Vehicle speed planning method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a vehicle speed planning method and device, electronic equipment and a storage medium. The method comprises the steps that a vehicle planning decision label corresponding to a target vehicle is determined according to collected surrounding information; determining target ST reference lines of target vehicles corresponding to different vehicle planning decision labels; smoothing the target ST reference line to generate a smooth ST reference line; and deriving the smooth ST reference line, and outputting a target speed planning curve. The method comprises the following steps: determining a planning decision label of a target vehicle by collecting surrounding information, generating a target ST reference line based on different labels, and finally outputting an optimal speed and acceleration curve through smooth processing and quadratic programming algorithm optimization. The method not only improves the decision precision and response speed of the intelligent driving system, but also enhances the stability and safety of the system, so that the vehicle can perform speed control more intelligently and efficiently in different driving scenes.

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and in particular to a vehicle speed planning method, device, electronic equipment and storage medium. Background Art

[0002] As one of the representatives of the new four modernizations of automobiles, the development momentum of technology is getting faster and faster. Among them, the autonomous driving technology of automobiles is the main battlefield of intelligent development. From the perspective of technology flow, autonomous driving technology can be divided into six modules: perception, positioning, prediction, decision-making, planning and control. Among them, the planning module plays a key role in connecting the upper and lower parts. Vehicles are divided into path planning and speed planning according to longitudinal and lateral movement. The purpose of speed planning is to plan a smooth speed sequence for the vehicle that meets the vehicle control requirements. It complements path planning and plays a key role in the lateral and longitudinal control of the rear-end vehicle. The mutual support of the two technologies can greatly improve the intelligence level of the vehicle and accelerate the commercialization process.

[0003] However, the speed planning in the related technology is limited to the narrow ST map assisted speed planning, which cannot take into account the complex road conditions in the real vehicle intelligent driving scenario, so the single vehicle speed planning cannot be applied to all scenarios. Summary of the invention

[0004] The purpose of the present invention is to provide a vehicle speed planning method to solve the problem of how to control the motor to improve the vehicle stability in the prior art.

[0005] In order to achieve the above object, the present invention discloses a vehicle speed planning method, the method comprising: Determine the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information; Determine the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; Smoothing the target ST reference line to generate a smoothed ST reference line; The smooth ST reference line is derived to output a target speed planning curve.

[0006] Optionally, determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information includes: When there is a first vehicle in front of the target vehicle in the driving lane, and the target vehicle maintains a preset following distance with the second vehicle, the vehicle planning decision label corresponding to the target vehicle is a following vehicle label; When the preset following distance corresponding to the target vehicle is changed, the vehicle planning decision tag corresponding to the target vehicle is a keep distance tag; When there is a third vehicle that changes lanes and cuts into the lane where the target vehicle is traveling within a preset range of the lane where the target vehicle is traveling, the vehicle planning decision label corresponding to the target vehicle is a cut-in label.

[0007] Optionally, determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision label is any one of the following label, the keeping distance label and the cutting-in label, obtaining a standard following distance; Determine a target following distance according to a pre-acquired first current scene of the target vehicle and the standard following distance, and construct a following distance line in a pre-set ST diagram for the target following distance; Determine a second current scene of the target vehicle according to the target following distance, and construct a transition reference line corresponding to the target vehicle in a preset ST diagram; A target ST reference line is constructed in a preset ST diagram according to the following distance line and the transition reference line.

[0008] Optionally, determining a target following distance according to the pre-acquired first current scene of the target vehicle and the standard following distance, and constructing a following distance line in a pre-set ST diagram for the target following distance comprises: When the current scenario is that the target vehicle is in a lane-changing state, obtaining a collision time between the target vehicle and a rear vehicle in a lane adjacent to the lane in which the target vehicle is traveling, and determining a rear vehicle safety level according to the collision time; If the safety level of the rear vehicle is in a dangerous state, calibrate a first bias coefficient according to a speed difference between the target vehicle and a front vehicle in a lane where the target vehicle is traveling; If the rear vehicle safety level is a safe state, calibrating the first bias coefficient according to the time headway intrusion level between the target vehicle and the front vehicle in the lane where the target vehicle is traveling; When the current scene is that the target vehicle is in a lane change end transition state, preliminarily calibrate the transition state time according to the intrusion time distance, obtain a first time offset coefficient according to the speed difference, a second time offset coefficient according to the intrusion rear vehicle time distance level, and obtain a time distance coefficient increment; generating the first bias coefficient according to the first time bias coefficient, the second time bias coefficient and a time interval coefficient increment; When the current scenario is that the target vehicle is in a non-lane-changing state, setting the first bias coefficient to a target value; adjusting the standard following distance according to the first bias coefficient to obtain a target following distance; A following distance line is constructed in a preset ST map for the target following distance.

[0009] Optionally, determining the rear vehicle safety level according to the collision time includes: When it is detected that the collision time is less than a first preset value, the safety level of the rear vehicle is in a dangerous state; When it is detected that the collision time is a negative number or infinite, an intrusion time level between the target vehicle and the rear vehicle is obtained, and when the intrusion time level is less than a second preset value, it is determined that the safety level of the rear vehicle is in a dangerous state.

[0010] Optionally, obtaining the intrusion time headway level between the target vehicle and the rear vehicle includes: The intrusion time-to-distance level is generated according to a pre-calibrated target following distance and a ratio of the distance between the target vehicle and a front vehicle or a rear vehicle in the lane.

[0011] Optionally, determining the second current scene of the target vehicle according to the target following distance, and constructing a transition reference line corresponding to the target vehicle in a preset ST diagram includes: When the second current scene is long-distance non-intrusion, a calibration lookup table is set according to a pre-acquired collision time between the target vehicle and the front vehicle, and a speed bias coefficient lookup table is set; Obtaining a velocity bias coefficient by linear interpolation according to the collision time; generating a reference speed of the target vehicle according to the speed bias coefficient, the current speed of the target vehicle and the current speed of the vehicle ahead; Constructing a first transition reference line corresponding to the target vehicle in a preset ST diagram according to the reference vehicle speed; When the second current scene is a close-range intrusion, obtaining a safety acceleration; Acquiring an acceleration bias according to the safety acceleration; generating a smooth acceleration according to the safety acceleration and the acceleration offset; A second transition reference line corresponding to the target vehicle is constructed in a preset ST diagram according to the smoothed acceleration.

[0012] Optionally, determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information includes: When the collected surrounding information shows that there is an obstacle in front of the target vehicle's driving lane, and the lanes on both sides corresponding to the target vehicle's driving lane are in an obstacle-free state, the vehicle planning decision label corresponding to the target vehicle is a deceleration passing label.

[0013] Optionally, the deceleration passing tag includes a first deceleration passing tag, and determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision tags includes: When the vehicle planning decision tag is the first deceleration passing tag, obtaining the deceleration passing time corresponding to the target vehicle; Obtaining a lower boundary distance of a lower boundary of the deceleration time map in a preset ST diagram; Obtaining a ratio of the lower boundary distance to the deceleration passing time to obtain a first constraint curve, wherein the first constraint curve includes a speed; The minimum value of the speed of the constraint curve and the pre-acquired speed limit of the target vehicle driving lane is taken to obtain a target ST reference line corresponding to the target vehicle.

[0014] Optionally, the deceleration passing tag includes a second deceleration passing tag, and determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision tags includes: When the vehicle planning decision label is the second deceleration passing label, obtaining the deceleration passing time corresponding to the target vehicle; Obtaining a lower boundary distance of a lower boundary of the deceleration time mapping in a preset coordinate system; Obtaining a ratio of the lower boundary distance to the deceleration passing time to obtain a first constraint curve, wherein the first constraint curve includes a speed; Taking the minimum value of the speed of the constraint curve and the pre-acquired speed limit of the target vehicle driving lane to obtain a second constraint curve; The intersection point of the pre-acquired following distance line and the second constraint curve is obtained to obtain a target ST reference line corresponding to the target vehicle.

[0015] Optionally, determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information includes: When the collected surrounding information indicates that there is a merging vehicle within a preset range of the target vehicle's driving lane, the vehicle planning decision label corresponding to the target vehicle is a merging label, and the merging label includes a merging deceleration label and a merging acceleration label.

[0016] Optionally, determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision tag is the merging deceleration tag, obtaining a predicted speed of the merging vehicle; Searching a pre-acquired safety distance calibration table for a target safety distance between the target vehicle and the merging vehicle corresponding to the predicted speed; The maximum value of the speed corresponding to the target vehicle and the speed corresponding to the merging vehicle is taken to obtain the target ST reference line corresponding to the target vehicle.

