A Cloud Planning Strategy for the Unmanned System in Open-pit Mining Areas

Abstract

The present invention relates to the technical fields of image processing and environmental modeling and simulation, and provides a cloud planning strategy for an unmanned system in open-pit mining areas. The method includes: obtaining the target positions to be reached by each vehicle of a formation fleet with n vehicles currently and the boundary of the planning area. If the current vehicle i > n, the planning ends; otherwise, through path planning and speed planning, the motion trajectory of the i-th vehicle is obtained. If there is no conflict, it is added to the directed weighted graph; otherwise, the conflict node is used as the boundary condition of the time interval, speed planning is performed on the existing planned path to obtain a trajectory and it is planned. Through evaluation, the optimal path is obtained. The present invention increases the 2D problem of multi-vehicle path planning to the 3D problem of multi-vehicle trajectory planning, and obtains the optimal multi-vehicle collaborative planning result considering both time and space dimensions, solves the heavy dependence of the current unmanned system in mining areas on high-precision maps, and has higher operating efficiency.

Description

Technical Field

[0001] The present invention relates to the technical fields of image processing and environmental modeling and simulation, and in particular to a cloud planning strategy for an unmanned system in an open-pit mining area. Background Art

[0002] Open-pit mine production operations, as a traditional industry, have characteristics such as high labor intensity, high work skill requirements, harsh production environments, and high safety risks for workers. Currently, the operation of mining areas also faces many challenges, such as a shortage of practitioners, rising costs brought about by tightened environmental protection and energy consumption control policies. With the large-scale implementation of 5G technology and the development of unmanned driving and vehicle networking technologies over the years, the advantages of unmanned operation systems in open-pit mining areas, such as safety, low cost, high efficiency, and environmental friendliness, have become increasingly prominent compared to traditional production methods.

[0003] As an important part of the unmanned system technology in open-pit mining areas, the working environment of unmanned transport vehicles is very different from the path planning of unmanned passenger cars on open roads in a structured high-precision map. The roads in the mining area are uneven and often change. With the start of loading and unloading operations at both ends, every movement of the electric shovel / excavator and every unloading operation will cause changes in the working environment of the loading and unloading areas. Unknown obstacles will also appear on the original high-precision map in the road driving area due to engineering construction. Moreover, open-pit mine production operations often involve multi-vehicle formation operations. In the road driving area, when fully loaded vehicle fleets and empty vehicle fleets drive through an intersection or a single-lane area together, there are no traffic lights and lane lines on the mining area roads. At both ends of loading and unloading, multiple vehicles operate together in an open area with irregular boundaries and no structured roads.

[0004] Due to the special working environment of unmanned transport vehicles in the mining area and the unique characteristics of the dynamics of mining trucks, the path planning algorithms and deployment strategies of unmanned systems in the mining area are quite different from those of unmanned passenger cars.

[0005] Chinese Patent Application No. CN201710863755.8, with an application date of September 22, 2017, titled "A Centralized Scheduling Method for Multi-Unmanned Vehicle Formations at Intersections", discloses a centralized scheduling method for unmanned vehicle formations on multi-lanes at intersections. For a collaborative intelligent transportation system with unmanned vehicles as mobile service carriers, this method performs centralized scheduling of multi-unmanned vehicle formations based on vehicle-to-infrastructure communication (V2I) and is deployed in the scheduling computers at each intersection.

[0006] The Chinese patent application "Open Area Multi-Unmanned Vehicle Avoidance Method, Device and Electronic Equipment" with the application number CN2020113456900.3 and the application date of November 25, 2020 discloses an open area multi-vehicle avoidance method, device, storage medium and point device, which belongs to the technical field of multi-unmanned vehicle avoidance. The open area is rasterized. During the process of multiple vehicles passing through the open area, the running trajectory is segmented into a front segment trajectory and the next segmented trajectory. Based on the occupancy of the segmented trajectory in the grid, the current service vehicle and the queue of the fleet to be served in the grid are updated. The priority right of way for multiple vehicles to occupy the desired grid is controlled through the queuing queue. Vehicles with the priority right of way can smoothly switch the trajectory to pass. Vehicles without the priority right of way wait after the current segmented trajectory is completed. The multi-unmanned vehicle avoidance in the open area is realized.

