is a brief schematic diagram of the process of trajectory completion in one embodiment of the present application;

[0027] Figure 7 It is a structural schematic diagram of a simulation trajectory data acquisition device in one embodiment of the present application;

[0028] Figure 8 It is a hardware structure diagram of an electronic device in one embodiment of the present application. DETAILED DESCRIPTION

[0029] In the embodiment of the present application, a simulation trajectory data acquisition method is proposed, which can be applied to management equipment, see Figure 1 FIG. 4 is a flow chart of the method, which may include:

[0030] Step 101: Acquire actual trajectory data of an actual vehicle during its driving process.

[0031] Step 102: At the configured data collection time, a simulated vehicle corresponding to the actual vehicle is established, and the fixed motion parameters are used as target motion parameters. The simulation start time is determined based on the data collection time and the acquired delay time. Starting from the simulation start time, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameters; during the driving process of the simulated vehicle, the simulated trajectory data of the simulated vehicle is obtained. Among them, the simulated position of the simulated vehicle at the simulation start time (the position in the simulated trajectory of the simulated vehicle is called the simulated position) corresponds to the actual position of the actual vehicle at the data collection time (that is, the position in the actual trajectory of the actual vehicle is called the actual position), that is, the simulated position at the simulation start time is the same as the actual position at the data collection time. Obviously, when the actual vehicle is at position A at the data collection time, after the simulated vehicle is established, the simulated vehicle is not at position A at the data collection time, but at position A at the simulation start time (such as data collection time + delay time), that is, it is delayed for a period of time to be at position A.

[0032] Step 103, when the actual vehicle is in the perception area in front of the detection blind area, multiple first update moments are obtained; at each first update moment, the predicted trajectory data of the simulated vehicle in the prediction period is predicted, and the predicted trajectory data is the trajectory data of the simulated vehicle when it is traveling in the traffic simulation model based on the target motion parameters, and the end time of the prediction period is determined based on the first update moment, and the start time of the prediction period is determined based on the first update moment and the delay duration. Among them, the perception area refers to the area where the actual trajectory data of the actual vehicle can be obtained, and the detection blind area refers to the area where the actual trajectory data of the actual vehicle cannot be obtained, that is, the actual trajectory of the actual vehicle is lost in the detection blind area.

[0033] Step 104 : determining a first motion parameter based on the predicted trajectory data of the prediction period and the actual trajectory data of the prediction period, and updating the first motion parameter to a target motion parameter of the simulated vehicle.

[0034] Exemplarily, if the actual trajectory data of the actual vehicle does not include the trajectory data between the first moment and the second moment, when the actual vehicle is in the detection blind spot, multiple second update moments can also be obtained; at each second update moment, the second motion parameter is determined based on the first average speed and the second average speed corresponding to the target section, and the second motion parameter is updated to the target motion parameter of the simulated vehicle. Wherein, the target section can be the section between the position of the actual vehicle at the first moment and the position of the actual vehicle at the second moment, the first average speed can represent the average speed of multiple actual vehicles when passing through the target section, and the second average speed can represent the average speed of all simulated vehicles in the target section at the second update moment. Wherein, multiple first update moments can be located between the simulation start moment and the target moment (when the actual vehicle is in the area between the simulation start moment and the target moment, it means that the actual vehicle is in the perception area in front of the detection blind spot); multiple second update moments can be located between the target moment and the second moment (when the actual vehicle is in the area between the target moment and the second moment, it means that the actual vehicle is in the detection blind spot); the target moment can be the sum of the first moment and the delay duration.

[0035] Exemplarily, when obtaining multiple first update moments, the first first update moment may be the simulation start moment, the interval between two adjacent first update moments may be the configured first simulation step length, and the first simulation step length may be smaller than the delay duration. When obtaining multiple second update moments, the first second update moment is the target moment, the interval between two adjacent second update moments is the configured second simulation step length, and the second simulation step length may be larger than the delay duration, or smaller than or equal to the delay duration.

[0036] Among them, when determining the starting time of the prediction period based on the first update time and the delay duration, the difference between the first update time and the delay duration is determined as the starting time of the prediction period; when determining the end time of the prediction period based on the first update time, it can be determined whether the first update time is greater than the first time; if not, the first update time can be determined as the end time of the prediction period; if so, the first time can be determined as the end time of the prediction period.

[0037] The delay time is determined based on the length of the road section in the detection blind spot. The longer the road section in the detection blind spot is, the longer the delay time is. The relationship between the road section length and the delay time is not limited in this embodiment. As long as the delay time is proportional to the road section length, the delay time can be determined based on the road section length.

[0038] Exemplarily, determining the first motion parameter based on the predicted trajectory data of the prediction period and the actual trajectory data of the prediction period may include but is not limited to: if the actual trajectory data includes the actual positions of multiple time points in the prediction period, the predicted trajectory data includes the predicted positions of multiple time points in the prediction period, then determining the distance (such as Euclidean distance) between the predicted position of each time point and the actual position of the time point; based on the distance corresponding to each time point, calculating the distance error between the predicted trajectory data and the actual trajectory data. Based on this, if the distance error is greater than a preset error threshold, the target motion parameter currently used by the simulated vehicle can be adjusted, and the adjusted motion parameter is determined as the first motion parameter of the simulated vehicle; or, if the distance error is not greater than the preset error threshold, the target motion parameter currently used by the simulated vehicle can be determined as the first motion parameter of the simulated vehicle.

[0039] Exemplarily, determining the second motion parameter based on the first average speed and the second average speed corresponding to the target road section may include, but is not limited to: if the target motion parameter currently used by the simulated vehicle includes the driving speed, then the driving speed in the target motion parameter may be adjusted based on the first average speed and the second average speed, and the adjusted motion parameter may be determined as the second motion parameter of the simulated vehicle.

[0040] Exemplarily, the driving speed in the target motion parameter may be adjusted using the following formula: in, Indicates the adjusted driving speed (the adjusted motion parameters include the adjusted driving speed), represents the driving speed in the target motion parameters, represents the second average speed, Indicates the first average speed.

[0041] Exemplarily, controlling a simulated vehicle to travel in a traffic simulation model based on target motion parameters may include, but is not limited to: during the driving process of an actual vehicle, establishing a simulated vehicle for associated vehicles (such as one or more associated vehicles) around the actual vehicle, and controlling the simulated vehicle of the associated vehicles to travel in the traffic simulation model; based on the simulation trajectory data of the simulated vehicle of the associated vehicles in the traffic simulation model, controlling the simulated vehicle of the actual vehicle to travel in the traffic simulation model based on the target motion parameters.

[0042] Exemplarily, after obtaining the simulated trajectory data of the simulated vehicle, the simulated trajectory data between the first moment and the second moment can be selected from the simulated trajectory data, and the selected simulated trajectory data can be fused with the actual trajectory data to obtain fused trajectory data, and the vehicle management of the actual vehicle can be implemented based on the fused trajectory data. Alternatively, after obtaining the simulated trajectory data of the simulated vehicle, the simulated trajectory data can be output, and the vehicle management of the actual vehicle can be implemented based on the simulated trajectory data. For example, vehicle management can include but is not limited to functions such as vehicle detection and traffic flow control, and there is no restriction on this vehicle management.

