An unmanned loader travel path planning method

Abstract

The present application is suitable for the technical field of loaders, and provides a driving path planning method for unmanned loaders. The present application can realize the unmanned driving function of articulated loaders, and thus can eliminate the dependence on drivers, thereby significantly improving the working efficiency of articulated loaders and effectively reducing the incidence of driving accidents. The present application adopts a multi-objective path planning strategy under constraints compared with the traditional single-objective shortest path planning method, aiming at the characteristics of complex structure and harsh working environment of articulated loaders. This innovation can better adapt to the needs of articulated loaders in harsh working environments, and improve the practicality and adaptability of path planning. The present application introduces a learning function and successfully constructs an adaptive Kriging Monte Carlo dynamics proxy model. The introduction of this model effectively solves the problem of long simulation calculation time, making the acquisition of vehicle characteristic parameters more rapid and efficient.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of loaders, and particularly relates to a driving path planning method for unmanned loaders. BACKGROUND

[0002] With the rapid development of science and technology, the intelligentization and unmanned trend of the engineering machinery field are increasingly evident. As a kind of efficient engineering operation equipment, the loader is widely used in non-structural environments such as construction sites and mines. The unmanned loader can realize the automatic transportation of materials and uninterrupted operation throughout the day, but in the non-structural environment, due to the uneven road surface, the loader may have a large attitude change during driving, and even face the risk of tipping. Therefore, planning a safe, efficient and executable path according to the non-structural environment map is the basis for ensuring the unmanned driving and autonomous operation of the loader.

[0003] In view of this challenge, the path planning of the loader automatic driving is particularly important. The traditional method usually simplifies the three-dimensional map into a two-dimensional map, ignoring the actual influence of the ground height fluctuation on the driving of the loader, and then requires the driving road surface to be a structured environment. For the structured terrain, due to its flat terrain and regular boundary, the 2D occupancy grid can effectively plan the optimal path, and the generated path only needs to consider avoiding obstacles. But for the unstructured environment, its boundary may be curved or incomplete, the ground is uneven, and there are a large number of obstacles with different sizes and shapes, plus the consideration of the instability of the terrain caused by the soil composition and the influence of bad weather, which makes the path planning method in the structured environment cannot be directly applied to the unstructured environment.

[0004] At present, the commonly used path planning methods include artificial potential field (APF), search-based method, sampling-based method, etc. But they still have limitations when facing the complexity of the unstructured environment. Especially in the case of large terrain fluctuation, the loader is easy to tip due to instability. In view of this, the present application proposes a driving path planning method for unmanned loaders. SUMMARY

[0005] The purpose of the present application is to provide a driving path planning method for unmanned loaders, which aims to solve the problems raised in the background.

[0006] The purpose of the present application is achieved by the following technical solutions:

[0007] A driving path planning method for unmanned loaders, comprising the following steps:

[0008] Step S1, a dynamic adaptive model is constructed to provide the articulated loader dynamics characteristic data for subsequent path planning;

[0009] Step S2, establishing a working environment: simplifying the vehicle working environment into a three-dimensional map to determine the initial position and target position of the path;

[0010] Step S3, constructing a target function and determining a constraint condition;

[0011] Step S4, outputting an optimal path by using a planner: extracting the top 10 non-dominated solutions after iteration of the multi-objective genetic algorithm, and then comparing and selecting an optimal solution as the optimal path output by the algorithm after comprehensive consideration;

[0012] Step S5, checking whether the output optimal path is reasonable by joint simulation; if yes, outputting the path, and if not, returning to step S4 for further optimization.

[0013] Further, the specific process of step S1 is as follows:

[0014] Based on the Kriging model and Monte Carlo simulation, a learning function is introduced to construct an adaptive Kriging Monte Carlo dynamic proxy model; in the dynamic proxy model, the learning function index U(x) is associated with the lower confidence boundary lcb function for optimization, and the lcb is defined as:

[0015]

[0016] In the formula, lcb(x) represents the lower limit function of the confidence degree, represents the model prediction mean, represents the model prediction deviation; taking b as 2, calculating and minimizing lcb as 0, at this time U(x) is equal to the value of b, corresponding to the model error probability Φ(-2) = 0.023, after training, the dynamic proxy model provides the articulated loader dynamic characteristic data for path planning.

[0017] Further, the specific process of step S2 is as follows:

[0018] The three-dimensional environment data is simplified into coordinate values x, y and z, and the coordinate values are stored in the map environment matrix; the design variables are the coordinates (x i ,y i ,z i ) of each path point, the starting point S coordinate is set as (x1, y1, z1), and the terminal point G coordinate is set as (x n ,y n ,z n ); the map is evenly divided into a grid in the x-y plane, and the remaining trajectory path points between the starting point and the terminal point are discretely distributed at the grid vertices, represented as (x i ,y i ,z i ); finally, the points are connected in order to form a complete trajectory.

