A pilot vehicle control method without vehicle-to-vehicle communication in off-road environment

Abstract

The present invention discloses a method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment, the method comprising the following steps: predicting the future driving trajectory of the pilot vehicle, completing the waypoint information of the predicted future driving trajectory to obtain the predicted trajectory; calculating the front pilot risk assessment parameters of the pilot vehicle and the rear pilot risk assessment parameters of the pilot vehicle based on the predicted trajectory; weightedly calculating the optimized control hyperparameters of the pilot vehicle according to the front pilot risk assessment parameters and the rear pilot risk assessment parameters, and adjusting the driving strategy of the pilot vehicle according to the optimized control hyperparameters; the present invention can perform front and back transformation on the sensor part of the unmanned vehicle, perform path environment perception in front of the vehicle and situation perception of the following vehicle behind the vehicle through the pilot vehicle, perform pilot risk assessment based on the multi-source information perceived, and finally perform adaptive optimization of the driving strategy of the pilot vehicle according to the assessment result to ensure stable following of the following vehicle.

Description

Technical Field

[0001] The present invention relates to the technical field of joint control of vehicle subsystems. Specifically, the present invention is applied to the field of vehicle following control, and in particular to a method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment. Background Art

[0002] At present, in the field of unmanned vehicle control, unmanned vehicle off-road communication-free following technology is widely used.

[0003] Specifically, the unmanned vehicle off-road non-communication following technology refers to: in a complex and changeable off-road environment, the unmanned vehicle can automatically drive and follow the pilot vehicle without relying on vehicle-to-vehicle communication; the above technology requires the following vehicle to accurately identify the pilot vehicle without external network information, and make a fast and accurate following response as the pilot vehicle changes its position, to ensure that the following vehicle can stably follow the route of the pilot vehicle; as shown in the attached Figure 1 As shown in the figure, the relationship between the leading vehicle and the following vehicle is relatively defined. Figure 1 Car No. 1 leads Car No. 2, and Car No. 2 follows Car No. 1. At the same time, Car No. 2 leads Car No. 3, and Car No. 3 follows Car No. 2, and so on. In this relatively defined relationship, there is no communication between cars. The car relies entirely on the sensors it carries to identify the information of the car in front and perform following operations.

[0004] Obviously, the unmanned vehicle off-road non-communication following technology can significantly reduce the vehicle's control requirements and expand the application scope of unmanned vehicles. For example, the technology's unmanned automatic following solution can prevent the driver from being exposed to dangerous environments. At the same time, the technology's non-communication control solution also has stronger concealment and robustness, suitable for use in more demanding environments, so that the following control technology is no longer limited by the level of communication links. In summary, this technology can significantly improve the safety of the fleet and the overall maximum speed, while reducing the high-precision requirements for the vehicle's positioning system.

[0005] However, high-quality technical effects face very high technical requirements and technical difficulties. For example, when the following vehicle loses the pilot vehicle under some complex working conditions, if the pilot vehicle does not take timely measures and drives out of the sensing range of the following vehicle, it will be difficult for the following vehicle to rely on its own vehicle capabilities to follow the pilot vehicle again. Especially under harsh conditions, when the following vehicle cannot be remotely controlled, the entire vehicle will be abandoned, resulting in huge losses. For the aforementioned complex working conditions, such as the attached Figure 2 and 3 As shown, Figure 2 In the first complex working condition, when the leading car turns right, the mountain on the right side of the road blocks the view of the following car's sensor; Figure 3The second complex working condition is an uphill and downhill scenario. In this scenario, vehicle No. 2 cannot sense vehicle No. 1, and vehicle No. 3 cannot sense vehicle No. 2. In the above two scenarios, the following vehicle is limited by its sensor range or blocked by the terrain and cannot sense its leading vehicle.

[0006] In addition, the existing unmanned vehicle non-communication following control technology mainly focuses on the sensor solution or following control method of the following vehicle, and does not study the control optimization method of the pilot vehicle, as follows:

[0007] (1) For example, Chinese patent CN110435654B: A vehicle following method, device and apparatus for an intelligent navigation system. This patent obtains the operating status parameters of the current vehicle and the target vehicle in front to control the current vehicle to follow the vehicle at a preset safety time interval and follow the periodic curve of the vehicle in front. However, this solution needs to be implemented based on communication between vehicles, and the solution mainly adjusts the control strategy of the following vehicle according to the motion parameters of the pilot vehicle.

[0008] (2) For example, Chinese patent CN113870316B: A method for reconstructing the path of a leading vehicle in a non-GPS following scenario. This patent generates a reference path based on the initial relative positions of the leading and trailing vehicles and determines it through cubic curve fitting. It then combines sensor positioning and odometer data, uses an extended Kalman filter algorithm to optimize the estimation of the relative pose, constructs a kinematic model, screens the path of the following vehicle, and generates an effective following path. This solution mainly focuses on the following path generation strategy of the following vehicle, and does not study the control scheme of the leading vehicle.

[0009] (3) For example, Chinese patent CN114037736B: A vehicle detection and tracking method based on an adaptive model. This patent pre-processes point cloud data to remove ground points and clusters, and simultaneously fuses point cloud data at multiple times based on GPS information to create a two-dimensional grid map, identify and mark dynamic objects, screen vehicle candidates, use motion consistency criteria to screen dynamic vehicles, determine their following vehicle relationship, and initialize speed information; finally, a labeled multi-Bernoulli filter is used for tracking; this solution also belongs to the study of control methods for following vehicles.