[0043] As can be seen from the above technical solutions, in the embodiment of the present application, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameters. During the driving process of the simulated vehicle, the simulated trajectory data of the simulated vehicle is obtained, such as the simulated trajectory data of the simulated vehicle in the detection blind spot, and then the trajectory data of the actual vehicle is completed based on the simulated trajectory data of the simulated vehicle in the detection blind spot, and the complete trajectory data of the actual vehicle is obtained, thereby realizing functions such as vehicle detection and traffic control. The driving behavior model parameters (i.e., target motion parameters) of the simulated vehicle can be dynamically adjusted to obtain accurate and reliable target motion parameters, improve the trajectory completion accuracy, and realize the smooth connection between the simulated trajectory and the real trajectory. Based on the high operating efficiency and data caching mechanism of the traffic simulation model, the real-time output of the result is improved, and the running trajectory of the actual vehicle involving the detection blind spot can be simulated and completed with a delay of seconds. By using the driving behavior model parameters with a higher degree of abstraction as the behavioral characteristics of the actual vehicle, considering the interactive motion between multiple actual vehicles, making full use of the trajectory data in the perception area before and after the detection blind spot, and considering the behavioral differences between the actual vehicle individuals, the accuracy of the trajectory completion result of the detection blind spot can be improved.

[0044] The above technical solutions of the embodiments of the present application are described below in combination with specific application scenarios.

[0045] Due to limited deployment conditions and insufficient equipment density, there will inevitably be detection blind spots on roads (such as expressways) (there is no positioning equipment inside the detection blind spots), resulting in the inability to obtain the trajectory data of the vehicle in the detection blind spots. The lack of trajectory data will affect functions such as vehicle detection and traffic control.

[0046] In view of the above findings, a method for acquiring simulated trajectory data is proposed in an embodiment of the present application, which can realize real-time completion of the trajectory of the detection blind spot based on a traffic simulation model (such as a microscopic traffic simulation model). For example, the simulated trajectory data of a simulated vehicle (such as the simulated trajectory data of a simulated vehicle in a detection blind spot) can be obtained, and then the trajectory data of the actual vehicle can be completed based on the simulated trajectory data of the simulated vehicle in the detection blind spot to obtain the complete trajectory data of the actual vehicle, thereby realizing functions such as vehicle detection and traffic flow control.

[0047] A simulation trajectory data acquisition method is proposed in an embodiment of the present application. The method can be applied to a management device (also referred to as a control device). The management device may include but is not limited to a server, a personal computer, a laptop computer, a smart phone, etc., and there is no restriction on the type of the management device.

[0048] Exemplarily, the method for acquiring simulation trajectory data may involve the following process:

[0049] First, obtain the actual trajectory data of the actual vehicle.

[0050] Exemplarily, vehicles actually traveling on the road can be called actual vehicles. During the driving process of the actual vehicles, the management device can obtain the trajectory data of the actual vehicles. For the convenience of distinction, the trajectory data of the actual vehicles can be called actual trajectory data, that is, the actual driving trajectory data.

[0051] The actual trajectory data may include multiple trajectory points. For each trajectory point, the trajectory point may include a time point and a location (such as a latitude and longitude location), indicating that the actual vehicle is at that location at that time point. Based on the actual trajectory data, the actual vehicle can be positioned, and functions such as vehicle detection and traffic flow control can be realized.

[0052] Exemplarily, in order to obtain the actual trajectory data of the actual vehicle, a large number of positioning devices (such as cameras, radar cameras, etc., there is no restriction on the type of positioning devices, as long as the trajectory points of the actual vehicle can be obtained) can be deployed on roads (such as highways, etc.). When the actual vehicle passes through these positioning devices, the positioning device can obtain the trajectory points (time points and positions) of the actual vehicle, and send the trajectory points to the management device, which summarizes the actual trajectory data based on the trajectory points sent by multiple positioning devices.

[0053] Exemplarily, during the driving process of an actual vehicle, the driving area of ​​the actual vehicle can be divided into a perception area and a detection blind area. The perception area is an area where a positioning device is deployed, that is, the trajectory points of the actual vehicle in the perception area can be obtained, and the actual trajectory data includes the trajectory points of the actual vehicle in the perception area. For example, when the actual vehicle is in the perception area, the position, speed, license plate and other information of the actual vehicle can be collected in real time by the positioning device, so as to obtain the trajectory points of the actual vehicle in the perception area. The detection blind area is an area where no positioning device is deployed, that is, the trajectory points of the actual vehicle in the detection blind area cannot be obtained, and the actual trajectory data does not include the trajectory points of the actual vehicle in the detection blind area. For example, when the actual vehicle is in the detection blind area, the position, speed, license plate and other information of the actual vehicle disappear from the current collected data.

[0054] For example, see Figure 2 As shown, it is a schematic diagram of the actual vehicle driving process. During the driving process of the actual vehicle, the actual vehicle passes through area A, area B and area C in sequence. The management device can obtain the actual trajectory data of the actual vehicle and cache the actual trajectory data of the actual vehicle. For example, area A and area C are perception areas, and area B is a detection blind area. Therefore, when the actual vehicle is in area A, the trajectory points of the actual vehicle can be obtained. When the actual vehicle is in area B, the trajectory points of the actual vehicle cannot be obtained. When the actual vehicle is in area C, the trajectory points of the actual vehicle can be obtained. Based on this, the actual trajectory data of the actual vehicle can include trajectory points in area A and trajectory points in area C.

[0055] Second, obtain the simulation trajectory data of the simulated vehicle.

[0056] Exemplarily, the traffic simulation model can also be called a microscopic traffic simulation model, and the movement behavior of the actual vehicle can be simulated in the traffic simulation model based on the driving behavior parameters to restore the traffic flow phenomenon in the real scene. When the traffic simulation model simulates the movement behavior of the actual vehicle, the driving environment in which the actual vehicle is currently located, the influence of the surrounding vehicles of the actual vehicle, and the driving behavior parameters of the actual vehicle (such as following parameters and lane changing parameters, etc.) can be referred to. The traffic simulation model has the advantages of high operating efficiency, being able to meet the needs of real-time applications, the simulation results meeting the traffic flow theory, and the simulation results meeting the physical constraints.

[0057] For example, a simulated vehicle can be established for an actual vehicle (i.e., a simulated vehicle is established for an actual vehicle in a traffic simulation model), so as to control the simulated vehicle to travel in the traffic simulation model (i.e., simulate the movement behavior of the actual vehicle and travel in the traffic simulation model), and during the travel of the simulated vehicle, the trajectory data of the simulated vehicle is obtained. For the sake of distinction, the trajectory data of the simulated vehicle can be referred to as simulated trajectory data, which is not the trajectory data of the actual travel, but the trajectory data of the simulated travel.

[0058] The simulation trajectory data may include a plurality of trajectory points. For each trajectory point, the trajectory point may include a time point and a position (such as a latitude and longitude position), indicating that the simulation vehicle is at the position at the time point.

[0059] In one possible implementation, see Figure 3 As shown, it is a schematic diagram of a process of obtaining the simulated trajectory data of the simulated vehicle. The process of obtaining the simulated trajectory data of the simulated vehicle may include:

[0060] Step 301: At the configured data collection moment, a simulated vehicle corresponding to the actual vehicle is established, and the configured fixed motion parameters are used as target motion parameters corresponding to the simulated vehicle.