[0007] The Chinese patent application "Parking and Docking Scheme Based on the Integration of Vehicle, Electric Shovel and Cloud for Unmanned Transportation in Mining Areas" with the application number CN202110581107.X and the application date of May 27, 2021 discloses a parking and docking method based on the integration of vehicle, electric shovel and cloud for unmanned transportation in mining areas. After the mining vehicle enters the loading area, it sends an entry application to the electric shovel. The electric shovel formulates the docking position and docking direction and sends them to the cloud platform. The cloud platform plans the forward path and the reverse path according to the collected boundary information, main path information and docking position direction, and sends them to the mining vehicle through the cloud method. The mining vehicle parks and docks according to the issued trajectory, and feeds back the status to the electric shovel for loading after the docking is completed. Summary of the Invention

[0008] In view of this, the present invention provides a cloud planning strategy for an unmanned system in open-pit mining areas to solve the problems in the prior art that the road driving area has a strong dependence on high-precision maps, the frequently changing road environment requires engineering personnel to update the maps frequently, the operation cost of the unmanned system in mining areas is very high, the spatial characteristics of the open area in the loading / unloading area are not fully utilized, the wide area around the formation vehicle is not utilized, and the operation efficiency of the fleet is relatively low.

[0009] The present invention provides a cloud planning strategy for an unmanned system in open-pit mining areas, including:

[0010] Step S1: Obtain the target position that each vehicle of the current formation fleet with n vehicles needs to reach and the boundary of the planning area;

[0011] Step S2: When i > n, the planning ends; otherwise, go to step S3, where i is the number of the current vehicle;

[0012] Step S3: Call the path planner to plan the path path_i for the i-th vehicle to its target position in the order of the vehicle numbers before and after;

[0013] Step S4, based on the path planning path_i, calls a speed planner to perform speed planning for the i-th vehicle on path_i according to boundary conditions, and obtains the motion trajectory traj_i of the i-th vehicle;

[0014] Step S5 calls a graph generator to discretize the trajectory traj_i into a sequence of 2D spatial nodes graph_i at fixed time intervals;

[0015] Step S6, when i == 1, proceeds to Step S7, otherwise proceeds to Step S8;

[0016] Step S7 adds the sequence of 2D spatial nodes graph_i to the directed weighted graph graph. The i-th vehicle's planning is successful, then i = i + 1, and returns to Step 2;

[0017] Step S8 checks whether there are node conflicts between the sequence of 2D spatial nodes graph_i and the current directed weighted graph graph. If there is no such node conflict, returns to Step S7, otherwise proceeds to Step S9;

[0018] Step S9 calls the speed planner, uses the conflicting nodes as boundary conditions for the time interval, and performs speed planning on the existing path planning to obtain the trajectory motion traj_i1;

[0019] Step S10 calls the path planner, uses the conflicting node area as an obstacle, dynamically corrects path_i to obtain path_i2, and performs speed planning for the i-th vehicle on path_i2 with the path centripetal acceleration limit and the maximum speed limit of the high-precision map as boundary conditions, and obtains the motion trajectory traj_i2 of the i-th vehicle;

[0020] Step S11 evaluates traj_i1 and the motion trajectory traj_i2 of the i-th vehicle. If cost(traj_i1) > cost(traj_i2), where cost() is a cost function, then proceeds to Step S12; otherwise proceeds to Step S13;

[0021] Step S12, if traj_i = traj_i2, returns to Step S7;

[0022] Step S13, if traj_i = traj_i1, returns to Step S7.

[0023] Further, the target position in Step S1 includes the end of the road right-of-way in the road driving area, the designated loading point of the electric shovel / excavator in the loading area, or the crushing station / dumping position in the unloading area. The planning area boundary includes the left and right road boundaries in the road driving area or the physical boundaries of the working areas at both ends of loading and unloading.