[0010] In summary, for off-road non-communication following convoys, only optimizing the following vehicle has certain limitations and cannot optimize the overall following performance of the convoy. Further auxiliary control optimization of the pilot vehicle should be performed. Therefore, for off-road non-communication following conditions or convoys, the auxiliary control optimization method for the pilot vehicle is a technical problem that needs to be solved urgently. Summary of the invention

[0011] The object of the present invention is to provide a method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment, thereby solving all or one of the above-mentioned problems existing in the prior art.

[0012] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0013] The present invention provides a method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment, comprising the following steps:

[0014] Driving trajectory prediction:

[0015] Predicting a future driving trajectory of the pilot vehicle according to the driving mode of the pilot vehicle, and completing waypoint information of the predicted future driving trajectory to obtain a predicted trajectory;

[0016] Pilot Risk Assessment:

[0017] Based on the predicted trajectory, calculating a forward pilot risk assessment parameter of the pilot vehicle and a rearward pilot risk assessment parameter of the pilot vehicle;

[0018] Pilot Optimization Control:

[0019] The optimized control hyperparameter of the pilot vehicle is weightedly calculated according to the front pilot risk assessment parameter and the rear pilot risk assessment parameter, and the driving strategy of the pilot vehicle is adjusted according to the optimized control hyperparameter.

[0020] In one embodiment, the driving trajectory prediction step further includes:

[0021] In response to the driving mode of the pilot vehicle being autonomous driving, the future driving path of the pilot vehicle is acquired, and the coordinates of the waypoints of the future driving path are converted to a geodetic coordinate system to obtain the future driving trajectory;

[0022] In response to the driving mode of the pilot vehicle being remote driving or manned driving, predicting the future driving trajectory of the pilot vehicle based on a vehicle kinematics model and a uniform acceleration model;

[0023] The environmental data of the pilot vehicle is collected, and the z coordinate of each waypoint in the future driving trajectory in the autonomous driving and remote driving or manned driving modes is completed based on the elevation information and terrain features in the environmental data to obtain the predicted trajectory.

[0024] In one embodiment, the pilot risk assessment step further comprises:

[0025] Calculating a terrain risk assessment parameter and a drivable distance assessment parameter based on the predicted trajectory as the forward navigation risk assessment parameter;

[0026] Based on the predicted trajectory, a following vehicle confidence risk assessment parameter and a following vehicle motion situation risk assessment parameter are calculated as the rear pilot risk assessment parameter.

[0027] In one embodiment, the step of calculating the terrain risk assessment parameter includes:

[0028] Based on the difference between the waypoint coordinates in the predicted trajectory and the current coordinates of the pilot vehicle, a total road height difference parameter corresponding to the terrain undulation in front of the pilot vehicle is obtained;

[0029] The total road height difference parameter is used as the terrain risk assessment parameter.

[0030] In one embodiment, the step of calculating the drivable distance risk assessment parameter includes:

[0031] Acquire environmental data of the pilot vehicle and identify obstacle position information in the environmental data; determine a number of visible points of the driving trajectory ahead of the pilot vehicle based on the predicted trajectory; determine the occlusion relationship between the visible points and the obstacle position information based on a polygon judgment algorithm; determine the maximum visible point among the visible points according to the occlusion relationship; connect the farthest drivable route of the pilot vehicle based on the maximum visible point and the predicted trajectory; calculate the length of the farthest drivable route to obtain the unobstructed farthest drivable distance of the pilot vehicle;

[0032] The unobstructed maximum drivable distance is used as the drivable distance risk assessment parameter.

[0033] In one embodiment, the step of calculating the following vehicle confidence risk assessment parameter includes:

[0034] Calling the backward perception module of the pilot vehicle to perform target detection on the following vehicle to obtain detection frame information of the following vehicle; extracting the detection confidence in the detection frame information;

[0035] The detection confidence is used as the following vehicle confidence risk assessment parameter.

[0036] In one embodiment, the step of calculating the risk assessment parameter of the following vehicle movement condition includes:

[0037] Invoking the backward sensing module of the pilot vehicle to monitor the motion state of the following vehicle to obtain the speed information of the following vehicle; calculating the deviation angle between the following vehicle and the pilot vehicle based on the earth coordinate system and the speed information of the following vehicle;

[0038] A steady navigation deviation angle threshold is set, and in response to the deviation angle being not less than the steady navigation deviation angle threshold, the deviation angle is used as a risk assessment parameter for the movement of the following vehicle.

[0039] In one embodiment, the step of calculating the risk assessment parameter of the following vehicle movement situation further includes:

[0040] In response to the deviation angle being less than the smooth pilot deviation angle threshold, calculating the relative speed and relative distance between the pilot vehicle and the following vehicle; and calculating the separation time of the following vehicle according to the relative speed and the relative distance;

[0041] The separation time is used as a risk assessment parameter for the movement of the following vehicle.