[0061] For example, during the driving process of the actual vehicle, the management device can configure the data collection time for the actual vehicle. There is no restriction on the data collection time, and it can be configured arbitrarily. For example, a certain time before the actual vehicle enters the detection blind spot is used as the data collection time, and the data collection time is recorded as t sim .

[0062] For example, a road network model can be built for a road section containing a detection blind spot (excluding an entrance or exit), that is, a traffic simulation model is started for the road section, and the traffic simulation model is used to restore the traffic flow phenomenon in the real scene. On this basis, at the data collection time t sim The management device can establish a simulated vehicle corresponding to the actual vehicle, that is, establish a simulated vehicle in the traffic simulation model, so as to control the simulated vehicle to travel in the traffic simulation model, that is, simulate the movement behavior of the actual vehicle and control the simulated vehicle to travel in the traffic simulation model.

[0063] Exemplarily, the target motion parameter is the motion parameter used when the simulated vehicle is traveling in the traffic simulation model, that is, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameter. For example, the target motion parameter may include but is not limited to a following parameter and / or a lane changing parameter, the following parameter may include a driving speed and a distance from a preceding vehicle, and the lane changing parameter may include a lane changing probability, and there is no restriction on the target motion parameter.

[0064] The driving speed indicates the speed adopted by the simulated vehicle when driving in the traffic simulation model, that is, the simulated vehicle is controlled to drive in the traffic simulation model at the driving speed. The front vehicle distance indicates the distance between the simulated vehicle and the front vehicle when driving in the traffic simulation model, that is, when the simulated vehicle is driving in the traffic simulation model, the distance between the simulated vehicle and the front simulated vehicle is controlled to be the front vehicle distance. The lane change probability indicates the probability of the simulated vehicle switching from one lane to another lane when driving in the traffic simulation model. For example, assuming that the simulated vehicle is driving in lane A, if the distance between the simulated vehicle and the front simulated vehicle is less than a threshold, when it is possible to switch to lane B, the lane change probability is used to decide whether to switch from lane A to lane B.

[0065] When establishing a simulated vehicle corresponding to an actual vehicle, the configured fixed motion parameters can be used as the target motion parameters corresponding to the simulated vehicle. For example, the configured fixed driving speed, fixed distance to the preceding vehicle, and fixed lane change probability can be used as the target motion parameters corresponding to the simulated vehicle. During the driving process of the simulated vehicle, the target motion parameters can be dynamically updated. After each update of the target motion parameters, the latest target motion parameters are used to control the simulated vehicle to drive in the traffic simulation model.

[0066] For example, at the data collection time t sim , the fixed motion parameter a1 is used as the target motion parameter, and the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameter a1. At a certain moment, if the target motion parameter is updated to the target motion parameter a2, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameter a2. At a certain moment, if the target motion parameter is updated to the target motion parameter a3, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameter a3, and so on.

[0067] For example, the management device may set the same data collection time t for multiple actual vehicles. sim The management device can also set the data collection time t for each actual vehicle separately. sim , there is no restriction on this.

[0068] Step 302: starting from the simulation start time, the interval between the simulation start time and the data collection time is the delay time, and controlling the simulated vehicle to travel in the traffic simulation model based on the target motion parameters.

[0069] For example, after establishing a simulated vehicle corresponding to an actual vehicle, the simulated vehicle is not immediately controlled to travel in a traffic simulation model based on target motion parameters, but instead waits for a period of time, starting from the simulation start time, and then controls the simulated vehicle to travel in the traffic simulation model based on target motion parameters.

[0070] For example, the delay time can be pre-configured. The delay time can be configured based on experience, and there is no restriction on the value of the delay time, such as 5 seconds, 10 seconds, etc. The delay time is recorded as t cache Alternatively, the delay time may be determined based on the length of the road section of the detection blind spot. The longer the road section of the detection blind spot is, the longer the delay time is. The shorter the road section of the detection blind spot is, the shorter the delay time is.

[0071] At the simulation startup time and data collection time t sim When the interval is the delay time, the simulation start time is t sim +t cache , that is, from the simulation start time t sim +t cache At the beginning, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameters. Based on the above processing, a buffer can be cached for a period of time (i.e., the delay time t cache ), and then update the target motion parameters based on the cached actual trajectory data (see subsequent embodiments for the updating process), so as to obtain accurate and reliable target motion parameters.

[0072] It should be noted that when controlling the simulated vehicle to travel in the traffic simulation model based on the target motion parameters, since the target motion parameters are continuously updated, each time the target motion parameters are updated, the updated target motion parameters are used to control the simulated vehicle to travel in the traffic simulation model.

[0073] Exemplary, when controlling the simulated vehicle to travel in the traffic simulation model based on the target motion parameters, multiple simulated vehicles can also be linked to travel together in the traffic simulation model. For example, in the driving process of the actual vehicle, a simulated vehicle can be established for the associated vehicles (such as one or more associated vehicles, which are also actual vehicles) around the actual vehicle. Based on this, the simulated vehicle of the associated vehicle can be controlled to travel in the traffic simulation model, and the simulated vehicle of the actual vehicle can be controlled to travel in the traffic simulation model based on the target motion parameters. For example, based on the simulation trajectory data of the simulated vehicle of the associated vehicle in the traffic simulation model, the simulated vehicle of the actual vehicle is controlled to travel in the traffic simulation model based on the target motion parameters, that is, the trajectory of the simulated vehicle of the associated vehicle in the traffic simulation model, which will affect the trajectory of the simulated vehicle of the actual vehicle in the traffic simulation model, that is, the trajectories of different simulated vehicles will affect each other.

[0074] For example, when controlling the simulated vehicle of the actual vehicle to travel in the traffic simulation model, it is necessary to refer to the trajectory of the simulated vehicle of the associated vehicle in the traffic simulation model to ensure that the trajectory of the simulated vehicle of the actual vehicle does not conflict with the trajectory of the simulated vehicle of the associated vehicle, and no collision or other abnormalities occur.

[0075] For example, in the same traffic simulation model, a corresponding simulated vehicle can be generated for each perceived actual vehicle. The movement of these simulated vehicles is constrained and affected by factors such as the distance and speed difference between each other, and they interact based on target motion parameters (such as driving speed, distance to the vehicle in front, lane change probability, etc.), thereby ensuring the rationality of the trajectory completion results while improving their accuracy.

[0076] For example, two real vehicles with different driving speeds enter the detection blind spot from the same lane one after another. The slow real vehicle A enters first, and the fast real vehicle B enters later. If their trajectories are completed separately, it is possible that the fast real vehicle B behind will directly "run over" the slow real vehicle A in front of it. Based on this, if the trajectories of real vehicles A and B are completed based on the traffic simulation model at the same time, the fast real vehicle B behind will be blocked by the slow real vehicle A in front, and thus choose to change lanes to overtake or brake to slow down. The behavior decision of real vehicle B will correspondingly affect other real vehicles around it. Finally, the completed trajectories of real vehicles A, B and other vehicles will not overlap or conflict.

[0077] Step 303: Acquire simulation trajectory data of the simulation vehicle during the driving process of the simulation vehicle.