[0019] Further, the specific process of step S3 is as follows:

[0020] Step S31, constructing a target function;

[0021] The spatial distance sum between adjacent path points is used as the path length cost, that is, the path driving length cost function F L is defined as:

[0022]

[0023] In the formula, n represents the number of path segments in the path, l i represents the spatial distance between the two path points at the ends of the i-th path segment, and is specifically represented as:

[0024]

[0025] The total driving time of each path segment is used as the path time cost, that is, the path driving time cost function F T is defined as:

[0026]

[0027] In the formula, n represents the number of path points contained in a path, t i represents the driving time between the two adjacent path points in the i-th path segment, and is specifically represented as:

[0028]

[0029] In the formula, l i represents the spatial distance between the two adjacent path points in the i-th path segment, v i and v i-1 represent the safety speed values calculated by the dynamic agent model at the two path points;

[0030] The smoothness cost F of a path is measured by evaluating the change in angle between adjacent straight line segments on each path: S

[0031]

[0032] In the formula, n represents the number of path points contained in a path, and are two adjacent vectors formed by q i-2 , q i-1 , q i

[0033]

[0034] In the formula, x​​i x-coordinate of the i-th path segment, y i y-coordinate of the i-th path segment, z i z-coordinate of the i-th path segment, x i-1 x-coordinate of the i-1-th path segment, y i-1 y-coordinate of the i-1-th path segment, z i-1 z-coordinate of the i-1-th path segment, x i-2 x-coordinate of the i-2-th path segment, y i-2 y-coordinate of the i-2-th path segment, z i-2 z-coordinate of the i-2-th path segment;

[0035] Step S32, determining a constraint condition;

[0036] The constraint condition includes a maximum articulation angle constraint, a maximum pitch angle constraint and a maximum roll angle constraint, the articulation angle constraint is ±30°, the pitch angle constraint is ±30°, and the roll angle constraint is ±20°;

[0037] The ground contact point of the path point q i (x i ,y i ,z i ) is q il (x il ,y il ,z il ), the obstacle crossing point is q ir (x ir ,y ir ,z ir ), and the roll angle φ i is:

[0038]

[0039] The maximum roll angle constraint is:

[0040] φ i ≤φ max

[0041] In the formula, φ ma x is the maximum roll angle, and L is the vehicle body width of the articulated loader;

[0042] Suppose that the angle between the path segment where the path point q i is located and the x-axis in the x-y plane is β, then q il and q ir are calculated from q i in the x-y plane, and are respectively represented as q il (x i -Δx,y i -Δy) and q ir (xi +Δx,y i +Δy), where the calculation formula of Δx and Δy is:

[0043]

[0044] On a certain planning path segment, the starting point q i-1 and the ending point q i constitute a vector The angle of the vector in space represents the pitch angle θ of the path segment i :

[0045]

[0046] The maximum pitch angle constraint is:

[0047] θ i ≤θ max

[0048] In the formula, θ max is the maximum pitch angle;

[0049] The articulation angle of the i-th path point q i is defined as ψ i For three adjacent path points q i-2 , q i-1 , q i , the articulation angle at the path point is described by the included angle of the projections of the corresponding vectors of the two adjacent path segments on the horizontal plane;

[0050]

[0051] The maximum articulation angle constraint is:

[0052] ψ i ≤ψ max

[0053] In the formula, ψ max is the maximum articulation angle;

[0054] Based on the constructed objective function and constraint conditions, the mathematical model of path planning is obtained:

[0055] min F(x) = [F L (x), F L (x), F S (x)] T

[0056]

[0057] In the formula, n represents the number of path points contained in a path, F(X) represents a multi-objective function, X represents a set of decision variables, and T represents an inversion operation.

[0058] Further, the process of step S4 is as follows:

[0059] By using the multi-objective genetic algorithm, the algorithm is iterated by combining the objective function and the constraint condition; after the iteration is completed, the first 10 non-inferior solutions are extracted from the obtained non-dominant sorting result; then the difference between the same objective function values in different non-inferior solutions is calculated, and a weight is assigned to each difference; by comprehensively considering the difference of the objective function values and the weight, an optimal solution is selected as the optimal path output by the algorithm; finally, the path point coordinates and speed values of the optimal path are stored.

[0060] Compared with the prior art, the beneficial effects of the present application are:

[0061] 1. The present application can realize the unmanned driving function of the articulated loader, and get rid of the dependence on the driver, thereby significantly improving the working efficiency of the articulated loader and effectively reducing the incidence of driving accidents.