[0017] Optionally, before the step of determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information, the method includes: Draw a rectangular frame of surrounding vehicles according to the predicted trajectory information of the surrounding vehicles of the target vehicle; Obtaining a set of projection distances between each preset point position of the target vehicle and a pre-acquired planned path of the target vehicle; Taking the minimum value in the projection distance set as the target projection distance; If the target projection distance is less than a preset threshold, the surrounding vehicles corresponding to the target projection distance are mapped to the ST map, wherein the ST map includes the mapping time and distance of the surrounding vehicles mapped to the ST map; Obstacle areas are drawn in the ST map according to the time and distance at which the surrounding vehicles are mapped to the ST map.

[0018] Optionally, determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision tag is the confluence acceleration tag, determining whether it is a mapping period according to the mapping time in the ST diagram; If yes, then the following distance line is obtained in advance, wherein the following distance line is the following distance line corresponding to the target ST reference line generated when the vehicle planning decision label is any one of the following label, the keeping distance label and the cutting label. If no, then the calibration query table is set according to the collision time between the target vehicle and the merging vehicle obtained in advance, and the speed bias coefficient query table is set; Obtaining a velocity bias coefficient by linear interpolation according to the collision time; generating a reference speed of the target vehicle according to the speed bias coefficient, the current speed of the target vehicle and the current speed of the merging vehicle; generating a merging transition reference line according to the reference vehicle speed; A uniform speed line corresponding to the target vehicle is set according to the reference vehicle speed; A target ST reference line corresponding to the target vehicle is constructed according to the uniform speed line and the confluence transition reference line.

[0019] The present invention discloses a vehicle speed planning device, the device comprising: A first determination module is used to determine a vehicle planning decision label corresponding to a target vehicle through collected surrounding information; A second determination module is used to determine the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; A smoothing processing module, used for smoothing the target ST reference line to generate a smoothed ST reference line; The output module is used to derive the smooth ST reference line and output a target speed planning curve.

[0020] An embodiment of the present invention further discloses an electronic device, comprising at least one processor and a memory communicatively connected to the at least one processor; The memory is used to store computer programs; The processor is used to implement the method described in the embodiment of the present invention when executing the program stored in the memory.

[0021] The embodiment of the present invention further discloses a computer-readable storage medium on which a computer program is stored. When executed by one or more processors, the processors are enabled to execute the method described in the embodiment of the present invention.

[0022] Beneficial effects of the present invention: The embodiment of the present invention determines the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information; determines the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; smoothes the target ST reference line to generate a smooth ST reference line; derives the smooth ST reference line, and outputs the target speed planning curve. That is, the embodiment of the present application abstracts and refines the intelligent driving longitudinal decision of different scenes into a variety of vehicle planning decision labels, generates different target ST reference lines based on different labels, and the target ST reference line at this time is a rough ST reference line, and then smoothes the target ST reference line, constructs the objective function about the ST reference line using the quadratic programming algorithm, calls the solver in combination with other constraints, and outputs the target speed planning curve that best meets the requirements, that is, the speed and acceleration curve can be output, that is, the planning decision label of the target vehicle is determined by collecting surrounding information, and the target ST reference line is generated based on different labels, and then optimized by smoothing and quadratic programming algorithms, and finally outputs the optimal speed and acceleration curve. This method not only improves the decision-making accuracy and response speed of the intelligent driving system, but also enhances the stability and safety of the system, so that the vehicle can be more intelligent and efficient in speed control in different driving scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A flow chart of the steps of a vehicle speed planning method provided in an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a vehicle speed planning device provided in an embodiment of the present invention; Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present invention; Figure 4 This is a block diagram of an exemplary longitudinal speed planning software architecture provided in an embodiment of the present invention; Figure 5 A schematic diagram of an exemplary speed planning technology route based on ST diagram reference lines provided in an embodiment of the present invention; Figure 6 A schematic diagram of a scenario in which an exemplary predicted trajectory is mapped into an ST graph provided by an embodiment of the present invention; Figure 7 A schematic diagram of an exemplary predicted trajectory mapped into an ST diagram provided by an embodiment of the present invention; Figure 8 A schematic diagram of an exemplary generalized vehicle following reference line and a planned ST curve provided in an embodiment of the present invention; Fig. 9 A schematic diagram of an exemplary close-range invasion and a reasonable ST reference line curve provided in an embodiment of the present invention; Fig.10 A schematic diagram of an exemplary smooth transition process during intrusion provided in an embodiment of the present invention; Fig.11 A schematic diagram of a typical scenario of decelerating through a label provided in an embodiment of the present invention; Fig.12 A schematic diagram of an ST diagram of an exemplary simple scenario deceleration through a label provided in an embodiment of the present invention; Fig.13 It is a schematic diagram of an ST diagram of an exemplary complex scene deceleration through a label provided in an embodiment of the present invention; Fig.14 A schematic diagram of an exemplary confluence scenario provided in an embodiment of the present invention; Fig.15 A schematic ST diagram of an exemplary confluence acceleration label provided in an embodiment of the present invention; Fig.16 A schematic diagram of an exemplary ST diagram of converging deceleration provided in an embodiment of the present invention; Fig.17 A schematic diagram of an output curve of a quadratic programming solver provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways according to different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0025] It should be noted that in the embodiments of the present application, reference is made to Figure 4 , Figure 4 It is an exemplary overall speed planning software framework, which mainly consists of four parts.

[0026] The first part is the common module of the software, including common functions, common constant parameters, common data structures and visualization, as well as error diagnosis information, which facilitates the output of error information in each link for analysis; The second part is the input adapter, which includes the signal reception of state machine information, target vehicle information, target information, etc., as well as some logical conversions. All input data constitute a data center Data_center; Part 3 is the core functional module of the speed planning software, including ST map target mapping, fusion speed limit and acceleration and arbitration, ST map path search, ST map reference line, speed optimizer, and warning information and post-processing; The fourth part is the output adapter, which includes the output to the vertical control, HMI and some debug signal outputs.

[0027] Reference Figure 6 , attached Figure 6 This is a schematic diagram of mapping the target vehicle to the ST graph. The S coordinate in the ST graph is the distance, and T is the corresponding time. Figure 6 Vehicles No. 3 and No. 2 are surrounding traffic vehicles. It should be noted that the subsequent static traffic vehicles, the front vehicle, the rear vehicle, the merging vehicle, and the descriptions of the first vehicle, the second vehicle, and the third vehicle, that is, except for the target vehicle, the rest are vehicles traveling within the preset range of the same road section as the target vehicle, and are also vehicles that have an impact on the speed planning of the target vehicle and need to be considered.

[0028] Back to Figure 6 , where the predicted trajectories corresponding to cars 2 and 3 are shown in the curve. Car 1 is the host car, i.e., the target vehicle in the follow-up. The vehicle no-entry warning sign in front of car 1 represents some static traffic vehicles, so car 1 plans a lane change path, such as Figure 3 The white curve is shown.

[0029] Analyze the impact of car 2 and car 3 on the white curve, combined with Figure 6 It can be seen that when Figure 6 When car No. 2 in the figure reaches position P1, the projection point F in the right front of the vehicle just touches the white track, and the length of the track occupied by the white curve is 0; when car No. 2 reaches position P2, the projection points of car No. 2 on the white curve are G1 and G2 respectively, and the length of the track occupied by G1 and G2 on the white curve is LG; as car No. 2 continues to drive, when it reaches position P3, the projection point H on the left rear of the vehicle just leaves the white track, and the length of the track occupied by the white curve becomes 0 again. Therefore, this process can be expressed by a two-dimensional graph of "occupied track length-time", that is, the ST graph preset in the embodiment of the present application, such as Figure 7 Therefore, combined with Figure 6 , 7 Let's see, when Figure 6 When car No. 2 reaches P1 at t4, it is reflected in Figure 7 The ST diagram is a point F; when car 2 drives to point P2 at t5, it is reflected in the ST diagram as a line segment of length LG. Therefore, after discretizing time, the trajectory of car 2 from position P1 to position P3 is mapped to a two-dimensional area in the ST diagram. In summary, Figure 6 As long as a monotonically increasing and continuously differentiable white curve is arbitrarily planned to bypass the two traffic vehicle areas, the safe lane change of vehicle No. 1, i.e. the target vehicle, can be guaranteed. The derivative of the white curve is the speed sequence curve. Therefore, it can be concluded that the purpose of the embodiment of the present application is to search for the obstacle avoidance curve in the ST graph, and then obtain the speed and acceleration planning based on the obstacle avoidance curve, that is, the speed planning process.