[0078] Exemplarily, since the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameters, during the driving process of the simulated vehicle, the management device can directly obtain the simulated trajectory data of the simulated vehicle, and the simulated trajectory data can include multiple trajectory points. For each trajectory point, the trajectory point can include a time point and a position (such as a latitude and longitude position), indicating that the simulated vehicle is at the position at the time point.

[0079] Since the driving trajectory of the simulated vehicle in the traffic simulation model is consistent with the driving trajectory of the actual vehicle in the real scene, when the driving area of ​​the actual vehicle is divided into the perception area and the detection blind area, the driving area of ​​the simulated vehicle can also be divided into the perception area and the detection blind area. Based on this, when the simulated vehicle is in the perception area, the trajectory points of the simulated vehicle in the perception area can be obtained. When the simulated vehicle is in the detection blind area, the trajectory points of the simulated vehicle in the detection blind area can also be obtained. For example, see Figure 2 As shown, during the driving process of the simulated vehicle, the simulated vehicle passes through area A, area B and area C in sequence, and the management device can obtain the simulated trajectory data of the simulated vehicle and cache the simulated trajectory data of the simulated vehicle.

[0080] For example, area A and area C are perception areas, and area B is a detection blind area. On this basis, when the simulated vehicle is in area A, the trajectory points of the simulated vehicle can be obtained, when the simulated vehicle is in area B, the trajectory points of the simulated vehicle can be obtained, and when the simulated vehicle is in area C, the trajectory points of the simulated vehicle can be obtained. Based on this, the simulated trajectory data for the simulated vehicle may include trajectory points in area A, trajectory points in area B, and trajectory points in area C.

[0081] Step 304: If the actual trajectory data of the actual vehicle does not include the trajectory data between the first moment and the second moment, then after obtaining the simulated trajectory data of the simulated vehicle, the simulated trajectory data between the first moment and the second moment can be selected from the simulated trajectory data, and the selected simulated trajectory data can be fused with the actual trajectory data to obtain fused trajectory data, and the vehicle management of the actual vehicle can be implemented based on the fused trajectory data. Alternatively, after obtaining the simulated trajectory data of the simulated vehicle, the simulated trajectory data can be output, and the vehicle management of the actual vehicle can be implemented based on the simulated trajectory data.

[0082] For example, the actual trajectory data of the actual vehicle may include trajectory points in area A and trajectory points in area C. Assuming that one trajectory point corresponds to every 0.1 seconds, the last trajectory point in area A corresponds to time t1, and the first trajectory point in area C corresponds to time t2, then the first moment may be time t1+0.1 seconds, and the second moment may be time t2-0.1 seconds. In addition, the actual trajectory data of the actual vehicle does not include the trajectory data between the first moment and the second moment (that is, the actual trajectory data does not include the trajectory points in area B).

[0083] For example, simulated trajectory data between the first moment and the second moment can be selected from the simulated trajectory data, and the simulated trajectory data may include trajectory points in area B (i.e., trajectory points generated by the simulated vehicle traveling in the traffic simulation model). In this way, the trajectory points in area B can be fused with the actual trajectory data to obtain fused trajectory data, that is, the fused trajectory data may include trajectory points in area A (located in the actual trajectory data), trajectory points in area B (located in the simulated trajectory data), and trajectory points in area C (located in the actual trajectory data), thereby realizing trajectory completion of the actual trajectory data.

[0084] Trajectory completion refers to the use of observed partial vehicle trajectories to fill in unobserved vehicle trajectories to restore the complete and coherent vehicle motion trajectory of a single vehicle or multiple vehicles. Among them, the trajectory refers to the spatiotemporal motion path generated by a vehicle while driving on the road, usually composed of a series of spatiotemporal coordinate points or trajectory segments, which can describe the vehicle's position, speed, direction and other motion state information.

[0085] Third, the dynamic adjustment process of target motion parameters.

[0086] Exemplarily, when the management device controls the simulated vehicle to travel in the traffic simulation model based on the target motion parameters, the target motion parameters can be dynamically adjusted, and each time the target motion parameters are updated, the updated target motion parameters are used to control the simulated vehicle to travel in the traffic simulation model.

[0087] In one possible implementation, see Figure 4 FIG. 4 is a flow chart of a dynamic adjustment process of target motion parameters. The dynamic adjustment process of target motion parameters may include:

[0088] Step 401, obtaining multiple first update moments, where the first update moment indicates the moment when the target motion parameter needs to be updated, that is, at each first update moment, the target motion parameter needs to be updated. Exemplarily, the first update moment can be understood as the moment when the actual vehicle is in the sensing area.

[0089] Exemplarily, the first update time may be located at the simulation start time t sim +t cache and the target time, and the target time can be the sum of the first time and the delay time. The first time is the time when the actual trajectory data starts to be missing, and the delay time is t cache When obtaining multiple first update times, the first first update time may be the simulation start time t sim +t cache , the last first update time can be the target time, and the interval between two adjacent first update times can be the configured first simulation step, and the first simulation step can be less than the delay time t cache For example, the first simulation step length can be configured based on experience, and there is no restriction on the value of the first simulation step length, which is less than the delay time t cache Then, such as 05. seconds, 1 second, etc., recorded as t step .

[0090] For example, the first update time can be the simulation start time t sim +t cache , the second first update time can be t sim +t cache +t step , the third first update time can be t sim +t cache +t step +t step , and so on, the last first update time can be the target time.

[0091] Step 402: at each first update moment (i.e., the current moment reaches the first update moment), predict the predicted trajectory data of the simulated vehicle in the prediction period, the predicted trajectory data is the trajectory data of the simulated vehicle when traveling in the traffic simulation model based on the target motion parameters, the end time of the prediction period is determined based on the first update moment, and the start time of the prediction period is determined based on the first update moment and the delay duration.

[0092] Exemplarily, for the start time of the prediction period, the difference between the first update time and the delay duration can be determined as the start time. For the end time of the prediction period, it can be determined whether the first update time is greater than the first time; if not, the first update time is determined as the end time; if so, the first time can be determined as the end time.

[0093] For example, for the first update time t sim +t cache , the end time of the prediction period is the first update time t sim +t cache , the starting time of the forecast period is t sim (First update time and delay time t cach The first update time t is reached at the current time. sim +t cache The actual trajectory data of the actual vehicle includes the data from t sim to sim +t cache Since the simulation starts at time t sim +t cache At the beginning, the simulated vehicle is controlled to travel in the traffic simulation model, so that at the current time, the first update time t sim +t cache When the simulated vehicle starts to drive in the traffic simulation model, it is assumed that the actual vehicle is at t sim At position P1, the simulated vehicle is controlled to travel from position P1 in the traffic simulation model.

[0094] At the current time, the first update time t is reached sim +t cache When the simulated vehicle is predicted to start from the starting time t sim Until the deadline t sim +t cache Obviously, since the simulated vehicle has just started to travel from position P1, that is, it only has the starting time t sim The simulation trajectory data without the starting time t sim Until the deadline t sim +t cacheTherefore, it is necessary to predict the simulated vehicle from the starting time t in advance. sim Until the deadline t sim +t cache The predicted trajectory data is pre-calculated from the starting time t sim Until the deadline t sim +t cache movement trajectory.