[0024] Further, the step S3 includes:

[0025] Step S31 determines a grid map of the minimum bounding box at the boundary of the planning area and the conflict node area from the current position of the vehicle to the target position;

[0026] Step S32 uses the Hybrid A* algorithm to heuristically plan a rough planned path from the current position of the vehicle to the target position with Dubins / ReedsShepp Curve;

[0027] Step S33 performs quadratic programming based on the rough planned path, with the weighted algebraic sum of the smoothing term, the path length term, the distance to the reference path term, and the closest distance to the obstacle term as the optimization objective function, and the restriction of the distance to the reference path and the maximum curvature limit of the vehicle kinematics as the constraint conditions, to obtain the optimized planned path path_i.

[0028] Further, the obtaining of the optimized planned path in step S33 includes:

[0029] Taking minf = w 1 cost1 + w 2 ost2 + w 3 cost3 + w 4 cost4 as the first objective function, where f is the objective function and minf is the minimum value of the objective function;

[0030]

[0031] Among them, x i and y i represent the abscissa and ordinate of the i-th path point to be optimized, x i-ref and y i-ref respectively represent the differences between the abscissa and ordinate of the i-th path point to be optimized and the abscissa x ref and ordinate y ref of the i-th reference path point; x i-collision , y i-collison respectively represent the differences between the abscissa and ordinate of the i-th path point to be optimized and the abscissa x collison , ordinate y collison of the obstacle point closest to the i-th path point to be optimized;

[0032] The following formula is used as the constraint condition:

[0033]

[0034] Among them, x l represents the lower bound of the abscissa of the optimization variable, x u represents the upper bound of the ordinate of the optimization variable, y lRepresents the lower bound of the ordinate of the optimization variable, y u Represents the upper bound of the ordinate of the optimization variable, stack i Represents the slack variable, Cur cstr Represents the maximum curvature;

[0035] Optimize the rough path planning to obtain a smoother and safer optimized planning path path_i, where cost1 is the smoothing term, cost2 is the path length term, cost3 is the term related to the reference path distance, and cost4 is the term of the closest distance to the obstacle.

[0036] Furthermore, the boundary conditions in step S4 include the path centripetal acceleration limit and the highest speed limit of the high-precision map.

[0037] Furthermore, step S4 includes:

[0038] Step S41 obtains the speed limit corresponding to the maximum centripetal acceleration of the vehicle according to the curvature k at the path point;

[0039]

[0040] Then the maximum speed limit at the path point

[0041] where a n is the tangential acceleration, is the longitudinal maximum speed limit at the path point,

[0042] Step S42 takes the motion trajectory traj_i of the i-th vehicle as the objective function, and the calculation formula is as follows:

[0043]

[0044] where L represents the objective function of speed planning, and j i+1 represents the jerk of the position of the (i + 1)-th optimized path point.

[0045] The boundary conditions are

[0046]

[0047] The variables to be optimized in the model are (v, a, j); where v i is the actual vehicle speed, a i is the acceleration, j i represents the acceleration change rate Jerk, i represents the path point index, is the desired vehicle speed, v i+1 represents the actual vehicle speed of the (i + 1)-th optimized path point, t i represents the time required to travel from the i-th optimized path point to the (i + 1)-th path point, ai+1 represents the actual acceleration of the (i + 1)-th path point;

[0048] represents the maximum speed of each path point (currently depending on centripetal acceleration, longitudinal maximum speed, and expected maximum mission vehicle speed);

[0049] and respectively represent the minimum and maximum vehicle accelerations at time t, depending on the driving force and braking force of the vehicle itself, taking (-1.0 m / s i -1.0 m / s 2 ); 2 )

[0050] and represent the minimum and maximum vehicle acceleration change rates at time t, currently tentatively (-3.0 m / s i -1.0 m / s 3 ); 3 )

[0051] represent the minimum and maximum vehicle position boundaries on the path at time t, i and the determination of is divided into two cases. When the current vehicle planning trajectory will collide with the movement trajectories of other existing vehicles in the time period t ∈ [t i , t i+1 , t i+1 .