[0042] In one embodiment, the weighted calculation of the optimization control hyperparameter of the pilot vehicle according to the front pilot risk assessment parameter and the rear pilot risk assessment parameter includes:

[0043] An activation function is used to map the front pilot risk assessment parameter, the rear pilot risk assessment parameter, and threshold parameters corresponding to the front pilot risk assessment parameter and the rear pilot risk assessment parameter;

[0044] The mapped parameters are weighted and calculated to obtain the optimized control hyperparameters.

[0045] In one embodiment, adjusting the driving strategy of the pilot vehicle according to the optimized control hyperparameters includes:

[0046] Multiplying the optimized control hyperparameter by the maximum speed of the pilot vehicle to obtain a desired speed of the pilot vehicle;

[0047] The pilot vehicle is controlled according to the expected vehicle speed, or a warning is given to the driver of the pilot vehicle.

[0048] The beneficial effects of the technical solution of the present invention are:

[0049] 1. The control method of the pilot vehicle without vehicle-to-vehicle communication in an off-road environment described in the present invention can realize the intelligent transformation of the pilot vehicle. In the absence of vehicle-to-vehicle communication, the pilot vehicle can not only perceive the environment and road conditions ahead, but also the following status of the rear vehicle, etc., and evaluate the risk of the rear following vehicle losing the target or the risk of the pilot vehicle failing to navigate based on the perceived multi-source information. According to the different assessed risk levels, the driving state and motion control strategy of the pilot vehicle are adaptively optimized to improve the stable navigation capability of the pilot vehicle during navigation, and at the same time improve the problem handling capability of the pilot vehicle under complex working conditions. Based on the active feedback adjustment of the pilot vehicle itself, the response time and space of the following vehicle are expanded to maximize the following success rate of the following vehicle and reduce the risk of navigation failure.

[0050] 2. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment described in the present invention can enable the pilot vehicle to perceive and analyze the ups and downs of the road ahead and the obstruction of obstacles, and use the perceived data as the basis for optimizing the pilot vehicle's driving strategy, thereby improving the pilot vehicle's navigation capability.

[0051] 3. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment described in the present invention can enable the pilot vehicle to perceive and analyze the position, speed, relative speed, deviation angle and detection confidence of the vehicle following behind, and use the perceived data as the basis for optimizing the pilot vehicle's driving strategy, thereby improving the pilot vehicle's navigation capability. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 It is a schematic diagram of a vehicle following a vehicle without communication in the background technology;

[0054] Figure 2 It is a schematic diagram of the vehicle sensor range under the first complex working condition in the background technology;

[0055] Figure 3 is a schematic diagram of the vehicle sensor range under the second complex working condition in the background technology;

[0056] Figure 4 is a flow chart of a method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0057] Figure 5is a detailed flow chart of the method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0058] Figure 6 is a schematic diagram of the body configuration of an unmanned vehicle in the pilot vehicle control method in an off-road environment without vehicle-to-vehicle communication according to Embodiment 1 of the present invention;

[0059] Figure 7 is a schematic diagram of a geodetic coordinate system in a pilot vehicle control method without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0060] Figure 8 Schematic diagram of the total height difference of the road surface in the method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0061] Fig. 9 1 is a schematic diagram of determining the maximum visual distance in the method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0062] Fig.10 It is a pseudo code schematic diagram for determining the maximum unobstructed drivable distance of the pilot vehicle in the pilot vehicle control method without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0063] Fig.11 Schematic diagram of the position relationship between point C and the convex polygon in the method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0064] Fig.12 is a schematic diagram of the deviation angle of the following vehicle in the method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment according to Embodiment 1 of the present invention;

[0065] The markings in the attached figure are as follows: 1. Pilot vehicle; 2. Follower vehicle. DETAILED DESCRIPTION

[0066] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more definite definition of the protection scope of the present invention.

[0067] In the description of the present invention, it should be noted that the embodiments described in the present invention are only part of the embodiments of the present invention, rather than all of the embodiments; based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention.

[0068] The terms "first", "second", etc. in the specification and claims of this article and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of this article described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0069] Embodiment 1: This embodiment provides a method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment. Figure 1 to Figure 12 As shown, the method includes: remodeling the sensor part of the unmanned vehicle, sensing the path environment in front of the vehicle and the situation of the following vehicle behind the vehicle through the pilot vehicle, conducting pilot risk assessment based on the multi-source information sensed, and finally adaptively optimizing the driving strategy of the pilot vehicle according to the assessment results, further improving the flexible pilot capability of the pilot vehicle under complex working conditions, and ensuring the stable following of the following vehicle. For example, when this method is applied to the two complex working conditions in the background technology, it can make Figure 2 Zhonglinghang vehicles actively reduce speed and steering angle speed before turning. Figure 3 Car No. 1 in the middle actively slows down before going downhill, leaving enough reaction time and space for the following vehicles and improving following stability.