[0030] Reference Figure 1 , shows a flowchart of a method for processing a high-precision map provided in an embodiment of the present invention, which specifically includes the following steps: Step 101, determining a vehicle planning decision label corresponding to a target vehicle through the collected surrounding information; Furthermore, before the step of determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information, the method includes: Draw a rectangular frame of surrounding vehicles according to the predicted trajectory information of the surrounding vehicles of the target vehicle; Obtaining a set of projection distances between each preset point position of the target vehicle and a pre-acquired planned path of the target vehicle; Taking the minimum value in the projection distance set as the target projection distance; If the target projection distance is less than a preset threshold, the surrounding vehicles corresponding to the target projection distance are mapped to the ST map, wherein the ST map includes the mapping time and distance of the surrounding vehicles mapped to the ST map; Obstacle areas are drawn in the ST map according to the time and distance at which the surrounding vehicles are mapped to the ST map.

[0031] It should be noted that in the embodiment of the present application, before obtaining the vehicle planning decision label, the ST diagram in the embodiment of the present application has been pre-constructed.

[0032] Specifically, an exemplary expression of the constructed ST diagram can be referred to Figure 7 First, calculate the longitudinal cumulative length of the planned path of the vehicle and match the length index number; receive the 8s predicted trajectory information of the surrounding vehicles at a time interval of 0.1s, and draw a rectangular frame of the surrounding vehicles for each trajectory point and heading angle of the surrounding vehicles; after taking the target vehicle as a rectangle, take the left rear, left body midpoint, left front, right front, right body midpoint, and right rear of the target vehicle as 6 judgment points, and calculate the projection distance from the point to the planned path of the vehicle; take the minimum value of the above 6 projection vertical distances. If the minimum value is less than the preset judgment threshold, it means that the surrounding vehicles will invade the planned path of the vehicle at some point in the future, and then map the obstacle vehicle to the ST map; if it needs to be mapped to the ST map, select the minimum index minIndex and the maximum index maxIndex of the projection point of the 6 judgment points on the path of the vehicle, and the two indexes correspond to the two distances of the planned path of the vehicle; finally, draw the obstacle area of ​​the ST map according to the time and distance mapped to the ST map.

[0033] It should be noted that in the embodiment of the present application, the time range of the ST diagram can be defined as 0-8s, the time discreteness is 0.1s, and the total number of points is 81.

[0034] After the ST diagram is constructed, you can search for pathways. Figure 7 , a total of three speed curves can be planned in the three areas: dashed line, solid line and dotted line. The three areas can be regarded as three search solution spaces, and the actual meanings of different solution spaces are different. For example, if the vehicle drives along the dashed line, it means that the vehicle is overtaking vehicle No. 3; if the vehicle drives along the dotted line, it means that the vehicle is avoiding vehicle No. 2. The specific path to take can be determined according to Figure 6 The upstream decision-making block in the architecture informs.

[0035] When the path is determined, an array structure can be used to store the upper and lower boundary information of the path. The structure information contains information such as index, time, upper boundary, lower boundary, etc., which is used for the subsequent generation of ST reference lines and constraints for quadratic programming solution.

[0036] Step 102, determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; When the ST path is determined, the position constraints are determined. In this path, a rough reference line can be generated in advance according to certain rules, that is, the target ST reference line, which is an array containing time t and distance s. After that, the rough ST reference line is smoothed according to a certain cost function and constraint conditions using a smoothing algorithm to generate a smooth ST curve that meets the requirements. The planned velocity curve and acceleration curve can be obtained by taking the first and second order derivatives of the ST curve.

[0037] It should be noted that in the above steps 101-102, the target ST reference lines corresponding to different labels are obtained based on the vehicle planning decision label. Therefore, it is first necessary to determine what the vehicle planning decision label corresponding to the target vehicle is in the current scenario.

[0038] Among them, there are many types of vehicle planning decision labels, mainly including following the vehicle (FOLLOW), keeping distance (KEEPDIS), slowing down and passing (SPEEDCUT), cutting in (CUTIN), merging acceleration (MERGEACCEL), merging deceleration (MERGEDECEL), etc. Different types of vehicle planning decision labels correspond to different longitudinal speed planning strategies.

[0039] Next, the details of generating the target ST reference line according to each vehicle planning decision label type will be explained.

[0040] First of all, the first one is the following tag, the keep distance tag and the cut-in tag.

[0041] In one embodiment, the above steps 101-102 may further include, that is, step 101, the vehicle planning decision label corresponding to the target vehicle is determined by collecting surrounding information, including: When there is a first vehicle in front of the target vehicle in the driving lane, and the target vehicle maintains a preset following distance with the second vehicle, the vehicle planning decision label corresponding to the target vehicle is a following vehicle label; When the preset following distance corresponding to the target vehicle is changed, the vehicle planning decision tag corresponding to the target vehicle is a keep distance tag; When there is a third vehicle that changes lanes and cuts into the lane where the target vehicle is traveling within a preset range of the lane where the target vehicle is traveling, the vehicle planning decision label corresponding to the target vehicle is a cut-in label.

[0042] It should be noted that in the embodiment of the present application, the following vehicle label, for example: suppose there is a dynamic vehicle traffic vehicle 100 meters ahead, and its speed is 60kph. According to a large number of tests, the following distance recognized by most drivers is 18m. If the decision label is following the vehicle, the target vehicle is expected to reach a position of 18m.

[0043] Keep distance tags are suitable for some scenarios where the standard following distance needs to be increased or decreased. For example, if a traffic vehicle is driving on the line for a long time but will not cut in, the target vehicle will keep a certain safety distance from it to ensure safety. This distance is generally less than the following distance.

[0044] The cut-in label means that a vehicle in the adjacent lane will change lanes and cut into the lane, affecting the acceleration of the target vehicle. Since the core processing ideas of the cut-in label, the distance keeping label and the following label are basically similar in the close-range intrusion and long-range following scenarios, the process of obtaining the target ST reference line is the same. Taking the following label as an example, the following label is the most common decision label, which requires maintaining a certain safe following distance from the vehicle in front.

[0045] Step 102 includes, that is, determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision label is any one of the following label, the keeping distance label and the cutting-in label, obtaining a standard following distance; Determine a target following distance according to a pre-acquired first current scene of the target vehicle and the standard following distance, and construct a following distance line in a pre-set ST diagram for the target following distance; Determine a second current scene of the target vehicle according to the target following distance, and construct a transition reference line corresponding to the target vehicle in a preset ST diagram; A target ST reference line is constructed in a preset ST diagram according to the following distance line and the transition reference line.

[0046] The safe following distance when following a vehicle driving straight in a single lane is obtained through a large number of test calibrations, so it can be called the standard following distance. When encountering other special scenarios such as lane changing, in order to balance safety and comfort, it may be necessary to reduce or increase the standard following distance. The reference line generation strategies for different scenarios are different.

[0047] Therefore, you can refer to Figure 5 First, a target following distance is determined based on a first current scene of the target vehicle and a standard following distance acquired in advance. The target following distance is a following distance obtained by adjusting the standard following distance based on an acquired bias coefficient.

[0048] Further, determining a target following distance according to the pre-acquired first current scene of the target vehicle and the standard following distance, and constructing a following distance line in a pre-set ST diagram for the target following distance includes: When the current scenario is that the target vehicle is in a lane-changing state, obtaining a collision time between the target vehicle and a rear vehicle in a lane adjacent to the lane in which the target vehicle is traveling, and determining a rear vehicle safety level according to the collision time; If the safety level of the rear vehicle is in a dangerous state, calibrate a first bias coefficient according to a speed difference between the target vehicle and a front vehicle in a lane where the target vehicle is traveling; If the rear vehicle safety level is a safe state, calibrating the first bias coefficient according to the time headway intrusion level between the target vehicle and the front vehicle in the lane where the target vehicle is traveling; When the current scene is that the target vehicle is in a lane change end transition state, preliminarily calibrate the transition state time according to the intrusion time distance, obtain a first time offset coefficient according to the speed difference, a second time offset coefficient according to the intrusion rear vehicle time distance level, and obtain a time distance coefficient increment; generating the first bias coefficient according to the first time bias coefficient, the second time bias coefficient and a time interval coefficient increment; When the current scenario is that the target vehicle is in a non-lane-changing state, setting the first bias coefficient to a target value; adjusting the standard following distance according to the first bias coefficient to obtain a target following distance; A following distance line is constructed in a preset ST map for the target following distance.

[0049] Further, determining the rear vehicle safety level according to the collision time includes: When it is detected that the collision time is less than a first preset value, the safety level of the rear vehicle is in a dangerous state; When it is detected that the collision time is a negative number or infinite, an intrusion time level between the target vehicle and the rear vehicle is obtained, and when the intrusion time level is less than a second preset value, it is determined that the safety level of the rear vehicle is in a dangerous state.