[0095] In the prediction simulation, the vehicle starts from the starting time t sim Until the deadline t sim +t cache When obtaining predicted trajectory data, the simulated vehicle can be simulated to travel in the traffic simulation model based on the target motion parameters (that is, not actually traveling in the traffic simulation model, but simulating and calculating the trajectory generated during the driving process), thereby obtaining the trajectory data of the simulated vehicle when traveling in the traffic simulation model, and using the trajectory data as predicted trajectory data.

[0096] It should be noted that since the target motion parameters are dynamically updated, the predicted trajectory data may not be consistent with the simulated trajectory data when the simulated vehicle is traveling in the traffic simulation model.

[0097] For example, for the first update time t sim +t cache +t step , the end time of the prediction period is the first update time t sim +t cache +t step , the starting time of the forecast period is t sim +t step At the current time, the first update time t is reached sim +t cache +t step The actual trajectory data includes the sim +t cache to sim +t cache +t step The trajectory data of . At the current time, the first update time t sim +t cache +t step When the simulation trajectory data of the simulated vehicle includes the data from t sim to sim +t step The trajectory data.

[0098] At the current time, the first update time t is reached sim +t cache +t step When the simulated vehicle is predicted to start from the starting time t sim +tstep Until the deadline t sim +t cache +t step The predicted trajectory data.

[0099] Similarly, for the first update time t3 (i.e., any first update time) between the first time t1 and the target time t2 (i.e., the sum of the first time and the delay time), that is, the first update time t3 is greater than the first time t1, then the difference between the first update time and the delay time is determined as the starting time, that is, t3 and the delay time t cache The difference between t1 and t2 is taken as the starting time. The first time t1 is determined as the end time.

[0100] Step 403: determine a first motion parameter based on the predicted trajectory data of the prediction period and the actual trajectory data of the prediction period, and update the first motion parameter to a target motion parameter of the simulated vehicle.

[0101] For example, at the current time, the first update time t sim +t cache The actual trajectory data includes the sim to sim +t cache The trajectory data of t sim to sim +t cache The trajectory data can therefore be used based on the prediction period (from t sim to sim +t cache ) and the actual trajectory data of the prediction period determine the first motion parameter. sim +t cache +t step The actual trajectory data includes the sim +t cache to sim +t cache +t ste The trajectory data of t sim +t cache to sim +t cache +t step The trajectory data can therefore be used based on the prediction period (from t sim +t cache to sim +t cache +t step ) and the actual trajectory data of the prediction period determine the first motion parameter, and so on.

[0102] Exemplarily, if the predicted trajectory data includes the predicted positions of each time point within the prediction period (the trajectory points in the predicted trajectory data are referred to as predicted positions, such as longitude and latitude positions), and the actual trajectory data includes the actual positions of each time point within the prediction period (the trajectory points in the actual trajectory data are referred to as actual positions, such as longitude and latitude positions), then for each time point, the distance between the predicted position of the time point and the actual position of the time point is determined (such as Euclidean distance, etc.), and there is no restriction on the method of determining this distance.

[0103] Based on the distance corresponding to each time point, the distance error between the predicted trajectory data and the actual trajectory data is calculated. For example, based on the distance corresponding to each time point, the average value of the distances corresponding to these time points can be calculated, and then the average value can be used as the distance error.

[0104] If the distance error is greater than a preset error threshold (which can be configured based on experience), the target motion parameters currently used by the simulated vehicle (such as driving speed, distance to the preceding vehicle, probability of lane change, and other parameters) can be adjusted, and the adjusted motion parameters can be determined as the first motion parameters of the simulated vehicle. For example, the driving speed currently used by the simulated vehicle can be reduced or increased, the distance to the preceding vehicle currently used by the simulated vehicle can be reduced or increased, the probability of lane change currently used by the simulated vehicle can be reduced or increased, and of course, the above are only examples of adjusting the target motion parameters, and there is no limitation on this adjustment method.

[0105] For example, the distance error can be input to a gradient descent algorithm, and the gradient descent algorithm adjusts the target motion parameters currently used by the simulated vehicle based on the distance error, and the adjustment goal is to make the distance error smaller and smaller, that is, the distance error is not greater than a preset error threshold. Alternatively, the distance error can be input to a heuristic algorithm, and the heuristic algorithm adjusts the target motion parameters currently used by the simulated vehicle based on the distance error, and the adjustment goal is to make the distance error smaller and smaller, that is, the distance error is not greater than a preset error threshold. Of course, the above are just two examples of adjustment algorithms, and there is no limitation to this.

[0106] If the distance error is not greater than a preset error threshold, the target motion parameter currently used by the simulated vehicle may be determined as the first motion parameter of the simulated vehicle, that is, the target motion parameter is kept unchanged.

[0107] In summary, it can be seen that for each first update moment, when the current moment reaches the first update moment, the first motion parameter can be obtained, and the first motion parameter is updated to the target motion parameter of the simulated vehicle. In this way, after the first update moment, when the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameter, it is driven based on the target motion parameter of the first update moment.

[0108] In one possible implementation, see Figure 5 FIG. 4 is a flow chart of a dynamic adjustment process of target motion parameters. The dynamic adjustment process of target motion parameters may include:

[0109] Step 501, obtaining multiple second update moments, the second update moment represents the moment when the target motion parameter needs to be updated, that is, at each second update moment, the target motion parameter needs to be updated. Exemplarily, the second update moment can be understood as the moment when the actual vehicle is in the detection blind spot.

[0110] Exemplarily, the plurality of second update times may be located between the target time and the second time, and the target time may be the time between the first time and the delay time t cache The first moment is the moment when the actual trajectory data starts to be missing, and the second moment is the moment when the actual trajectory data stops being missing. For example, when obtaining multiple second update moments, the first second update moment can be the target moment, the last second update moment can be the second moment, and the interval between two adjacent second update moments can be the configured second simulation step, and the second simulation step can be equal to the first simulation step t step , can also be greater or less than the first simulation step t step For example, the second simulation step length can be configured based on experience, and there is no restriction on the value of the second simulation step length, such as 0.5 seconds, 1 second, etc.

[0111] Step 502: At each second update time (i.e., the current time reaches the second update time), obtain the first average speed corresponding to the target section. If the actual trajectory data of the actual vehicle does not include the trajectory data between the first time and the second time, then the target section may be the section between the position of the actual vehicle at the first time and the position of the actual vehicle at the second time. In addition, the first average speed may represent the average speed of multiple actual vehicles when passing through the target section.

[0112] For example, the first average speed of all historical vehicles (i.e., actual vehicles) passing through the target road section within a historical time window can be calculated. For example, the first average speed can be determined by the following formula (1). Of course, formula (1) is only an example and there is no limitation on this determination method.

[0113]

[0114] In formula (1), Represents the first average speed. n represents the total number of historical vehicles that passed the target road section within a historical time window. represents the time when the i-th historical vehicle enters the target section, It indicates the position where the i-th historical vehicle enters the target section. The position can be the starting position of the target section or other positions of the target section, and there is no restriction on this.