[0052] The beneficial effects of the present invention compared with the prior art are as follows:

[0053] 1. Increase the 2D problem of multi-vehicle path planning to the 3D problem of multi-vehicle trajectory planning, and obtain the optimal multi-vehicle collaborative planning result considering both the time dimension and the space dimension, solve the problem that the multi-vehicle collaboration in the current mine unmanned system depends on high-precision maps and the utilization of wide road driving areas is insufficient, and enable multi-vehicles to have higher operation efficiency when running in the same road driving area.

[0054] 2. The present invention proposes a path planning method that can efficiently obtain a smooth, safe, and followable vehicle driving planning path from the current vehicle position to the target position in an open space without relying on a reference path, can solve the heavy dependence of the current mine unmanned system on high-precision maps, reduce the workload of on-site engineering personnel for frequently collecting reference paths, and reduce the operation cost of the mine unmanned system.

[0055] 3. The present invention provides a trajectory planning method, which, while considering the operation efficiency of a vehicle on a path, takes into account fuel consumption savings and trajectory collision safety, solves the problem of damage to key vehicle components caused by emergency braking and frequent acceleration / braking due to the non-smooth expected speed sequence followed by the control module in the current driverless system, and at the same time enables a smoother and safer following process during the operation of a vehicle fleet, reduces the collision risk during the simultaneous operation of multiple vehicles, saves fuel, and reduces the operation cost of the unmanned system in the mining area. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] To more clearly illustrate the technical solutions in the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following-described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0057] Figure 1 is a flowchart of a cloud planning strategy for an unmanned system for open-pit mining areas provided by the present invention.

[0058] Figure 2 is a schematic diagram showing that the movement trajectory of vehicle i does not collide with the movement trajectory of vehicle i - 1 provided by the present invention;

[0059] Figure 3 is a schematic diagram of a position space detour path in the case where the movement trajectory of vehicle i collides with the movement trajectory of vehicle i - 1 provided by the present invention;

[0060] Figure 4a is a schematic diagram of Hybrid A* + Dubins Curve path planning provided by the present invention;

[0061] Figure 4b is a schematic diagram of another Hybrid A* + ReedsShepp Curve path planning provided by the present invention;

[0062] Figure 5 is a schematic diagram of the curves of ReedsShepp Curve before and after optimization provided by the present invention;

[0063] Figure 6 is a schematic diagram showing that the planned trajectory of the current vehicle collides with the movement trajectories of other existing vehicles provided by the present invention at t ∈ [t i , t i+1 ;

[0064] Figure 7 is a schematic diagram showing that the planned trajectory of the current vehicle collides with the trajectory of an existing vehicle provided by the present invention at t ∈ [t i , t i+1 ​ Determined schematic diagram;

[0065] Figure 8 It is a schematic diagram of the test path provided by the present invention;

[0066] Figure 9 It is a schematic diagram of the curvature of the test path provided by the present invention;

[0067] Figure 10 It is a schematic diagram of the planned effect of the desired speed curve provided by the present invention;

[0068] Figure 11 It is a schematic diagram of the local enlarged detail view of the desired speed curve provided by the present invention. Detailed implementation manners

[0069] In the following description, for the purpose of illustration rather than limitation, specific details such as specific system structures and technologies are presented in order to thoroughly understand the embodiments of the present invention. However, those skilled in the art should clearly understand that the present invention can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present invention.

[0070] Next, a cloud planning strategy for an unmanned system for open-pit mining areas according to the present invention will be described in detail with reference to the accompanying drawings.

[0071] Figure 1 It is a flowchart of a cloud planning strategy for an unmanned system for open-pit mining areas provided by the present invention. As Figure 1 shown, the cloud planning strategy includes:

[0072] Step S1: Obtain the target position that each vehicle in the current formation fleet with n vehicles needs to reach and the boundary of the planning area;

[0073] In step S1, the target position includes the end of the road right-of-way in the road driving area, the designated loading point of the electric shovel / excavator in the loading area, or the crusher / dump position in the unloading area, and the boundary of the planning area includes the left and right road boundaries in the road driving area or the physical boundaries of the working areas at both ends of the loading and unloading.