[0070] It should be noted that the focus of this application is to optimize the motion control of the pilot vehicle based on the multi-source perception results of the detection, rather than the conventional technical means in this field such as environmental perception algorithms, vehicle identification and tracking algorithms, etc. Figure 4 and Figure 5 As shown, the steps of this method are specifically as follows:

[0071] S100, pilot vehicle configuration:

[0072] In this step, the pilot vehicle is modified and configured: first, according to the conventional configuration of unmanned vehicle sensors, several laser radars, millimeter wave radars and cameras are configured in front of the pilot vehicle; second, several laser radars, millimeter wave radars and cameras are configured behind the pilot vehicle, and the pilot vehicle body is equipped with an inertial navigation system, and the vehicle's control signals are obtained through bus information; the specific vehicle module configuration is as follows: Figure 6 As shown; Based on the above configuration, the pilot vehicle can predict the route through inertial navigation and vehicle control signals, use multiple sensors in the front to estimate the road conditions ahead, and use multiple sensors in the rear to detect the status of the following vehicle;

[0073] It should be noted that the present application does not impose any limitation on the type and configuration of the front and rear sensors of the pilot vehicle configured above. It is sufficient to adopt a module that can accurately perceive the surrounding environment, including but not limited to configuring a number of visual sensors around the pilot vehicle and reconstructing the surrounding environment through BEV perception.

[0074] S200, coordinate system configuration:

[0075] In this step, a geodetic coordinate system O-XYZ is established, with the X axis pointing due east, the Y axis pointing due north, the Z axis pointing vertically upward, and the origin O being a fixed point on the earth; this coordinate system is made consistent with the coordinate system of the pilot vehicle navigation system to ensure that the control strategy of the pilot vehicle and the reference trajectory provided by the pilot vehicle navigation system are measured by the same coordinate system. The specific coordinate system is as follows: Figure 7 shown.

[0076] S300, driving trajectory prediction:

[0077] In this step, the driving mode of the pilot vehicle is determined, and in response to the pilot vehicle being in an autonomous driving, remote driving or manned driving mode, the future driving trajectory of the pilot vehicle is predicted, and the waypoint information in the future driving trajectory is completed;

[0078] S301, in response to the pilot vehicle being autonomously driven, using the vehicle navigation module to obtain the future driving path of the pilot vehicle, converting the coordinates of each waypoint on the path to the aforementioned geodetic coordinate system to obtain the future driving trajectory Futurepoints of the pilot vehicle; defining the nth waypoint in Futurepoints as W n (W n.x , W n.y , W n.z , W n.v , W n.θ ), which contains the location information, speed information and heading angle information of the waypoint;

[0079] S302: In response to the pilot vehicle being remotely driven or manned, a vehicle kinematics model is used in conjunction with a uniform acceleration model to predict the future driving trajectory of the pilot vehicle, as follows:

[0080] (1) Construct a two-degree-of-freedom kinematic model of the pilot vehicle: Let x and y be the positions of the pilot vehicle in the earth coordinate system, and use the inertial navigation to obtain the speed of the pilot vehicle. , heading angle θ and angular velocity ω; Based on the above parameters, the kinematic equation of the pilot vehicle is as follows:

[0081] ;

[0082] in:

[0083] The derivatives corresponding to the positions x and y, respectively, refer to the velocity components of the pilot vehicle in the X and Y directions;

[0084] is the derivative of the heading angle θ, used to represent the angular velocity ω;

[0085] (2) Based on , the acceleration components of the pilot vehicle in the X and Y directions are obtained as follows:

[0086] ;

[0087] (3) Use the uniform acceleration model to predict the pilot vehicle along the time series k1 to k n The future driving trajectory Futurepoints, the state space of the uniform acceleration model is ;

[0088] in:

[0089] The time interval k is an adjustable parameter;

[0090] State Vector =[x, , y, , ] T ;

[0091] A is the state transfer matrix, which describes the evolution of the state with the interval time k, as follows:

[0092] ;

[0093] B is the control input matrix, as follows:

[0094] ;

[0095] is the noise, as follows:

[0096] ;

[0097] for The covariance matrix of is as follows:

[0098] ;

[0099] (4) After prediction, the future driving trajectory Futurepoints with n predicted waypoints is obtained. The nth waypoint in Futurepoints is W n (W n.x, W n.y , W n.z , W n.v , W n.θ );

[0100] It should be noted that in the Futurepoints when the pilot vehicle is remotely driven or manned, the z coordinate of each waypoint is missing, so it needs to be completed in the subsequent steps;

[0101] It should also be noted that since there is no need to consider the impact of trajectory waypoints that are too far away from the vehicle, the maximum prediction distance is set to 50 meters; when the straight-line distance between the predicted position and the current pilot vehicle position is 50 meters, trajectory prediction is no longer performed;

[0102] It should also be noted that the uniform acceleration model trajectory prediction method is not required to be unique, and other prediction methods can also be used; at the same time, the farthest prediction distance is also adaptively set according to the specific situation.

[0103] S303, updating or completing the z coordinate of each waypoint in the future driving trajectory Futurepoint of the pilot vehicle under all driving conditions, as follows:

[0104] (1) Using environmental perception algorithms, collect and process multi-source environmental data from the pilot vehicle to clarify elevation information and terrain features. For example, a bilinear interpolation method is used to establish a grid terrain model around the pilot vehicle, where each grid cell corresponds to a specific elevation value to form terrain surface data.