[0115] In formula (1), represents the time when the i-th historical vehicle leaves the target section, and represents the vehicles in a historical time window, t window is the historical time window size, It indicates the position where the i-th historical vehicle leaves the target section. The position can be the end position of the target section or other positions of the target section. There is no limitation on this.

[0116] Step 503: At each second update time (i.e., the current time reaches the second update time), obtain the second average speed corresponding to the target section. The target section may be the section between the position of the actual vehicle at the first time and the position of the actual vehicle at the second time, and the second average speed may represent the average speed of all simulated vehicles at the target section at the second update time.

[0117] For example, simulated vehicles can be established for multiple actual vehicles in the traffic simulation model, and these simulated vehicles can be controlled to travel in the traffic simulation model. Based on this, the simulated vehicles at the target section at the current moment (i.e., the second update moment) can be determined, and the speed average of all simulated vehicles at the target section at the current moment, i.e., the second average speed, can be determined. For example, the second average speed can be determined by the following formula (2). Of course, formula (2) is only an example, and there is no limitation on this determination method.

[0118]

[0119] In formula (2), represents the second average speed. m represents the total number of all simulated vehicles on the target road section at the current moment. Represents the driving speed of the i-th simulated vehicle at the current moment.

[0120] Step 504: determine a second motion parameter based on the first average speed corresponding to the target road section and the second average speed corresponding to the target road section, and update the second motion parameter to the target motion parameter of the simulated vehicle.

[0121] Exemplarily, if the target motion parameters currently used by the simulated vehicle include driving speed, the driving speed in the target motion parameters can be adjusted based on the first average speed and the second average speed (the remaining parameters in the target motion parameters remain unchanged, such as the distance to the vehicle in front and the probability of lane change, and the like), and the adjusted motion parameters are determined as the second motion parameters of the simulated vehicle.

[0122] Exemplarily, the following formula (3) may be used to adjust the vehicle speed in the target motion parameter. Of course, formula (3) is only an example and there is no limitation on this adjustment method.

[0123]

[0124] In formula (3), It can represent the adjusted driving speed (the adjusted motion parameters include the adjusted driving speed), It can represent the driving speed in the target motion parameters (i.e. the parameters before adjustment), can represent the second average speed, The first average speed can be represented.

[0125] In summary, it can be seen that for each second update moment, when the current moment reaches the second update moment, the second motion parameter can be obtained, and the second motion parameter is updated to the target motion parameter of the simulated vehicle. In this way, after the second update moment, when the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameter, it is driven based on the target motion parameter of the second update moment.

[0126] In a possible implementation, when the second moment is reached, the actual trajectory data begins to include the trajectory points at the second moment, that is, the actual vehicle re-perceives the area and re-acquires multiple first update moments, the first first update moment can be the second moment and the delay time t cache Repeat the above steps.

[0127] After the actual vehicle leaves the detection blind area and reenters the perception area, the actual trajectory data of the actual vehicle reappears the trajectory based on Figure 4 The target motion parameter dynamic adjustment method is used to reduce the error between the predicted trajectory data and the actual trajectory data of the corresponding simulated vehicle, thereby achieving a smooth continuation of the simulated trajectory data.

[0128] Exemplarily, the processing of the perception area after the detection blind spot is the same as the processing of the perception area before the detection blind spot, but the parameter adjustment of the perception area before the detection blind spot is for more accurate prediction within the detection blind spot, while the parameter adjustment of the perception area after the detection blind spot is for the connection between the prediction and the actual trajectory, so as to achieve a smooth continuation of the simulated trajectory data.

[0129] In summary, when the actual vehicle is in the perception area before the detection blind spot, the target motion parameters of the corresponding simulated vehicle are dynamically adjusted based on the actual trajectory data. When the actual vehicle is in the detection blind spot, its motion trajectory is simulated and completed based on the target motion parameters obtained by dynamic parameter adjustment and the expected driving speed (i.e., the second motion parameter) considering the historical average and individual differences. When the actual vehicle is in the perception area after the detection blind spot, the simulated trajectory data and the actual trajectory data are smoothly connected.

[0130] In one possible implementation, see Fig. 6A The figure shows the overall logic diagram of trajectory completion.

[0131] First, a road network model is built for the road section containing the detection blind spot (excluding the entrance and exit), and the traffic simulation model is run. The simulation time t is set. sim is the current actual time t real Then, the data in the sensing area t is collected and cached. sim to sim +t cache Real-time vehicle trajectory data for a period of time, that is, actual trajectory data.

[0132] Then, based on the cached data and taking into account the actual vehicle position, simulated vehicle loading, dynamic parameter adjustment, target speed setting, etc. are performed. sim The actual vehicle newly sensed at time t is loaded into the traffic simulation model. sim to sim +t cache For actual vehicles that were in the sensing area before and after the detection blind spot during the period, the target motion parameters of the simulated vehicle corresponding to the actual vehicle are dynamically adjusted. sim to sim +t cache For the actual vehicles that are in the detection blind zone during the time period, a target speed based on historical average and individual differences is set for the simulated vehicles corresponding to the actual vehicles.

[0133] Then, simulate and output the time of all simulated vehicles on the road section at t sim The position, speed and other information at the moment. Update the simulation time t sim t sim +t step, and start the next simulation step cycle, t step is the first simulation step.

[0134] It should be noted that the above steps are performed at the actual time t real =t sim +t cache The simulation will be executed only after the data collection and caching are completed. Therefore, the time t sim Compared to actual time t real Existence cache of delay.

[0135] In one possible implementation, see Figure 6B As shown, it is a brief flowchart of trajectory completion.

[0136] During the simulation stepping process: when the actual vehicle is in the perception area before the blind spot, the individual model parameters (i.e., target motion parameters) of the simulated vehicle are dynamically adjusted based on the cached data. When the actual vehicle is in the detection blind spot, the simulation is completed based on the driving behavior model and the expected driving speed. When the actual vehicle is in the perception area after the blind spot, the simulation trajectory and the real trajectory are smoothly connected.

[0137] It can be seen from the above technical solutions that in the embodiments of the present application, the driving behavior model parameters (i.e., target motion parameters) of the simulated vehicle can be dynamically adjusted to obtain accurate and reliable target motion parameters, thereby improving the accuracy of trajectory completion and achieving a smooth connection between the simulated trajectory and the real trajectory. Based on the high operating efficiency and data caching mechanism of the traffic simulation model, the real-time output of the result is improved, and the running trajectory of the actual vehicle involved in the detection blind spot can be simulated and completed with a delay of seconds. By using driving behavior model parameters with a higher degree of abstraction as the behavioral characteristics of the actual vehicle, considering the interactive motion between multiple actual vehicles, making full use of the trajectory data in the perception area before and after the detection blind spot, and considering the behavioral differences between individual actual vehicles, the accuracy of the trajectory completion results of the detection blind spot can be improved.

[0138] A dynamic calibration method for individual driving behavior model parameters based on vehicle trajectory data caching and acceptable error threshold is proposed. It is more flexible and can dynamically adjust the driving behavior model parameters of the simulated vehicle to improve the trajectory completion accuracy while achieving a smooth connection between the simulated trajectory and the real trajectory.