[0074] Step S2: When i > n, the planning ends; otherwise, proceed to step S3, where i is the current vehicle;

[0075] Step S3: Call the path planner to perform path planning path_i for the i-th vehicle to its target point in the order of the vehicle formation;

[0076] Step S31: Determine the grid map of the minimum bounding box in the conflict node area at the boundary of the planning area from the current position of the vehicle to the end position;

[0077] Step S32 uses the Hybrid A* algorithm to heuristically plan a rough planned path from the current position of the vehicle to the target position with Dubins / ReedsShepp Curve;

[0078] Figure 4a is a schematic diagram of the Hybrid A*+Dubins Curve path planning provided by the present invention, Figure 4b is another schematic diagram of the Hybrid A*+ReedsShepp Curve path planning provided by the present invention.

[0079] Figure 5 is a schematic diagram of the ReedsShepp Curve before and after optimization provided by the present invention.

[0080] Based on the rough planned path of the target position, with the weighted algebraic sum of the smoothing term, the path length term, the distance term from the reference path and the closest distance term to the obstacle as the optimization objective function, and with the restriction of the distance from the reference path and the maximum curvature limit of the vehicle kinematics as the constraint conditions, quadratic programming is performed to obtain the optimized planned path.

[0081] The obtaining of the optimized planned path in step S33 includes:

[0082] Taking minf = w 1 cost1 + w 2 ost2 + w 3 cost3 + w 4 cost4 as the first objective function, where the first objective function represents the optimized planned path path_i,

[0083]

[0084] Constraint conditions:

[0085] The quadratic programming model of is to optimize the path planning to obtain a smoother and safer optimized planned path path_i.

[0086] Among them, cost1 is the smoothing term, cost2 is the path length term, cost3 is the distance term from the reference path, cost4 is the distance term from the closest obstacle, and the boundary conditions are the 4 constraint conditions.

[0087] Based on the path planning path_i, call the speed planner, and according to the boundary conditions, perform speed planning for the i-th vehicle on path_i to obtain the motion trajectory traj_i of the i-th vehicle;

[0088] In step S4, the boundary conditions include path centripetal acceleration limit and the maximum speed limit of the high-precision map.

[0089] In step S41, the speed limit corresponding to the maximum centripetal acceleration of the vehicle is obtained according to the curvature k at the path point;

[0090]

[0091] Then the maximum speed limit at the path point

[0092] where a n is the tangential acceleration, is the longitudinal maximum speed limit at the path point,

[0093] The calculation formula of the second objective function in step S42 is as follows:

[0094]

[0095] The boundary conditions are

[0096]

[0097] The variables to be optimized in the model are (v, a, j); where v i is the actual vehicle speed, a i is the acceleration, j i represents the acceleration change rate Jerk, i represents the path point index, is the desired vehicle speed,

[0098] represents the maximum speed of each path point (currently depending on the centripetal acceleration, longitudinal maximum speed, and desired maximum mission vehicle speed);

[0099] and respectively represent the minimum and maximum vehicle accelerations at time t i , depending on the driving force and braking force of the vehicle itself, taking (-1.0m / s 2 -1.0m / s 2 );

[0100] and represent the minimum and maximum vehicle acceleration change rates at time t i , currently tentatively (-3.0m / s 3 -1.0m / s 3 );

[0101] Figure 2 is a schematic diagram showing that the movement trajectory of vehicle i provided by the present invention does not collide with the movement trajectory of vehicle i-1.

[0102] Figure 3 It is a schematic diagram of the position space detour path in the case where the motion trajectory of vehicle i provided by the present invention collides with the motion trajectory of vehicle i-1.

[0103] Figure 6 It is a schematic diagram of the collision between the planned trajectory of the current vehicle provided by the present invention and the motion trajectories of other existing vehicles when t ∈ [t i , t i+1 .

[0104] Figure 7 It is a schematic diagram of the collision between the planned trajectory of the current vehicle and the trajectory of the existing vehicle provided by the present invention when t ∈ [t i , t i+1 . Determined schematic diagram.

[0105] Represents the position boundaries of the minimum and maximum vehicles on the path at time t i . The determination of is divided into two cases where the planned trajectory of the current vehicle will collide with the motion trajectories of other existing vehicles when t ∈ [t i , t i+1 .