[0050] Therefore, in the first case, when the current scene is that the target vehicle is in a lane-changing state, the relative position and speed difference between the target vehicle and the vehicle in the target lane need to be considered. The specific steps are as follows: First, determine the safety level of the vehicle behind in the adjacent lane.

[0051] When the target vehicle changes lanes, it must consider not only the vehicle in front of it but also the vehicle behind it. When the vehicle behind is dangerous (closer), the target vehicle needs to generate a large enough acceleration in a short time to quickly change lanes and avoid being rear-ended by the vehicle behind.

[0052] The specific steps are as follows: 1) Calculate the collision time between the target vehicle and the rear vehicle in the target lane. If the collision time is less than 6, it can be considered that the target vehicle starting to change lanes will cause great pressure on the rear vehicle, and the rear vehicle may not brake enough and cause a rear-end collision, so it is considered very dangerous; 2) If the speed of the target vehicle is very close to that of the rear vehicle, the calculated collision time may be negative or infinite, resulting in invalid judgment. In this case, the judgment is made based on the time headway level of the target vehicle invading the rear vehicle (the calculation of time headway intrusion can be referred to above). When invade_class<0.2, it is also considered to be relatively dangerous.

[0053] Secondly, if there is danger behind, the bias coefficient is calibrated according to the speed difference between the target vehicle and the front vehicle in the target lane.

[0054] For example, let the speed difference (obstacle vehicle speed - target vehicle speed) lookup table be spd_err_kph_table = {-20.0, -15.0, -10.0, -7.0, -5.0, -2.0, 0.0, 2.0, 5.0, 7.0, 10.0, 15.0}, and let the bias coefficient table be gap_coeff_table = {1.0, 0.8, 0.6, 0.5, 0.4, 0.3,0.2, 0.16, 0.1, 0.05,0.03, 0.02}. The above calibration table is interpreted as follows: the greater the speed difference, the more dangerous it is. In order to ensure the safety of following the vehicle, the bias coefficient should be closer to 1, such as a scenario where the target vehicle changes lanes at high speed to a stationary vehicle in front of the adjacent lane; conversely, it means it is very safe, and the bias coefficient should be closer to 0. For example, when there is a fast-approaching vehicle behind the adjacent lane and the vehicle just overtakes the target vehicle, the target vehicle can slowly accelerate behind it. At this time, although the time-distance intrusion is large, the driver does not feel it is dangerous because the speed of the vehicle is much faster than that of the target vehicle. Therefore, the bias coefficient should be small enough to prevent the target vehicle from being unable to follow its acceleration or even braking.

[0055] Secondly, if the rear is safe, the bias coefficient is calibrated according to the time distance between the target vehicle and the vehicle in front of the target lane. If the rear is safe, the target vehicle can control the speed of the target vehicle more "calmly", as long as the target vehicle does not hit the front vehicle, and does not need to care whether the rear vehicle is the rear-end target vehicle. At this time, the relative distance between the target vehicle and the front vehicle in the target lane can be combined for calibration. For example, it can be divided into four levels according to the time-distance intrusion level: invade_class<0.6, 0.6<=invade_class&&invade_class<1.2, 1.2<= invade_class&&invade_class<2.5, invade_class>2.5. Each level branch recalibrates the gap_coeff_table in 4.1.2. In this way, the parameter decoupling calibration of whether there is danger in the rear can be carried out, which is convenient for engineering application.

[0056] The second is when the current scene is that the target vehicle is in a transition state of lane change completion. Generally speaking, the lane change is completed when the lateral position of the center point of the target vehicle is within a certain distance threshold range of the center line of the target lane. At this time, the longitudinal direction may still intrude into the standard following distance of the vehicle in front, so it is necessary to slowly increase the following distance at a safe and comfortable deceleration. The specific operation steps are as follows: Preliminary calibration of the transition state time transition_STATe_time based on the intrusion time interval Generally speaking, the more intrusion time intervals there are, the longer the time to pull apart will be. Therefore, it can be calibrated according to the current intrusion time interval level. For example, let the intrusion time interval level query table be time_gap_invade_class_table = {0.1, 0.2, 0.3,0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0}, and let the transition state time query table be transition_STATe_time_table = {10.0, 9.0, 8.0, 7.0, 6.0,5.0, 4.0, 3.0, 2.0, 0.01}. Explanation: If the current intrusion is half of the time interval, the slow pull apart time obtained from the table is 6s. Calculate the time offset coefficient time_offset_coeff_spd_err based on the speed difference. Step 1) only analyzes the intrusion time interval level. Different speed differences also have a certain impact on the slow pull-off time. Generally speaking, the smaller the target speed is than the target vehicle, the shorter the pull-off time is, and the greater the deceleration is for a quick pull-off.

[0057] Calculate the time offset coefficient time_offset_rr_obj_timegap_invade based on the time gap level of the invading rear vehicle. Generally speaking, after the lane change is completed, if there is a car close behind, if the target vehicle decelerates too quickly at this time, it may cause the rear vehicle to have no time to brake and cause a rear-end collision. Therefore, the time for the target vehicle to slowly pull away is also affected by the degree of danger of the rear vehicle. Therefore, the time gap level of the target vehicle invading the rear vehicle can be calculated, a one-dimensional query table can be established, and then the time offset coefficient based on the time gap level of the invading rear vehicle can be calibrated.

[0058] Calculate the increment of the time distance coefficient when the vehicle is slowly moving away. When the vehicle is slowly moving away, the standard following distance bias coefficient must increase gradually. Therefore, the increment calculation method for each cycle is shown in Formula 1: delt_coeff = (1 - time_gap_invade_class) / (transition_STATe_time *time_offset_coeff_spd_err * time_offset_rr_obj_timegap_invade) *delt_t (Formula 1) Among them, in the above formula 1, time_gap_invade_class represents the time gap level of the target vehicle's current intrusion into the front vehicle, delt_t represents the software operation cycle, time_offset_rr_obj_timegap_invade is the time offset coefficient based on the time gap level, transition_STATe_time is the time parameter in the transition state time query table, and time_offset_coeff_spd_err is the time offset coefficient based on the speed difference.

[0059] Calculate the first bias coefficient of this cycle, formula 2 is as follows: gap_coeff = gap_coeff_laST + delt_coeff (Formula 2) In the above formula 2, gap_coeff is the first bias coefficient, and delt_coeff is the time interval coefficient increment.

[0060] In the third, when the lane is not changed and the lane change is not in the transition state.

[0061] When there is no lane change and the vehicle is not in the transition state of lane change termination, it indicates that this is a single-lane cruise following scenario, which is the most common and ordinary scenario and is processed according to the first bias coefficient gap_coeff = 1.

[0062] It should be noted that in the embodiment of the present application, when calculating the bias coefficient and the like, two parameters are involved, the first being the collision time, and the second being the intrusion time interval level.

[0063] Specifically, the time to collision (TTC) refers to the time of collision with the vehicle in front, which represents the degree of danger with the vehicle in front. The smaller the value, the more dangerous it is. Calculating the collision time in advance can provide a quantitative indicator basis for the degree of danger for subsequent calibration parameters. The calculation steps are as follows: Considering that the longer the 8s prediction trajectory is, the lower the accuracy is, the higher confidence prediction trajectory interval is determined to be the first 2s; For the future time i, the predicted trajectory is attached with the predicted speed , the target vehicle's speed prediction is calculated according to the uniform acceleration formula, that is, the following formula 3: (Formula 3) Calculate the collision time at each time point in the first 2 seconds, compare and take the smaller one, and calculate according to formula 4: (Formula 4) Among them, in formula 4, and Represent the relative distance and relative speed (m / s) between the target vehicle and the front vehicle respectively.

[0064] Further, the obtaining of the intrusion time headway level between the target vehicle and the rear vehicle includes: The intrusion time-to-distance level is generated according to a pre-calibrated target following distance and a ratio of the distance between the target vehicle and a front vehicle or a rear vehicle in the lane.

[0065] It should be noted that in the embodiment of the present application, the intrusion headway level also represents the degree of danger to the preceding vehicle, and the smaller the value, the more dangerous it is. Calculating the intrusion headway level in advance can also provide a quantitative indicator basis for the degree of danger for subsequent calibration parameters. The calculation steps are as follows: Calibrate the target following distance. Generally, intelligent driving vehicles will set the following distance level on the human-machine interface. At the same steady-state following speed, different distance levels correspond to different following distances. The most appropriate following distance can be obtained through a large number of tests and calibrations. For example, if the current gear is 1st gear and the speed of the front vehicle is 60km / h, the calibrated target steady-state following distance is target_pos = 18m.

[0066] Calculate the intrusion level. For example, assuming that the distance between the front vehicle and the target vehicle is actual_pos = 9m, the intrusion time interval level invade_class is calculated as shown in the following formula 5.