[0139] Driving behavior model parameters are abstracted from the vehicle's trajectory data as its behavior characteristics, and the interaction between vehicles is considered through the traffic simulation model to predict and complete the trajectory in the blind spot in real time. The driving behavior model parameters can reflect the vehicle's more advanced behavior characteristics (rather than simple motion states such as speed and vehicle head direction), and can restore and simulate its motion trajectory under other different conditions, thereby improving the accuracy of trajectory completion. At the same time, all vehicles on the road section are simulated, and their motion behaviors are constrained through the interaction between vehicles to ensure the reasonable accuracy of trajectory completion. Based on the high operating efficiency and data caching mechanism of traffic simulation, the real-time performance of trajectory completion can be improved. Considering the average passing speed of the detection blind spot over a period of time and the difference between individual vehicles, different expected driving speeds are set for vehicles in the detection blind spot, so as to make fuller use of the trajectory data in the sensing area before and after the detection blind spot, and considering the dynamic changes of traffic conditions in the detection blind spot, the accuracy of trajectory completion can be improved.

[0140] Based on the same application concept as the above method, a simulation trajectory data acquisition device is proposed in the embodiment of the present application, see Figure 7 FIG. 1 is a schematic diagram of the structure of the device, and the device may include:

[0141] The actual data processing module 71 is used to obtain the actual track data of the actual vehicle during the driving process of the actual vehicle, and store the actual track data of the actual vehicle;

[0142] The simulation data processing module 72 is used to establish a simulation vehicle corresponding to the actual vehicle at the configured data collection time, and use fixed motion parameters as target motion parameters; determine the simulation start time based on the data collection time and the acquired delay time, and control the simulation vehicle to travel in the traffic simulation model based on the target motion parameters starting from the simulation start time; wherein the simulation position of the simulation vehicle at the simulation start time corresponds to the actual position of the actual vehicle at the data collection time; during the driving process of the simulation vehicle, obtain the simulation trajectory data of the simulation vehicle, and store the simulation trajectory data of the simulation vehicle;

[0143] The parameter data processing module 73 is used to obtain multiple first update moments when the actual vehicle is in the perception area in front of the detection blind spot; at each first update moment, predict the predicted trajectory data of the simulated vehicle in the prediction period, the predicted trajectory data is the trajectory data of the simulated vehicle when it is traveling in the traffic simulation model based on the target motion parameters, the end time of the prediction period is determined based on the first update moment, and the start time of the prediction period is determined based on the first update moment and the delay duration; determine the first motion parameter based on the predicted trajectory data of the prediction period and the actual trajectory data of the prediction period, and update the first motion parameter to the target motion parameter of the simulated vehicle.

[0144] Exemplarily, if the actual trajectory data does not include trajectory data between the first moment and the second moment, the parameter data processing module 73 is also used to obtain multiple second update moments when the actual vehicle is in the detection blind spot; at each second update moment, a second motion parameter is determined based on the first average speed and the second average speed corresponding to the target section, and the second motion parameter is updated to the target motion parameter of the simulated vehicle; wherein the target section is the section between the position of the actual vehicle at the first moment and the position of the actual vehicle at the second moment, the first average speed represents the average speed of multiple actual vehicles when passing through the target section, and the second average speed represents the average speed of all simulated vehicles in the target section at the second update moment; wherein the multiple first update moments are located between the simulation start moment and the target moment; the multiple second update moments are located between the target moment and the second moment; the target moment is the sum of the first moment and the delay duration.

[0145] Exemplarily, when obtaining multiple first update moments, the first first update moment is the simulation start moment, the interval between two adjacent first update moments is the configured first simulation step, and the first simulation step is smaller than the delay duration; when obtaining multiple second update moments, the first second update moment is the target moment, the interval between two adjacent second update moments is the configured second simulation step; wherein, when determining the start moment of the prediction period based on the first update moment and the delay duration, the difference between the first update moment and the delay duration is determined as the start moment.

[0146] When determining the end time of the prediction time period based on the first update time, determine whether the first update time is greater than the first time; if not, determine the first update time as the end time; if so, determine the first time as the end time; wherein the delay duration is determined based on the length of the road section in the detection blind spot, and the longer the road section in the detection blind spot is, the longer the delay duration is.

[0147] Exemplarily, the parameter data processing module 73 is specifically used to determine the first motion parameter based on the predicted trajectory data of the prediction period and the actual trajectory data of the prediction period: if the actual trajectory data includes the actual positions of multiple time points within the prediction period, and the predicted trajectory data includes the predicted positions of multiple time points within the prediction period, then determine the distance between the predicted position of each time point and the actual position of the time point; based on the distance corresponding to each time point, calculate the distance error between the predicted trajectory data and the actual trajectory data; if the distance error is greater than a preset error threshold, adjust the target motion parameter currently used by the simulated vehicle, and determine the adjusted motion parameter as the first motion parameter of the simulated vehicle; or, if the distance error is not greater than the preset error threshold, determine the target motion parameter currently used by the simulated vehicle as the first motion parameter of the simulated vehicle.

[0148] Exemplarily, when the parameter data processing module 73 determines the second motion parameter based on the first average speed and the second average speed corresponding to the target road section, it is specifically used to: if the target motion parameter currently used by the simulated vehicle includes the driving speed, then the driving speed in the target motion parameter is adjusted based on the first average speed and the second average speed, and the adjusted motion parameter is determined as the second motion parameter of the simulated vehicle; wherein, the parameter data processing module 73 uses the following formula to adjust the driving speed in the target motion parameter: in, Indicates the adjusted driving speed. represents the driving speed in the target motion parameter, represents the second average speed, represents the first average speed.

[0149] Exemplarily, when the simulation data processing module 72 controls the simulated vehicle to travel in the traffic simulation model based on the target motion parameters, it is specifically used to: during the driving process of the actual vehicle, establish a simulated vehicle for the associated vehicles around the actual vehicle, and control the simulated vehicle of the associated vehicles to travel in the traffic simulation model; based on the simulation trajectory data of the simulated vehicle of the associated vehicle in the traffic simulation model, control the simulated vehicle of the actual vehicle to travel in the traffic simulation model based on the target motion parameters.

[0150] Exemplarily, the simulation data processing module 72 is also used to select the simulation trajectory data between the first moment and the second moment from the simulation trajectory data after acquiring the simulation trajectory data, and fuse the selected simulation trajectory data with the actual trajectory data to obtain fused trajectory data, and implement vehicle management for the actual vehicle based on the fused trajectory data; or, implement vehicle management for the actual vehicle based on the simulation trajectory data.

[0151] Based on the same application concept as the above method, an electronic device (such as a management device) is proposed in the embodiment of the present application, see Figure 8 As shown, it includes: a processor 81 and a machine-readable storage medium 82, the machine-readable storage medium 82 stores machine-executable instructions that can be executed by the processor 81; the processor 81 is used to execute the machine-executable instructions to implement the simulation trajectory data acquisition method disclosed in the above example of this application.

[0152] Based on the same application concept as the above method, an embodiment of the present application also provides a machine-readable storage medium, on which a number of computer instructions are stored. When the computer instructions are executed by a processor, the simulation trajectory data acquisition method disclosed in the above example of the present application can be implemented.