[0105] (2) Complete Futurepoints based on the obtained elevation information and terrain features. For example, assuming that the pilot vehicle is driving close to the ground, based on the aforementioned terrain surface data, calculate the time series k1 to k n Complete the z coordinates of the n predicted waypoints of the pilot vehicle in Futurepoints in sequence, and then obtain the complete coordinate value of each predicted waypoint n Futurepoints(W n.x ,W n.y , W n.z , W n.v , W n.θ ), the completed Futurepoints are the final predicted trajectory;

[0106] It should be noted that W n.z It may also be a relative height value between the coordinate point and the pilot vehicle obtained based on the elevation value of the terrain surface, and it does not necessarily have to be an absolute height value in the geodetic coordinate system.

[0107] S400, terrain risk assessment parameter calculation:

[0108] In this step, based on the future driving trajectory Futurepoints, the total road height difference parameter representing the terrain undulation in front of the pilot vehicle is calculated. , using this parameter as a terrain risk assessment parameter to avoid navigation failures caused by complex terrain ahead, as follows:

[0109] (1) Let the z-axis component W of the coordinates of each waypoint in Futurepoints be n.z The z-axis component of the current coordinates of the pilot vehicle Take the difference and get H1, H2, ..., H n ;

[0110] (2) Take H1, H2, …, H n The absolute value of and sum to get the total height difference of the road surface The calculation principle of the total height difference of the road surface is shown as follows: Figure 8 , and its calculation formula is as follows:

[0111] ;

[0112] (3) Set the stable navigation height difference threshold for the total road height difference as ;when Greater than When the pilot vehicle has uphill and downhill slopes in its future driving trajectory, The larger it is, the greater the road undulations in the future driving trajectory will be;

[0113] It should be noted that Make adaptive adjustments based on actual road conditions, driving conditions and vehicle performance.

[0114] S500, calculation of drivable distance evaluation parameters:

[0115] In this step, the maximum unobstructed driving distance of the pilot vehicle is calculated based on the future driving trajectory Futurepoints and the occlusion relationship between the front path of the pilot vehicle and environmental obstacles. , using this parameter as a risk assessment parameter for the drivable distance, thereby avoiding the situation where the pilot fails due to environmental obstacles ahead or the following vehicle loses the pilot target, so that the pilot vehicle is not blocked by obstacles and always stays within the sensor range of the following vehicle, as follows:

[0116] (1) Obtaining obstacle positions: Obtain the perception data of the multi-source sensors on the pilot vehicle, call the obstacle recognition algorithm to process the perception data, and then identify the obstacles in the environment, and describe the corresponding obstacles in the form of convex polygons; if there are j obstacles in total, each obstacle contains a number of vertices, and through the environment perception algorithm and coordinate system conversion, the vertex coordinate values ​​of all obstacles in the geodetic coordinate system can be obtained; the obstacle recognition algorithm is not specifically limited in this method;

[0117] (2) Generate visible distance: Traverse all predicted waypoints in Futurepoints, connect each predicted waypoint with the current coordinate point of the pilot vehicle, and obtain the visible line segments L1, L2, ..., L n ; Using L1~L n Indicates the visible distance from the pilot vehicle to each prediction point;

[0118] (3) Determine the visible point: traverse the visible line segment L1~L n , divide each visible line segment according to the length d1, and obtain the visible line segment L1~L n Several visible points L on n _Judge1, L n _Judge2, …, L n _Judge i ; Among them, d1 is adjustable and can be adaptively adjusted according to the vehicle size and actual driving conditions;

[0119] (4) Determine whether the visible point is blocked by the aforementioned obstacle position: along the visible line segment L1~L n In the order of, the polygon judgment algorithm (Method 1 is used as an example later) is used to judge each visible point L on each visible line segment one by one n _Judge i Is it blocked by an obstacle? If the return value of a visible point is "in" during the traversal process, it is judged that the visible point is blocked by an obstacle, and the traversal is stopped at this time. The traversal principle is shown as follows Fig. 9 As shown;

[0120] (5) Determine the maximum unobstructed driving distance of the pilot vehicle: When the visible point L is detected n _Judge i When blocked by an obstacle, exit the traversal and set the predicted point W corresponding to the visible point (n-1) Defined as the maximum visual point MVP (Maximum Visual Point), the n-1 value is stored in the number of visual points Num; then, the current coordinate of the pilot vehicle P lead (x lead ,y lead) is connected to waypoint W1, and the subsequent waypoints are connected in sequence, and the connection ends when it is connected to the MVP point; finally, the sum of the distances of the connected Num line segments is calculated, and this result is used as the maximum unobstructed drivable distance of the pilot vehicle. , and its calculation formula is as follows:

[0121] ;

[0122] in, It is negatively correlated with the degree of occlusion of the road ahead, that is, The smaller it is, the more serious the obstruction of the road ahead; see Fig.10 , that is, about Pseudo code of the calculation process;

[0123] It should be noted that the aforementioned polygon judgment algorithm is a method for judging whether a point is inside a convex polygon. The obstacle judgment method is not limited to uniqueness. One optional judgment logic (Method 1 mentioned above) is as follows: In the same plane, judge whether a visible point exists inside a convex polygon obstacle composed of m vertices: define a convex polygon obstacle as P1~P m Vertices, a visible point C(x C ,y C ) is connected to every two adjacent vertices of the convex polygon obstacle, and can form Fig. 9 There are m triangles as shown; then, the areas of the m triangles are accumulated and summed, and compared with the area of ​​the convex polygon; if the sum of the areas of the m triangles is greater than the area of ​​the convex polygon, it is determined that the visible point C is outside the convex polygon obstacle; if the sum of the areas of the m triangles is not greater than the area of ​​the convex polygon, it is determined that the visible point C is inside or at the edge of the convex polygon obstacle;