[0067] (Formula 5) In the above formula 5, target_pos is the pre-calibrated target following distance, actual_pos is the actual distance between the target vehicle and the vehicle in front or behind the target vehicle during driving, and invade_class is the intrusion time-to-distance class.

[0068] Further, the determining the second current scene of the target vehicle according to the target following distance and constructing a transition reference line corresponding to the target vehicle in a preset ST diagram includes: When the second current scene is long-distance non-intrusion, a calibration lookup table is set according to a pre-acquired collision time between the target vehicle and the front vehicle, and a speed bias coefficient lookup table is set; Obtaining a velocity bias coefficient by linear interpolation according to the collision time; generating a reference speed of the target vehicle according to the speed bias coefficient, the current speed of the target vehicle and the current speed of the vehicle ahead; Constructing a first transition reference line corresponding to the target vehicle in a preset ST diagram according to the reference vehicle speed; When the second current scene is a close-range intrusion, obtaining a safety acceleration; Acquiring an acceleration bias according to the safety acceleration; generating a smooth acceleration according to the safety acceleration and the acceleration offset; A second transition reference line corresponding to the target vehicle is constructed in a preset ST diagram according to the smoothed acceleration.

[0069] It should be noted that in the embodiments of the present application, as shown in the attached Figure 8 It can be seen that the target following distance is far, and a curve OA needs to be generated for transition. If the starting point of the target following distance just passes through point O, no transition curve is needed. Therefore, it mainly involves two situations: long-distance non-intrusion and close-distance intrusion.

[0070] The first one is to calculate the first transition reference line when there is no invasion at a long distance.

[0071] As attached Figure 8 It can be seen that when the target following line at point 0 is greater than 0, it can be considered a long-distance non-intrusion scenario. At this time, the target vehicle is still far away from the target following vehicle, and the main factor that determines the degree of danger is not the distance factor, but the speed difference between the target vehicle and the traffic vehicle in front. The general operation steps for this scenario are as follows: Calculate the collision time. If the collision time is less than 100 and greater than 0, it can be considered that the scene is relatively dangerous and braking is required. At this time, the calibration query table is set according to the collision time, collision time_table = {2,4,6,8,10,12, 20, 50, 100}, and then the speed bias coefficient query table coeff_table = {0,0.2,0.4,0.6,0.8,0.9, 0.92, 0.95, 1.0} is set. According to the corresponding collision time, linear interpolation is performed to obtain the speed bias coefficient coeff. Finally, according to the speed bias coefficient, the target vehicle's expected speed is calculated: (Formula 6) In the above formula 6, represents the target vehicle speed, Represents the speed of the traffic vehicle, and coeff is the speed bias coefficient.

[0072] If the collision time does not meet the above conditions, it can be considered that the traffic vehicle ahead is far away and has a high speed. Therefore, dynamic calibration can be performed based on the speed difference. For example, if the speed difference lookup table is spd_kph_err_table = {-2.0, 0.0, 2.0, 5.0, 8.0, 10.0, 15.0}, and the speed difference bias coefficient lookup table is coeff_table = {0.2, 0.3, 0.5, 0.6, 0.7, 0.9, 1.0}, the speed that the target vehicle is expected to reach can be obtained by looking up the table.

[0073] After obtaining the reference speed, according to the uniform speed calculation formula, the following can be generated: Figure 8 Curve 1 in the figure is the first transition reference line.

[0074] The second method is to calculate the second transition reference line when intrusion like Fig. 9 As shown in the figure, some s are less than 0, indicating that when the traffic vehicle cuts in, the target vehicle has a certain amount of intrusion. Therefore, when dealing with this scenario, there are two points to consider: 1) Considering that the generalized following reference line should not have a situation where s is less than 0 (the target vehicle cannot reverse at a reverse speed), the dotted line ABC is processed as a broken line 0BC; 2) Since the slope of the OB segment is 0, if qp is directly used to smooth this segment of the broken line, a particularly large deceleration will be generated. Therefore, for close-range intrusion, the broken line segment OBC needs to be smoothed in advance according to a certain acceleration.

[0075] Therefore, first, calculate the safe acceleration safe_accel, which represents a theoretical safe acceleration value when the target vehicle interacts with the traffic vehicle. The specific steps are as follows: First, generate a series of pending acceleration tables according to acc_table = {2,1.5,1,0.5,0,-0.5,-1,-1.5,-2}.

[0076] Second, starting from an acceleration of 2, generate a uniform acceleration curve and check whether the curve intersects with the traffic vehicle mapping polygon.

[0077] Third, if they intersect, take the next acceleration 1.5 and perform the same operation as the second step above. If they do not intersect, output the maximum safe acceleration for the target vehicle to interact with the traffic vehicle. Fig.10 As shown, let curve 1 be the curve obtained according to the acceleration of 0.5, curve 2 be the uniform speed line of the current speed of the target vehicle, and curve 3 be the curve obtained according to the acceleration of -0.5. From the above, it can be seen that the maximum safe acceleration for the target vehicle to interact with the traffic vehicle is -0.5.

[0078] Next, the acceleration offset offset_accel is calculated.

[0079] Considering the accuracy of the predicted trajectory and the driver's need to have a certain psychological expectation of the safety threshold, the above safety acceleration is only a perfect theoretical value and cannot represent the various complex scenarios encountered in engineering applications. Therefore, it is necessary to reasonably bias the above safety acceleration to ensure a safe and comfortable transition. The specific steps to obtain the bias acceleration are as follows: First, if safe_accel>0.5, it means it is very safe. The typical scenario is as follows: a fast car coming from behind cuts in at a close distance. At this time, the target speed is high or the acceleration is very large, and the safe acceleration of the target vehicle is also large. Consider two aspects: If the calculated collision time is less than 12, you can refer to collision_table = {4.0, 6.0, 8.0, 10.0,12.0}, offset_accel_table = {-1.8, -1.6, -1.4, -1.2, -1} to calibrate the offset acceleration. If the above conditions are not met, calibration is performed according to the speed difference between the target vehicle and the traffic vehicle. For reference, spd_kph_err_table = {-10.0, -8.0, -6.0, -4.0, -2.0, 0.0,2.0, 4.0, 6.0, 8.0, 10.0, 15.0}; offset_accel_table = {-1.0, -0.9, -0.8, -0.6, -0.4, -0.3,-0.2, -0.2, 0.0,0.0, 0.0, 0.0}.

[0080] Second, if the safety acceleration is greater than safe_accel>-0.5 and safe_accel<=0.5, this is a scenario that only requires slight deceleration. Typical scenarios are as follows: The vehicle in the side lane is close to the same speed as the target vehicle, and the cutin is close, and the intrusion time interval may continue for a long time. At this time, safe_accel is likely to be calculated inaccurately, and it will jump at -0.5, 0.0, and 0.5, so these three acceleration segments need to be considered together. If the calculated collision time is less than 12, refer to the corresponding steps in the first method. If the above conditions are not met, consider dividing the grades again according to the speed difference: 1) If the absolute value of the speed difference is less than 4km / h, it may be impossible to open the distance for a long time, causing a sense of oppression, so it is necessary to increase the acceleration bias here to force deceleration; 2) Otherwise, it can be floated up and down according to the speed difference calibration table in the first, and the calibration can be improved through actual vehicle test problems.

[0081] Third, if the safety acceleration is greater than safe_accel>-0.5, the typical scenario is as follows: cutin when the vehicle in the side lane is much lower than the target vehicle. The solution for this scenario is basically similar to the first part. The difference is mainly that the various calibration parameters are different, which will not be repeated.

[0082] Secondly, the smooth acceleration smooth_accel is obtained by adding the safety acceleration and the acceleration offset, that is: (Formula 7) Among them, in the above formula 7, smooth_accel is the smooth acceleration, safe_accel is the safe acceleration, and offset_accel is the acceleration offset.

[0083] It should be noted that, as attached Fig.10 As shown in the figure, after obtaining the smooth acceleration, a uniform acceleration line OD is generated and intersects with the line segment ABC at point D. The final generalized following vehicle ST reference line is ODC.

[0084] In another embodiment, the above steps 101-102 may further include, that is, step 101, the vehicle planning decision label corresponding to the target vehicle is determined by collecting surrounding information, including: When the collected surrounding information shows that there is an obstacle in front of the target vehicle's driving lane, and the lanes on both sides corresponding to the target vehicle's driving lane are in an obstacle-free state, the vehicle planning decision label corresponding to the target vehicle is a deceleration passing label.

[0085] When a decision is made to issue a slow-down passing tag to a certain traffic vehicle, a slow-down passing time will be issued at the same time. The meaning of this time is: during this time, lane change, obstacle avoidance or deviation will be initiated horizontally, and longitudinally, it is only necessary to ensure that there is no collision with the traffic vehicle within this time.