[0153] The above-mentioned machine-readable storage medium may be any electronic, magnetic, optical or other physical storage device, which may contain or store information, such as executable instructions, data, etc. For example, the machine-readable storage medium may be: RAM (Radom Access Memory), volatile memory, non-volatile memory, flash memory, storage drive (such as hard disk drive), solid state drive, any type of storage disk (such as CD, DVD, etc.), or similar storage medium, or a combination thereof.

[0154] Based on the same application concept as the above method, an embodiment of the present application also provides a computer program product, which may include a computer program. When the computer program is executed by a processor, it implements the simulation trajectory data acquisition method disclosed in the above example of the present application.

[0155] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0156] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for acquiring simulation trajectory data, characterized in that: The method comprises: During the driving process of the actual vehicle, obtaining actual trajectory data of the actual vehicle; At the configured data collection time, a simulated vehicle corresponding to the actual vehicle is established, and fixed motion parameters are used as target motion parameters; a simulation start time is determined based on the data collection time and the acquired delay duration, and starting from the simulation start time, the simulated vehicle is controlled to travel in the traffic simulation model based on the target motion parameters; wherein the simulated position of the simulated vehicle at the simulation start time corresponds to the actual position of the actual vehicle at the data collection time; During the driving process of the simulated vehicle, obtaining simulated trajectory data of the simulated vehicle; When the actual vehicle is in the sensing area in front of the detection blind area, obtaining a plurality of first update moments; At each first update time, predict the predicted trajectory data of the simulated vehicle in the prediction period, the predicted trajectory data is the trajectory data of the simulated vehicle when it travels in the traffic simulation model based on the target motion parameters, the end time of the prediction period is determined based on the first update time, and the start time of the prediction period is determined based on the first update time and the delay time; A first motion parameter is determined based on the predicted trajectory data of the predicted period and the actual trajectory data of the predicted period, and the first motion parameter is updated to a target motion parameter of the simulated vehicle.

2. The method according to claim 1, characterized in that If the actual trajectory data does not include trajectory data between the first moment and the second moment, the method further includes: When the actual vehicle is in the detection blind area, a plurality of second update moments are obtained; at each second update moment, a second motion parameter is determined based on a first average speed and a second average speed corresponding to a target section, and the second motion parameter is updated to a target motion parameter of the simulated vehicle; wherein the target section is a section between a position of the actual vehicle at a first moment and a position of the actual vehicle at a second moment, the first average speed represents an average speed of a plurality of actual vehicles when passing through the target section, and the second average speed represents an average speed of all simulated vehicles in the target section at the second update moment; Wherein, the plurality of first update times are located between the simulation start time and the target time; The plurality of second update time points are located between the target time point and the second time point; The target time is the sum of the first time and the delay duration.

3. The method according to claim 2, characterized in that When obtaining multiple first update moments, the first first update moment is the simulation start moment, the interval between two adjacent first update moments is the configured first simulation step length, and the first simulation step length is less than the delay duration; When obtaining multiple second update moments, the first second update moment is the target moment, and the interval between two adjacent second update moments is the configured second simulation step length; Wherein, when determining the starting time of the prediction period based on the first update time and the delay time, the difference between the first update time and the delay time is determined as the starting time; When determining the end time of the prediction period based on the first update time, determining whether the first update time is greater than the first time; if not, determining the first update time as the end time; if so, determining the first time as the end time; The delay time is determined based on the length of the road section of the detection blind spot. The longer the length of the road section of the detection blind spot is, the longer the delay time is.

4. The method according to claim 1, characterized in that: The determining the first motion parameter based on the predicted trajectory data of the predicted period and the actual trajectory data of the predicted period comprises: If the actual trajectory data includes actual positions of multiple time points within the prediction period, and the predicted trajectory data includes predicted positions of multiple time points within the prediction period, determining the distance between the predicted position at each time point and the actual position at the time point; and calculating the distance error between the predicted trajectory data and the actual trajectory data based on the distance corresponding to each time point; If the distance error is greater than a preset error threshold, adjusting the target motion parameter currently used by the simulated vehicle, and determining the adjusted motion parameter as the first motion parameter of the simulated vehicle; Alternatively, if the distance error is not greater than the preset error threshold, the target motion parameter currently used by the simulated vehicle is determined as the first motion parameter of the simulated vehicle.

5. The method according to claim 2, characterized in that: The determining of the second motion parameter based on the first average speed and the second average speed corresponding to the target road section includes: If the target motion parameter currently used by the simulated vehicle includes a driving speed, adjusting the driving speed in the target motion parameter based on the first average speed and the second average speed, and determining the adjusted motion parameter as a second motion parameter of the simulated vehicle; The following formula is used to adjust the driving speed in the target motion parameter: in, Indicates the adjusted driving speed. represents the driving speed in the target motion parameter, represents the second average speed, represents the first average speed.

6. The method according to any one of claims 1 to 5, characterized in that: The controlling the simulated vehicle to travel in the traffic simulation model based on the target motion parameter comprises: During the driving process of the actual vehicle, a simulated vehicle is established for associated vehicles around the actual vehicle, and the simulated vehicle of the associated vehicle is controlled to drive in a traffic simulation model; Based on the simulation trajectory data of the simulation vehicle of the associated vehicle in the traffic simulation model, the simulation vehicle of the actual vehicle is controlled to travel in the traffic simulation model based on the target motion parameters.

7. The method according to claim 2, characterized in that After obtaining the simulated trajectory data of the simulated vehicle, the method further includes: Selecting simulated trajectory data between the first moment and the second moment from the simulated trajectory data, fusing the selected simulated trajectory data with the actual trajectory data to obtain fused trajectory data, and implementing vehicle management for the actual vehicle based on the fused trajectory data; Alternatively, vehicle management is implemented for the actual vehicle based on the simulation trajectory data.

8. A simulation trajectory data acquisition device, characterized in that: The device comprises: An actual data processing module, used for acquiring actual track data of the actual vehicle during the driving process of the actual vehicle, and storing the actual track data of the actual vehicle; A simulation data processing module is used to establish a simulated vehicle corresponding to the actual vehicle at a configured data collection time, and use fixed motion parameters as target motion parameters; determine a simulation start time based on the data collection time and the acquired delay time, and control the simulated vehicle to travel in a traffic simulation model based on the target motion parameters starting from the simulation start time; wherein the simulated position of the simulated vehicle at the simulation start time corresponds to the actual position of the actual vehicle at the data collection time; during the travel of the simulated vehicle, obtain simulated trajectory data of the simulated vehicle, and store the simulated trajectory data of the simulated vehicle; A parameter data processing module is used to obtain multiple first update moments when the actual vehicle is in the perception area in front of the detection blind spot; at each first update moment, predict the predicted trajectory data of the simulated vehicle in the prediction period, the predicted trajectory data is the trajectory data of the simulated vehicle when it is traveling in the traffic simulation model based on the target motion parameters, the end time of the prediction period is determined based on the first update moment, and the start time of the prediction period is determined based on the first update moment and the delay duration; determine the first motion parameter based on the predicted trajectory data of the prediction period and the actual trajectory data of the prediction period, and update the first motion parameter to the target motion parameter of the simulated vehicle.

9. A machine-readable storage medium storing machine-executable instructions that can be executed by a processor; wherein: The processor is configured to execute the machine executable instructions to implement the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor; The processor is used to execute machine executable instructions to implement the method described in any one of claims 1-7.

Images