[0124] It should be noted that the judgment formula of the polygon judgment algorithm is as follows:

[0125] A. Calculate the area of ​​a triangle by using the coordinates of its vertices ,by For example, C(x C ,y C ), P g =(X Pg , Y Pg )、P g+1 =(X P(g+1) , Y P(g+1) ),have:

[0126] ;

[0127] B. Calculate the area of ​​convex polygons using the Gaussian area formula ,by For example, P1=(X P1 , Y P1 )、P g =(X Pg , Y Pg )、P m =(X Pm , Y Pm ),have:

[0128] ;

[0129] C. Compare the sum of the triangle area and the area of ​​the convex polygon. represents the area of ​​the corresponding triangle, Representing the area of ​​a convex polygon, we have:

[0130] ;

[0131] D. Determination of the return value Comp: When the sum of the triangle areas is greater than the convex polygon area, the visible point is judged to be outside the convex polygon obstacle, and the return value Comp of the visible point is set to out; when the sum of the triangle areas is less than the convex polygon area, the visible point is judged to be inside the convex polygon obstacle, and the return value Comp of the visible point is set to in.

[0132] S600, Calculation of following vehicle confidence risk assessment parameters:

[0133] In this step, the backward perception module of the pilot vehicle detects the following vehicle and obtains the detection confidence of the following vehicle. , the confidence level is used to measure the accuracy of the perception results, accurately perceive the impact of environmental factors on the following stability (for example, in severe weather conditions such as sandstorms, the confidence level of the perception module will be reduced), and then help the pilot vehicle to make real-time targeted motion adjustments based on the confidence level, thereby improving the pilot success rate, as follows:

[0134] (1) Since the leading vehicle cannot directly obtain the confidence information of the following vehicle under no communication conditions, it is necessary to read the real-time data of the leading vehicle's backward perception module, such as using a detector such as YOLO to monitor the following vehicle in real time and obtain the detection box information Tar along the time series, Tar = {Tar1, ..., Tar G}, and the coordinates of the following vehicle in the geodetic coordinate system P follow (x follow ,y follow );

[0135] In the detection frame information:

[0136] At a certain moment Tar G =(XTar , Y Tar , W Tar , H Tar , C Tar , V xTar , V yTar );

[0137] X Tar is the horizontal coordinate of the center of the bounding box of the detection box in the perception environment;

[0138] Y Tar is the vertical coordinate of the center of the bounding box of the detection box in the perceived environment;

[0139] W Tar is the detection box width;

[0140] H Tar is the detection box height;

[0141] C Tar is the confidence of the detection box;

[0142] V xTar is the speed of the following vehicle along the x-axis;

[0143] V yTar is the speed of the following vehicle along the y-axis;

[0144] (2) When confidence Less than the preset stable pilot detection threshold When the vehicle is detected, it is judged that there may be severe weather such as sandstorm or fog, or there may be emergencies such as the following vehicle being blocked or out of the sensor detection range; It is negatively correlated with the perceived status, i.e. The smaller it is, the worse the perception.

[0145] S700, Calculation of risk assessment parameters for following vehicle movement:

[0146] In this step, the rearward sensing module of the pilot vehicle monitors and calculates the motion state of the following vehicle in real time, and uses the calculated deviation angle of the following vehicle and the relative motion relationship between the following vehicle and the pilot vehicle as risk assessment parameters for the following vehicle motion situation, specifically the deviation angle of the following vehicle and the time when the following vehicle separates Based on the risk assessment parameters of the following vehicle movement, it can avoid the following failure caused by the following vehicle being unable to maintain high speed or unable to drive due to mechanical failure or other unexpected conditions in off-road environments without vehicle communication. The details are as follows:

[0147] S701, deviation angle of following vehicle Calculation:

[0148] (1) Based on the following vehicle V in the aforementioned detection frame information xTar and V yTar , calculate the following vehicle speed , the angle between the extension line of the velocity direction of the following vehicle and the X-axis in the geodetic coordinate system is defined as ;

[0149] (2) Connect the coordinates of the leading vehicle and the following vehicle, and define the angle between the extended line of the connecting line and the X-axis as , the angle is shown as Fig.12 As shown;

[0150] (3) Based on the formula , calculate the deviation angle of the following vehicle ;

[0151] (4) The threshold of the steady pilot deviation angle under the preset pilot state is ,when Greater than When the pilot fails, it is judged that there is a risk of pilot failure; It is positively correlated with the possibility of pilot failure. The larger the value, the higher the possibility of pilot failure. Make adaptive adjustments based on actual road conditions, driving conditions and vehicle performance.