[0086] As attached Fig.11 As shown in the figure, let A be the vehicle, there is a traffic vehicle B in front of this lane, the target vehicle changes lanes to the left lane according to navigation needs, and there is a traffic vehicle C in front of the left lane. When car C is not considered first, the influence of different mapping methods of the predicted trajectory of car B on speed planning is discussed, as shown in the figure. Fig.12 shown.

[0087] (1) When deciding to give a following car label to car B, its 8s mapping area should be Fig.12 The solid line mapping polygon on the left and the dotted line mapping polygon on the right together constitute. Assuming that the constant speed line of the target vehicle's current speed is shown as dotted line 1, it will intersect with the mapping polygon at time t1, that is, a collision will occur at time t1. Therefore, the ST curve planned according to the following vehicle label is shown as solid line 2. It can be seen that in order to avoid collision with low-speed targets, the slope of solid line 2 is small, which means that more deceleration is required.

[0088] (2) When the decision is made to give a deceleration passing label with time t2, the dotted polygon part after time t2 can be removed, and then a "scratch-the-edge" curve 3 is planned from above time t2. The target vehicle can ensure that it will not collide with traffic vehicles within time t2 by strictly following this curve.

[0089] It can be seen from the three curves that the only curve 3 where the target vehicle can achieve a slight and comfortable deceleration while changing lanes is curve 3.

[0090] Further, step 102 may include: the deceleration passing tag includes a first deceleration passing tag, and the determining of the target ST reference line of the target vehicle corresponding to different vehicle planning decision tags includes: When the vehicle planning decision tag is the first deceleration passing tag, obtaining the deceleration passing time corresponding to the target vehicle; Obtaining a lower boundary distance of a lower boundary of the deceleration time map in a preset ST diagram; Obtaining a ratio of the lower boundary distance to the deceleration passing time to obtain a first constraint curve, wherein the first constraint curve includes a speed; The minimum value of the speed of the constraint curve and the pre-acquired speed limit of the target vehicle driving lane is taken to obtain a target ST reference line corresponding to the target vehicle.

[0091] The first deceleration pass tag is a simple scene deceleration pass tag.

[0092] Back to Fig.11 , when car C is not considered and only car B exists in the scene, this is called a simple scene of deceleration. The steps for generating the ST reference line are as follows: According to the deceleration time t2 decided, the trajectory mapping part after t2 is removed; The longitudinal distance corresponding to the lower boundary of the mapping at time t2 is s2, which defines the speed of deceleration through the constraint curve. As follows: (Formula 8) In the above formula 8, s2 is the longitudinal distance corresponding to the lower boundary of the mapping at time t2, and t2 is the deceleration passing time.

[0093] When there are no other mapped obstacles in the ST diagram, the speed of the constraint curve will be compared with the current road speed limit and the smaller one will be taken. The curve after comparison is the generalized ST reference line.

[0094] Further, step 102 may further include: the deceleration passing tag includes a second deceleration passing tag, and the determining of the target ST reference line of the target vehicle corresponding to different vehicle planning decision tags includes: When the vehicle planning decision label is the second deceleration passing label, obtaining the deceleration passing time corresponding to the target vehicle; Obtaining a lower boundary distance of a lower boundary of the deceleration time mapping in a preset coordinate system; Obtaining a ratio of the lower boundary distance to the deceleration passing time to obtain a first constraint curve, wherein the first constraint curve includes a speed; Taking the minimum value of the speed of the constraint curve and the pre-acquired speed limit of the target vehicle driving lane to obtain a second constraint curve; The intersection point of the pre-acquired following distance line and the second constraint curve is obtained to obtain a target ST reference line corresponding to the target vehicle.

[0095] When considering car C, and car C is the target of the following vehicle, since there are other mapped obstacles in the ST graph, this can be called a slow-down passing complex scene, corresponding to the second slow-down passing label and the target ST reference line. The generation steps are as follows: like Fig.13 As shown, first generate a constraint curve according to a simple scenario, then compare it with the speed limit line and take the smaller one to get dotted line 1; generate the target following line according to the following car label, that is, dotted line 2; dotted line 1 and dotted line 2 intersect at point A, then the broken line segment OAB is the generalized following car ST reference line.

[0096] In another embodiment, the above steps 101-102 may further include, that is, step 101, the vehicle planning decision label corresponding to the target vehicle is determined by collecting surrounding information, including: When the collected surrounding information indicates that there is a merging vehicle within a preset range of the target vehicle's driving lane, the vehicle planning decision label corresponding to the target vehicle is a merging label, and the merging label includes a merging deceleration label and a merging acceleration label.

[0097] like Fig.14 The figure shows a typical merging scenario. Let the target vehicle be car A and the traffic vehicle be car B. The decision module will make a decision in advance whether to drive in front of car B (overtaking) or behind (avoiding) based on the relative speed of the target vehicle and the traffic vehicle and the distance to the merging point. At this time, the decision behavior of overtaking can be named as the merging acceleration label, and the decision behavior of avoiding can be named as the merging deceleration label.

[0098] Further, step 102 may include that determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision tag is the merging deceleration tag, obtaining a predicted speed of the merging vehicle; Searching a pre-acquired safety distance calibration table for a target safety distance between the target vehicle and the merging vehicle corresponding to the predicted speed; The maximum value of the speed corresponding to the target vehicle and the speed corresponding to the merging vehicle is taken to obtain the target ST reference line corresponding to the target vehicle.

[0099] When the target vehicle overtakes other vehicles in a merging scenario, it needs to be able to maintain a sufficiently large safety distance from the other vehicles to avoid being rear-ended. Fig.15 As shown, the specific steps for generating ST reference lines are as follows: First, above the traffic vehicle mapping polygon, according to the predicted speed of the traffic vehicle, based on the empirical calibration table (for example, 60kph, the safe following distance is about 18m), the safety distance s1 is obtained, which represents the distance that the target vehicle is expected to maintain with the following vehicle when it reaches this position. Connect each distance point to generate a dotted line segment AB.

[0100] Since the predicted trajectory of the traffic vehicle does not intersect with the target vehicle before time t1, it is not mapped to the ST graph, so the dotted line endpoint A is temporarily extended to the left to C.

[0101] Since the slope of the dotted line BC is 0, in order to ensure that the final generalized following line is close to the planned curve, a guide line needs to be generated to connect BC. Here, considering the larger speed of the target vehicle and the traffic vehicle, a uniform speed line OD is generated, which then intersects with BC at point D.

[0102] Finally, the broken line segment ABDO is the generalized ST reference line for following the vehicle, and the planned ST curve is shown as the solid line in the figure.

[0103] Further, step 102 may also include that determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision tag is the merging acceleration tag, determining whether it is a mapping period according to a mapping time in a preset ST diagram; If yes, then the following distance line is obtained in advance, wherein the following distance line is the following distance line corresponding to the target ST reference line generated when the vehicle planning decision label is any one of the following label, the keeping distance label and the cutting label. If no, then the calibration query table is set according to the collision time between the target vehicle and the merging vehicle obtained in advance, and the speed bias coefficient query table is set; Obtaining a velocity bias coefficient by linear interpolation according to the collision time; generating a reference speed of the target vehicle according to the speed bias coefficient, the current speed of the target vehicle and the current speed of the merging vehicle; generating a merging transition reference line according to the reference vehicle speed; A uniform speed line corresponding to the target vehicle is set according to the reference vehicle speed; A target ST reference line corresponding to the target vehicle is constructed according to the uniform speed line and the confluence transition reference line.

[0104] When the target vehicle avoids other dynamic traffic vehicles in a merging scenario, it must fully consider safety and comfort. Fig.13 As shown, the specific steps for generating ST reference lines are as follows: Generate the ST reference line of the mapping interval. First, under the traffic vehicle mapping polygon, refer to the experience calibration table and the following vehicle label to obtain the standard following line AB. If there is a close vehicle following behind the traffic vehicle B, it means that the target vehicle may cause a rear-end collision if it avoids behind the vehicle B. The target following line that indents and follows the vehicle B should be adjusted according to the degree of danger of the following vehicle, that is, the following distance should be offset. The degree of danger of the following vehicle can be implemented according to the strategy of considering the rear target when changing lanes above, which will not be repeated.

[0105] Calculate the collision time. Assume that the time and distance when the traffic vehicle is first mapped into the ST graph are t1 and s1 respectively. Assume that the target vehicle will travel at a constant speed in the future. Calculate the collision time at the future time t1: (Formula 9) Among them, in the above formula 9, represents the target vehicle speed, represents the predicted speed of the traffic vehicle at time t1, Represents the length of line segment AE. Note that when calculating the future collision time, the future time should be added to represent the estimated collision time.