[0106] Step S5 calls the graph generator to discretize the trajectory traj_i into a sequence of multiple 2D space nodes graph_i at fixed time intervals;

[0107] In step S6, when i == 1, step S7 is performed; otherwise, step S8 is performed.

[0108] In step S7, the 2D space node sequence graph_i is added to the directed weighted graph graph. The planning of the i-th vehicle is successful, then i = i + 1, and it returns to step 2;

[0109] In step S8, it is checked whether there are node conflicts between the 2D space node sequence graph_i and the current directed weighted graph graph. If there are no node conflicts, it returns to step S7; otherwise, step S9 is performed.

[0110] Figure 8 It is a schematic diagram of the test path provided by the present invention;

[0111] Figure 9 It is a schematic diagram of the curvature of the test path provided by the present invention;

[0112] Figure 10 It is a schematic diagram of the planning effect of the desired speed curve provided by the present invention;

[0113] Figure 11It is a schematic diagram of a partial enlarged detail view of the expected speed curve provided by the present invention.

[0114] Step S9 calls the speed planner, uses the conflict node as the boundary condition of the time interval, and performs speed planning on the existing planned path to obtain the trajectory traj_i1;

[0115] Step S10 calls the path planner, uses the conflict node area as an obstacle, dynamically corrects path_i to obtain path_i2, and performs speed planning on the i-th vehicle on path_i2 according to the path centripetal acceleration limit, the highest speed limit of the high-precision map, etc. as boundary conditions to obtain the motion trajectory traj_i2 of the i-th vehicle;

[0116] Step S11 evaluates traj_i1 and the motion trajectory traj_i2 of the i-th vehicle. If cost(traj_i1)>cost(traj_i2), then proceed to step S12; otherwise, proceed to step S13;

[0117] If in step S12 traj_i = traj_i2, return to step S7;

[0118] If in step S13 traj_i = traj_i1, return to step S7.

[0119] All of the above optional technical solutions can be combined arbitrarily to form optional embodiments of the present application, which will not be elaborated herein one by one.

[0120] The following is an embodiment of the device of the present invention, which can be used to execute the method embodiment of the present invention. For the details not disclosed in the embodiment of the device of the present invention, please refer to the method embodiment of the present invention.

[0121] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0122] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A cloud planning strategy for an unmanned system in open-pit mining areas, characterized in that, it includes: Step S1: Obtain the target position that each vehicle in the current formation fleet with n vehicles needs to reach and the boundary of the planning area; Step S2: When i > n, the planning ends, otherwise go to Step S3, where i is the current vehicle number; Step S3: Call the path planner to perform path planning path_i for the i-th vehicle to its target position in the order of the vehicle numbers before and after; Step S4: Based on the path planning path_i, call the speed planner to perform speed planning for the i-th vehicle on path_i according to the boundary conditions to obtain the motion trajectory traj_i of the i-th vehicle; Step S5: Call the graph generator to discretize the trajectory traj_i into a sequence of 2D space nodes graph_i at fixed time intervals; Step S6: When i == 1, go to Step S7, otherwise go to Step S8; Step S7: Add the 2D space node sequence graph_i to the directed weighted graph graph. The i-th vehicle's planning is successful, then i = i + 1, and return to Step 2; Step S8: Check whether there are node conflicts between the 2D space node sequence graph_i and the current directed weighted graph graph. If there are no node conflicts, return to Step S7, otherwise go to Step S9; Step S9: Call the speed planner, use the conflict node as the boundary condition of the time interval, and perform speed planning on the existing path planning to obtain the trajectory motion traj_i1; Step S10: Call the path planner, use the conflict node area as an obstacle, dynamically correct the path_i to obtain path_i2, and perform speed planning for the i-th vehicle on path_i2 with the path centripetal acceleration limit and the maximum speed limit of the high-precision map as the boundary conditions to obtain the motion trajectory traj_i2 of the i-th vehicle; Step S11: Evaluate traj_i1 and the motion trajectory traj_i2 of the i-th vehicle. If cost(traj_i1) > cost(traj_i2), where cost() is the cost function, then go to Step S12; otherwise go to Step S13; Step S12: If traj_i = traj_i2, return to Step S7; Step S13: If traj_i = traj_i1, return to Step S7.