[0152] S702, time to separate from the following vehicle Calculation:

[0153] (1) according to The size of is calculated in different ways, for example, in one case, when Greater than When setting , and order ;in, is the threshold of stable piloting time under the preset piloting state; Make adaptive adjustments based on actual road conditions, driving conditions and vehicle performance;

[0154] (2) In another case, when Less than When the vehicle information of the pilot vehicle and the following vehicle is used, the parameters Calculation is performed to quantify the current relative motion relationship between the following vehicle and the leading vehicle, and further evaluate the loss risk of the following vehicle. The specific calculation steps are as follows:

[0155] (2.1) Through the formula Calculate the pilot vehicle speed Speed ​​of following vehicle The relative speed between ;

[0156] (2.2) In step S500, the known coordinates of the pilot vehicle are P lead (x lead ,y lead ), in step S600, the coordinates of the following vehicle are known to be P follow (x follow ,y follow ), so the relative distance between the leading vehicle and the following vehicle is calculated based on the above known coordinates , the calculation formula is as follows:

[0157] ;

[0158] (2.3) Based on the relative distance between the lead vehicle and the following vehicle and relative speed Calculate the time it takes for the following vehicle to disengage , the calculation formula is as follows:

[0159] ;

[0160] (2.4) Introduce the stable pilot time threshold under the pilot state preset in the previous step ,when Less than When the following vehicle is judged to have insufficient mobility, there is a risk of following failure; the urgency of the situation increases with The value decreases and increases.

[0161] S800, pilot car optimization control hyperparameter calculation:

[0162] In this step, the control optimization hyperparameters of the pilot vehicle for stable driving without communication are calculated based on the five follower vehicle loss risk assessment parameters and three stable following thresholds obtained in steps S400 to S700. ; Based on the calculated pilot vehicle optimization control hyperparameters , combined with the specific driving status, the corresponding control strategy is executed on the pilot vehicle, as follows:

[0163] (1) Use the Sigmoid function to convert the total height difference of the road surface , Steady navigation height difference , predicted driving distance , Detection box confidence , Steady Pilot Detection Threshold , Following vehicle deviation angle , Steady Pilot Deviation Angle , Breakaway time And stable pilot time Map to [1 / 2,1];

[0164] (2) According to the influence of the above parameters on the actual driving conditions, adjust the corresponding weights of the parameters respectively. , the risk assessment parameters are minimized to obtain the optimal control hyperparameters of the pilot vehicle , the calculation formula is as follows:

[0165] ;

[0166] (3) Optimizing control hyperparameters based on the pilot vehicle The control methods include: reducing speed, reducing steering angle speed control or issuing warning information to the driver, such as: Multiply it by the maximum speed of the pilot vehicle to get the expected speed of the pilot vehicle ,Will As vehicle warning information, it is released to the driver, or automatically based on Implement speed limit control for autonomous pilot vehicles.

[0167] In summary, the leading vehicle controlled by this method can perceive the road conditions ahead and the status of the following vehicle, and take timely driving strategies to control the vehicle speed before the following vehicle is lost based on the perceived multi-source information, thereby reserving time and space for the following vehicle to catch up and adjust, improving the success rate of the following vehicle and improving the normal driving stability of the entire off-road communication-free convoy.

[0168] It should be noted that the above examples are only for explaining the present invention and cannot limit the protection scope of the present invention.

[0169] Embodiment 2: This embodiment is based on the same inventive concept as the method for controlling a pilot vehicle without vehicle-to-vehicle communication in an off-road environment described in Embodiment 1, and provides a pilot vehicle control system without vehicle-to-vehicle communication in an off-road environment, including:

[0170] A driving trajectory prediction module is used to: predict the future driving trajectory of the pilot vehicle according to the driving mode of the pilot vehicle, and complete the waypoint information of the predicted future driving trajectory to obtain a predicted trajectory;

[0171] A pilot risk assessment module, used to: calculate a forward pilot risk assessment parameter of the pilot vehicle and a rearward pilot risk assessment parameter of the pilot vehicle based on the predicted trajectory;

[0172] The pilot optimization control module is used to: calculate the optimization control hyperparameters of the pilot vehicle according to the forward pilot risk assessment parameters and the rearward pilot risk assessment parameters by weight, and adjust the driving strategy of the pilot vehicle according to the optimization control hyperparameters.

[0173] Embodiment 3: This embodiment provides a computer-readable storage medium, including:

[0174] The storage medium is used to store computer software instructions used to implement the method for controlling a pilot vehicle without communication between vehicles in an off-road environment described in the above-mentioned embodiment 1, which includes a program for executing the above-mentioned method for controlling a pilot vehicle without communication between vehicles in an off-road environment; specifically, the executable program can be built into the control system for the pilot vehicle without communication between vehicles in an off-road environment described in embodiment 2, so that the control system for the pilot vehicle without communication between vehicles in an off-road environment can implement the method for controlling a pilot vehicle without communication between vehicles in an off-road environment described in embodiment 1 by executing the built-in executable program.

[0175] In addition, the computer-readable storage medium of this embodiment may adopt any combination of one or more computer-readable storage media, wherein the computer-readable storage medium includes electrical, optical, electromagnetic, infrared or semiconductor systems, devices or components, or any combination thereof.

[0176] It should be understood that in the various embodiments of this document, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this document.

[0177] It should also be understood that in the embodiments of this article, the term "and / or" is only a description of the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0178] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this article.

[0179] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0180] In the several embodiments provided herein, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or can be electrical, mechanical or other forms of connection.

[0181] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of this article.