[0106] Calculate the constraint speed. If the collision time is less than 12 and greater than 0, it can be considered that the scene is relatively dangerous and braking is required. Set the collision time query table, collision time_table = {2,4,6,8,10,12}, and set the speed bias coefficient query table coeff_table = {0,0.2,0.4,0.6,0.8,0.9}. According to the corresponding collision time, linear interpolation is used to obtain the speed bias coefficient coeff. Finally, according to the speed bias coefficient, calculate the target vehicle's expected speed: (Formula 10) In the above formula 10, represents the target vehicle speed, Represents the speed of traffic vehicles.

[0107] If the collision time is not less than 12 and is not greater than 0, it means that the speed of the traffic vehicle is greater than the speed of the target vehicle, and the standard following time distance can be directly used to generate the ST reference line.

[0108] The uniform speed line based on the current speed of the target vehicle intersects AE at D. According to the expected speed target_spd, a uniform speed ST line is generated that intersects AB at C (it can also intersect line segment CD at a certain point). Then line segment OFB is the reference line that QP finally needs to smooth. The smoothing effect is shown by the solid line in the figure.

[0109] Step 103, smoothing the target ST reference line to generate a smoothed ST reference line; Step 104, deriving the smoothed ST reference line and outputting a target speed planning curve.

[0110] It should be noted that in the embodiment of the present application, a smoothing algorithm is used to smooth this rough ST reference line according to a certain cost function and constraint conditions to generate a smooth ST curve that meets the requirements. The planned velocity curve and acceleration curve can be obtained by taking the first and second order derivatives of the ST curve.

[0111] Specifically, the ST reference line is smoothed using the quadratic programming algorithm. The general expression of the quadratic programming algorithm is as follows: (Formula 11) In formula 11, is a vector, is a matrix.

[0112] A total of 8 segments of quintic polynomial curves are selected, and the quadratic programming algorithm needs to find the solution that satisfies the cost function under the constraints, that is, a total of 6×8 coefficients need to be solved. For speed planning application scenarios, the cost function can be quantitatively expressed from a variety of mathematical models such as comfort and smoothness, and the above formula is converted to: (Formula 12) In formula 12, Represents the weight of each cost function. The first item in the brackets represents the acceleration cost function, and the second item represents the jerk cost function. Focusing on the scenario of following the ST reference line, it is hoped that the deviation of the planned smooth ST curve from the ST reference line should be as small as possible, so the relative position deviation is introduced based on equation 12; (Formula 13) The third cost function is called the position deviation term, and its specific expression is as follows: (Formula 14) When the current position of the vehicle differs greatly from the target reference line position, the position deviation term has a larger proportion. Therefore, the quadratic programming algorithm will be affected by the position error term when smoothing the position sequence, so that the planned curve can meet the reference line position as much as possible, thereby gradually approaching the target speed limit value in a gentler way.

[0113] In addition, in the embodiments of the present application, refer to Fig.17 , taking the following vehicle tag as an example, the following set of road measured data: there is a traffic vehicle No. 118 about 38m in front of the target vehicle, with a speed of about 10m / s and an acceleration of about -0.1m / ss. According to the method of the long-distance non-intrusive following scene described above, the standard following distance can be calibrated to be about 15m. Therefore, this scene belongs to the "long-distance non-intrusive following scene", so combined with the collision time, speed difference and other information, the bias coefficient and transition curve can be further calibrated to obtain the final ST reference line.

[0114] The ST curve, VT curve, and AT curve output by the quadratic programming solver are as follows: Fig.17 As shown in the figure. From the ST and VT curves, it can be seen that the target vehicle has been decelerating within the 8-second planning space, and when the target following distance is reached, the planned vehicle speed is basically equal to that of the leading vehicle, ensuring safety. On the other hand, from the AT curve, it can be seen that the planned deceleration does not exceed -1.5, which also ensures comfort.

[0115] The embodiment of the present invention determines the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information; determines the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; smoothes the target ST reference line to generate a smooth ST reference line; derives the smooth ST reference line, and outputs the target speed planning curve. That is, the embodiment of the present application abstracts and refines the intelligent driving longitudinal decision of different scenes into a variety of vehicle planning decision labels, generates different target ST reference lines based on different labels, and the target ST reference line at this time is a rough ST reference line, and then smoothes the target ST reference line, constructs the objective function about the ST reference line using the quadratic programming algorithm, calls the solver in combination with other constraints, and outputs the target speed planning curve that best meets the requirements, that is, the speed and acceleration curve can be output, that is, the planning decision label of the target vehicle is determined by collecting surrounding information, and the target ST reference line is generated based on different labels, and then optimized by smoothing and quadratic programming algorithms, and finally outputs the optimal speed and acceleration curve. This method not only improves the decision-making accuracy and response speed of the intelligent driving system, but also enhances the stability and safety of the system, so that the vehicle can be more intelligent and efficient in speed control in different driving scenarios.

[0116] Reference Figure 2 , shows a vehicle speed planning device provided by an embodiment of the present invention, the device comprising: The first determination module 201 is used to determine the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information; A second determination module 202 is used to determine the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; A smoothing processing module 203, used for smoothing the target ST reference line to generate a smoothed ST reference line; The output module 204 is used to derive the smoothed ST reference line and output a target speed planning curve.

[0117] The embodiment of the present invention determines the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information; determines the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; smoothes the target ST reference line to generate a smooth ST reference line; derives the smooth ST reference line, and outputs the target speed planning curve. That is, the embodiment of the present application abstracts and refines the intelligent driving longitudinal decision of different scenes into a variety of vehicle planning decision labels, generates different target ST reference lines based on different labels, and the target ST reference line at this time is a rough ST reference line, and then smoothes the target ST reference line, constructs the objective function about the ST reference line using the quadratic programming algorithm, calls the solver in combination with other constraints, and outputs the target speed planning curve that best meets the requirements, that is, the speed and acceleration curve can be output, that is, the planning decision label of the target vehicle is determined by collecting surrounding information, and the target ST reference line is generated based on different labels, and then optimized by smoothing and quadratic programming algorithms, and finally outputs the optimal speed and acceleration curve. This method not only improves the decision-making accuracy and response speed of the intelligent driving system, but also enhances the stability and safety of the system, so that the vehicle can be more intelligent and efficient in speed control in different driving scenarios.

[0118] The embodiment of the present invention further provides an electronic device, such as Figure 3 As shown, it includes a processor 301, a device interface 302, a memory 303 and a bus 304; Memory 303, used for storing computer programs; The processor 301 is used to execute the program stored in the memory 303 to implement the following steps: Determine the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information; Determine the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; Smoothing the target ST reference line to generate a smoothed ST reference line; The smooth ST reference line is derived to output a target speed planning curve.

[0119] The bus mentioned in the above terminal can be a Peripheral Component Interconnect (PCI) bus or an Extended Industrial Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0120] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0121] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0122] The present invention also provides a storage medium. When instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the vehicle speed planning method of the aforementioned embodiment.

[0123] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0124] The algorithm and display provided herein are not inherently related to any particular computer, virtual device or other equipment. According to the above description, it is obvious that the structure required for constructing this type of device is. In addition, the present invention is not directed to any specific programming language. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the description of the above specific language is to disclose the best mode of the present invention.

[0125] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0126] Similarly, it should be understood that in order to streamline the present invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the following intention: that the claimed invention requires more features than those expressly recited in each claim. Rather, as reflected in the claims below, inventive aspects lie in less than all of the features of the individual embodiments disclosed above. Therefore, the claims that follow the detailed description are hereby expressly incorporated into the detailed description, with each claim itself serving as a separate embodiment of the present invention.

[0127] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into a plurality of submodules or subunits or subcomponents. All features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device so disclosed may be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0128] The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the sorting device according to the present invention. The present invention may also be implemented as a device or apparatus program for executing part or all of the methods described herein. Such a program for implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0129] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

[0130] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices, apparatuses and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0131] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0132] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0133] It should be noted that the various data-related processes in the embodiments of the present application are all carried out in compliance with the corresponding data protection laws and policies of the country where the device is located, and with the authorization given by the owner of the corresponding device.

Claims

1. A vehicle speed planning method, characterized in that: The method comprises: Determine the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information; Determine the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; Smoothing the target ST reference line to generate a smoothed ST reference line; The smooth ST reference line is derived to output a target speed planning curve.

2. The method according to claim 1, characterized in that Determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information includes: When there is a first vehicle in front of the target vehicle in the driving lane, and the target vehicle maintains a preset following distance with the second vehicle, the vehicle planning decision label corresponding to the target vehicle is a following vehicle label; When the preset following distance corresponding to the target vehicle is changed, the vehicle planning decision tag corresponding to the target vehicle is a keep distance tag; When there is a third vehicle that changes lanes and cuts into the lane where the target vehicle is traveling within a preset range of the lane where the target vehicle is traveling, the vehicle planning decision label corresponding to the target vehicle is a cut-in label.