2. The cloud planning strategy according to claim 1, characterized in that, in Step S1, the target position includes the end of the road right-of-way in the road driving area, the designated loading point of the electric shovel / excavator in the loading area, or the crusher / dump position in the unloading area, and the boundary of the planning area includes the left and right road boundaries in the road driving area or the physical boundaries of the working areas at both ends of loading and unloading.

3. The cloud planning strategy according to claim 1, characterized in that, Step S3 includes: S31: Determine the grid map of the minimum bounding box at the boundary of the planning area and the conflict node area from the current position of the vehicle to the target position; S32 Use the Hybrid A* algorithm to heuristically plan a rough planned path from the current position of the vehicle to the target position using Dubins / ReedsShepp Curve; S33 Based on the rough planned path, use the weighted algebraic sum of the smoothing term, path length term, distance from the reference path term, and closest distance to the obstacle term as the optimization objective function, and use the restriction of the distance from the reference path and the maximum curvature limit of vehicle kinematics as the constraint conditions to perform quadratic programming to obtain the optimized planned path path_i.

4. The cloud planning strategy according to claim 3, characterized in that the obtaining of the optimized planned path in step S33 includes: Let minf = w 1 cost1 + w 2 cost2 + w 3 cost3 + w 4 cost4 be the objective function, where f is the objective function, minf is the minimum value of the objective function, and w 1 , w 2 , w 3 , w 4 respectively represent the weights of the four optimization terms cost1, cost2, cost3, and cost4 in the objective function, representing the importance of each term in the objective function; Among them, x i and y i represent the abscissa and ordinate of the i-th path point to be optimized. x i-ref and y i-ref respectively represent the differences between the abscissa and ordinate of the i-th path point to be optimized and the abscissa x ref and ordinate y ref of the i-th reference path point; x i-collision , y i-collison respectively represent the differences between the abscissa and ordinate of the i-th path point to be optimized and the abscissa x collison and ordinate y collison of the obstacle point closest to the i-th path point to be optimized; using the following formula as the constraint condition: where x l represents the lower bound of the abscissa of the optimization variable, x u represents the upper bound of the ordinate of the optimization variable, y l represents the lower bound of the ordinate of the optimization variable, y u represents the upper bound of the ordinate of the optimization variable, stack i represents the slack variable, Cur cstr represents the maximum curvature; optimize the rough planned path to obtain a smoother and safer optimized planned path path_i, where cost1 is the smoothing term, cost2 is the path length term, cost3 is the distance from the reference path term, and cost4 is the closest distance to the obstacle term.

5. The cloud planning strategy according to claim 1, characterized in that the boundary conditions in step S4 include the path centripetal acceleration limit and the highest speed limit of the high-precision map.

6. The cloud planning strategy according to claim 1, characterized in that step S4 includes: step S41 Obtain the speed limit corresponding to the maximum centripetal acceleration of the vehicle according to the curvature k at the path point; Then the maximum speed limit on the path point where a n is the tangential acceleration, is the longitudinal maximum speed limit at the path point, step S42 Use the motion trajectory traj_i of the i-th vehicle as the objective function, and the calculation formula is as follows: Among them, L represents the objective function of speed planning, and j i+1 indicates that the jerk boundary condition for the position of the (i + 1)-th optimized path point is The variables to be optimized in the model are (v, a, j); where v i is the actual vehicle speed, a i is the acceleration, j i represents the acceleration change rate, i represents the path point index, is the desired vehicle speed, v i+1 represents the actual vehicle speed at the (i + 1)-th optimized path point, t i represents the time required to travel from the i-th optimized path point to the (i + 1)-th path point, a i+1 represents the actual acceleration at the (i + 1)-th path point; and respectively represent the minimum and maximum vehicle accelerations at time t i instant; and represent the minimum and maximum vehicle acceleration change rates at time t i respectively; represents the minimum and maximum vehicle position boundaries on the path at time t i ​

Images