[0182] In addition, each functional unit in each embodiment of this invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional unit.

[0183] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this article is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of this article. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0184] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment, characterized in that: The following steps are involved: Driving trajectory prediction: Predicting a future driving trajectory of the pilot vehicle according to the driving mode of the pilot vehicle, and completing waypoint information of the predicted future driving trajectory to obtain a predicted trajectory; Pilot Risk Assessment: Based on the predicted trajectory, calculating a forward pilot risk assessment parameter of the pilot vehicle and a rearward pilot risk assessment parameter of the pilot vehicle; Pilot Optimization Control: The optimized control hyperparameter of the pilot vehicle is weightedly calculated according to the front pilot risk assessment parameter and the rear pilot risk assessment parameter, and the driving strategy of the pilot vehicle is adjusted according to the optimized control hyperparameter.

2. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 1, characterized in that: The driving trajectory prediction step further comprises: In response to the driving mode of the pilot vehicle being autonomous driving, obtaining a future driving path of the pilot vehicle, converting the coordinates of the waypoints of the future driving path into a geodetic coordinate system, and obtaining a predicted future driving trajectory; In response to the driving mode of the pilot vehicle being remote driving or manned driving, predicting the future driving trajectory of the pilot vehicle based on a vehicle kinematics model and a uniform acceleration model; The environmental data of the pilot vehicle is collected, and the coordinates of each waypoint in the future driving trajectory in the autonomous driving and remote driving or manned driving modes are completed based on the elevation information and terrain features in the environmental data to obtain the predicted trajectory.

3. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 1, characterized in that: The pilot risk assessment step further includes: Based on the predicted trajectory, calculating a terrain risk assessment parameter and a drivable distance risk assessment parameter as the forward pilot risk assessment parameter; Based on the predicted trajectory, a following vehicle confidence risk assessment parameter and a following vehicle motion situation risk assessment parameter are calculated as the rear pilot risk assessment parameter.

4. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 3, characterized in that: The step of calculating the terrain risk assessment parameters comprises: Based on the difference between the waypoint coordinates in the predicted trajectory and the current coordinates of the pilot vehicle, a total road height difference parameter corresponding to the terrain undulation in front of the pilot vehicle is obtained; The total road height difference parameter is used as the terrain risk assessment parameter.

5. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 3, characterized in that: The step of calculating the drivable distance risk assessment parameter comprises: Acquire environmental data of the pilot vehicle and identify obstacle position information in the environmental data; determine a number of visible points of the driving trajectory ahead of the pilot vehicle based on the predicted trajectory; determine the occlusion relationship between the visible points and the obstacle position information based on a polygon judgment algorithm; determine the maximum visible point among the visible points according to the occlusion relationship; connect the farthest drivable route of the pilot vehicle based on the maximum visible point and the predicted trajectory; calculate the length of the farthest drivable route to obtain the unobstructed farthest drivable distance of the pilot vehicle; The unobstructed maximum drivable distance is used as the drivable distance risk assessment parameter.

6. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 3, characterized in that: The calculation step of the following vehicle confidence risk assessment parameter includes: Calling the backward perception module of the pilot vehicle to perform target detection on the following vehicle to obtain detection frame information of the following vehicle; extracting the detection confidence in the detection frame information; The detection confidence is used as the following vehicle confidence risk assessment parameter.

7. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 3, characterized in that: The step of calculating the risk assessment parameter of the following vehicle movement situation includes: Invoking the backward sensing module of the pilot vehicle to monitor the motion state of the following vehicle to obtain the speed information of the following vehicle; calculating the deviation angle between the following vehicle and the pilot vehicle based on the earth coordinate system and the speed information of the following vehicle; A steady navigation deviation angle threshold is set, and in response to the deviation angle being not less than the steady navigation deviation angle threshold, a steady navigation time threshold is set, and the deviation angle and the steady navigation time threshold are used as risk assessment parameters for the following vehicle movement situation.

8. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 7, characterized in that: The step of calculating the risk assessment parameter of the following vehicle movement situation further includes: In response to the deviation angle being less than the smooth pilot deviation angle threshold, calculating the relative speed and relative distance between the pilot vehicle and the following vehicle; and calculating the separation time of the following vehicle according to the relative speed and the relative distance; The deviation angle and the separation time are used as risk assessment parameters for the movement of the following vehicle.

9. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 1, characterized in that: The weighted calculation of the optimization control hyperparameter of the pilot vehicle according to the front pilot risk assessment parameter and the rear pilot risk assessment parameter includes: An activation function is used to map the front pilot risk assessment parameter, the rear pilot risk assessment parameter, and threshold parameters corresponding to the front pilot risk assessment parameter and the rear pilot risk assessment parameter; The mapped parameters are weighted and calculated to obtain the optimized control hyperparameters.

10. The method for controlling a pilot vehicle without inter-vehicle communication in an off-road environment according to claim 1, characterized in that: The adjusting the driving strategy of the pilot vehicle according to the optimized control hyperparameters includes: Multiplying the optimized control hyperparameter by the maximum speed of the pilot vehicle to obtain a desired speed of the pilot vehicle; The pilot vehicle is controlled according to the expected vehicle speed, or a warning is given to the driver of the pilot vehicle.

Images