3. The method according to claim 2, characterized in that Determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision label is any one of the following label, the keeping distance label and the cutting-in label, obtaining a standard following distance; Determine a target following distance according to a pre-acquired first current scene of the target vehicle and the standard following distance, and construct a following distance line in a pre-set ST diagram for the target following distance; Determine a second current scene of the target vehicle according to the target following distance, and construct a transition reference line corresponding to the target vehicle in a preset ST diagram; A target ST reference line is constructed in a preset ST diagram according to the following distance line and the transition reference line.

4. The method according to claim 3, characterized in that The step of determining a target following distance according to the pre-acquired first current scene of the target vehicle and the standard following distance, and constructing a following distance line in a pre-set ST diagram for the target following distance comprises: When the current scenario is that the target vehicle is in a lane-changing state, obtaining a collision time between the target vehicle and a rear vehicle in a lane adjacent to the lane in which the target vehicle is traveling, and determining a rear vehicle safety level according to the collision time; If the safety level of the rear vehicle is in a dangerous state, calibrate a first bias coefficient according to a speed difference between the target vehicle and a front vehicle in a lane where the target vehicle is traveling; If the rear vehicle safety level is a safe state, calibrating the first bias coefficient according to the time headway intrusion level between the target vehicle and the front vehicle in the lane where the target vehicle is traveling; When the current scene is that the target vehicle is in a lane change end transition state, preliminarily calibrate the transition state time according to the intrusion time distance, obtain a first time offset coefficient according to the speed difference, a second time offset coefficient according to the intrusion rear vehicle time distance level, and obtain a time distance coefficient increment; generating the first bias coefficient according to the first time bias coefficient, the second time bias coefficient and a time interval coefficient increment; When the current scenario is that the target vehicle is in a non-lane-changing state, setting the first bias coefficient to a target value; adjusting the standard following distance according to the first bias coefficient to obtain a target following distance; A following distance line is constructed in a preset ST map for the target following distance.

5. The method according to claim 4, characterized in that Determining the rear vehicle safety level according to the collision time includes: When it is detected that the collision time is less than a first preset value, the safety level of the rear vehicle is in a dangerous state; When it is detected that the collision time is a negative number or infinite, an intrusion time level between the target vehicle and the rear vehicle is obtained, and when the intrusion time level is less than a second preset value, it is determined that the safety level of the rear vehicle is in a dangerous state.

6. The method according to claim 5, characterized in that The obtaining of the intrusion time headway level between the target vehicle and the rear vehicle comprises: The intrusion time-to-distance level is generated according to a pre-calibrated target following distance and a ratio of the distance between the target vehicle and a front vehicle or a rear vehicle in the lane.

7. The method according to claim 3, characterized in that Determining the second current scene of the target vehicle according to the target following distance, and constructing a transition reference line corresponding to the target vehicle in a preset ST diagram comprises: When the second current scene is long-distance non-intrusion, a calibration lookup table is set according to a pre-acquired collision time between the target vehicle and the front vehicle, and a speed bias coefficient lookup table is set; Obtaining a velocity bias coefficient by linear interpolation according to the collision time; generating a reference speed of the target vehicle according to the speed bias coefficient, the current speed of the target vehicle and the current speed of the vehicle ahead; Constructing a first transition reference line corresponding to the target vehicle in a preset ST diagram according to the reference vehicle speed; When the second current scene is a close-range intrusion, obtaining a safety acceleration; Acquiring an acceleration bias according to the safety acceleration; generating a smooth acceleration according to the safety acceleration and the acceleration offset; A second transition reference line corresponding to the target vehicle is constructed in a preset ST diagram according to the smoothed acceleration.

8. The method according to claim 1, characterized in that Determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information includes: When the collected surrounding information shows that there is an obstacle in front of the target vehicle's driving lane, and the lanes on both sides corresponding to the target vehicle's driving lane are in an obstacle-free state, the vehicle planning decision label corresponding to the target vehicle is a deceleration passing label.

9. The method according to claim 8, characterized in that The deceleration passing tag includes a first deceleration passing tag, and the determining of the target ST reference line of the target vehicle corresponding to different vehicle planning decision tags includes: When the vehicle planning decision tag is the first deceleration passing tag, obtaining the deceleration passing time corresponding to the target vehicle; Obtaining a lower boundary distance of a lower boundary of the deceleration time map in a preset ST diagram; Obtaining a ratio of the lower boundary distance to the deceleration passing time to obtain a first constraint curve, wherein the first constraint curve includes a speed; The minimum value of the speed of the constraint curve and the pre-acquired speed limit of the target vehicle driving lane is taken to obtain a target ST reference line corresponding to the target vehicle.

10. The method according to claim 8, characterized in that The deceleration passing tag includes a second deceleration passing tag, and the determining of the target ST reference line of the target vehicle corresponding to different vehicle planning decision tags includes: When the vehicle planning decision label is the second deceleration passing label, obtaining the deceleration passing time corresponding to the target vehicle; Obtaining a lower boundary distance of a lower boundary of the deceleration time mapping in a preset coordinate system; Obtaining a ratio of the lower boundary distance to the deceleration passing time to obtain a first constraint curve, wherein the first constraint curve includes a speed; Taking the minimum value of the speed of the constraint curve and the pre-acquired speed limit of the target vehicle driving lane to obtain a second constraint curve; The intersection point of the pre-acquired following distance line and the second constraint curve is obtained to obtain a target ST reference line corresponding to the target vehicle.

11. The method according to claim 1, characterized in that: Determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information includes: When the collected surrounding information indicates that there is a merging vehicle within a preset range of the target vehicle's driving lane, the vehicle planning decision label corresponding to the target vehicle is a merging label, and the merging label includes a merging deceleration label and a merging acceleration label.

12. The method according to claim 11, characterized in that Determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision tag is the merging deceleration tag, obtaining a predicted speed of the merging vehicle; Searching a pre-acquired safety distance calibration table for a target safety distance between the target vehicle and the merging vehicle corresponding to the predicted speed; The maximum value of the speed corresponding to the target vehicle and the speed corresponding to the merging vehicle is taken to obtain the target ST reference line corresponding to the target vehicle.

13. The method according to claim 11, characterized in that Determining the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels includes: When the vehicle planning decision tag is the merging acceleration tag, determining whether it is a mapping period according to a mapping time in a preset ST diagram; If yes, then the following distance line is obtained in advance, wherein the following distance line is the following distance line corresponding to the target ST reference line generated when the vehicle planning decision label is any one of the following label, the keeping distance label and the cutting label. If no, then the calibration query table is set according to the collision time between the target vehicle and the merging vehicle obtained in advance, and the speed bias coefficient query table is set; Obtaining a velocity bias coefficient by linear interpolation according to the collision time; generating a reference speed of the target vehicle according to the speed bias coefficient, the current speed of the target vehicle and the current speed of the merging vehicle; generating a merging transition reference line according to the reference vehicle speed; A uniform speed line corresponding to the target vehicle is set according to the reference vehicle speed; A target ST reference line corresponding to the target vehicle is constructed according to the uniform speed line and the confluence transition reference line.

14. The method according to claim 1, characterized in that Before the step of determining the vehicle planning decision label corresponding to the target vehicle through the collected surrounding information, the method includes: Draw a rectangular frame of surrounding vehicles according to the predicted trajectory information of the surrounding vehicles of the target vehicle; Obtaining a set of projection distances between each preset point position of the target vehicle and a pre-acquired planned path of the target vehicle; Taking the minimum value in the projection distance set as the target projection distance; If the target projection distance is less than a preset threshold, the surrounding vehicles corresponding to the target projection distance are mapped to the ST map, wherein the ST map includes the mapping time and distance of the surrounding vehicles mapped to the ST map; Obstacle areas are drawn in the ST map according to the time and distance at which the surrounding vehicles are mapped to the ST map.

15. A vehicle speed planning device, characterized in that: The device comprises: A first determination module is used to determine a vehicle planning decision label corresponding to a target vehicle through collected surrounding information; A second determination module is used to determine the target ST reference line of the target vehicle corresponding to different vehicle planning decision labels; A smoothing processing module, used for smoothing the target ST reference line to generate a smoothed ST reference line; The output module is used to derive the smooth ST reference line and output a target speed planning curve.

16. An electronic device, characterized in that: include: at least one processor; a memory communicatively coupled to the at least one processor; The memory has a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the vehicle speed planning method according to any one of claims 1 to 14.

17. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the vehicle speed planning method according to any one of claims 1 to 14 is implemented.

Images