[0062] 2. The present application adopts a multi-objective path planning strategy under constraint conditions compared with the traditional single-objective shortest path planning method, aiming at the complex structure characteristics and harsh working environment of the articulated loader. This innovation can better adapt to the needs of the articulated loader in harsh working environments, and improve the practicality and adaptability of path planning.

[0063] 3. The present application introduces a learning function and successfully constructs an adaptive kriging monte carlo dynamics agent model. The introduction of this model effectively solves the problem of long simulation calculation time, making the acquisition of vehicle characteristic parameters more rapid and efficient. BRIEF DESCRIPTION OF DRAWINGS

[0064] Figure 1 The path planning flowchart of the method of the present application.

[0065] Figure 2 The path planning environment map.

[0066] Figure 3 The side tilt angle diagram when the articulated loader is crossing obstacles; wherein (a) is the change of the front vehicle body when the articulated loader is crossing obstacles, (b) is the change of the rear vehicle body when the articulated loader is crossing obstacles, and (c) is the side tilt angle calculation.

[0067] Figure 4 The path planning framework diagram.

[0068] Figure 5 The environment map in the form of a three-dimensional grid map.

[0069] Figure 6 The joint simulation diagram.

[0070] Figure 7 Trajectory planning results for flat terrain; wherein (a) is a trajectory intuitive diagram, (b) is a length convergence curve, (c) is a time convergence curve, and (d) is a smoothness convergence curve.

[0071] Figure 8 Trajectory planning results for regular obstacle terrain; wherein (a) is a trajectory intuitive diagram, (b) is a length convergence curve, (c) is a time convergence curve, and (d) is a smoothness convergence curve.

[0072] Figure 9 Trajectory planning results for extreme obstacle terrain; wherein (a) is a trajectory intuitive diagram, (b) is a length convergence curve, (c) is a time convergence curve, and (d) is a smoothness convergence curve. DETAILED DESCRIPTION

[0073] In order to have a clearer understanding of the technical features, objectives and beneficial effects of the present application, the technical solutions of the present application are described in detail below, but cannot be understood as limiting the scope of implementation of the present application.

[0074] The specific implementation of the present application is described in detail below in combination with specific embodiments.

[0075] One embodiment of the present application provides a driving path planning method of an unmanned loader, a flowchart of which is shown as Figure 1 The method comprises the following steps:

[0076] Step S1, a dynamic adaptive model is constructed to provide articulated loader dynamic characteristic data for subsequent path planning;

[0077] Step S2, a working environment is established: the vehicle working environment is simplified into a three-dimensional map to determine the initial position and target position of the path;

[0078] Step S3, a target function and a constraint condition are constructed;

[0079] Step S4, an optimal path is output by the planner: the non-dominated solutions of the top 10 non-inferior solutions after iteration of the multi-objective genetic algorithm are extracted and compared by weighting, and an optimal solution is selected as the optimal path output by the algorithm after comprehensive consideration;

[0080] Step S5, whether the output optimal path is reasonable is verified by joint simulation; if it is reasonable, the path is output, and if it is not reasonable, the step S4 is returned for further optimization.

[0081] Specifically as follows:

[0082] (1) Step S1, a dynamic adaptive model is constructed to provide articulated loader dynamic characteristic data for subsequent path planning.

[0083] Based on the Kriging model and Monte Carlo simulation, a learning function is introduced to construct an adaptive Kriging Monte Carlo simulation (AK-MCS) dynamic proxy model. In the dynamic proxy model, the learning function indicator U(x) is associated with the lower confidence limit lcb function for optimization, and the lcb is defined as:

[0084]

[0085] wherein lcb(x) represents the lower limit function of confidence, represents the model prediction mean, represents the model prediction deviation; taking b as 2, the lcb is calculated and minimized to 0, at this time U(x) is equal to the value of b, corresponding to the model error probability Φ(-2) = 0.023. After training, the dynamic proxy model provides the articulated loader dynamic characteristic data for path planning.

[0086] (2) Step S2, establishing a working environment: simplifying the vehicle working environment into a three-dimensional map to determine the initial position and target position of the path.

[0087] As shown in Figure 2 , the three-dimensional environment data is simplified into coordinate values x, y, and z, and the coordinate values are stored in the map environment matrix; the design variables are the coordinates of each path point (x i ,y i ,z i ), the starting point S coordinate is set as (x1, y1, z1), and the end point G coordinate is set as (x n ,y n ,z n ); the map is uniformly divided into a grid in the x-y plane, and the remaining trajectory path points between the starting point and the end point are discretely distributed at the grid vertices, represented as (x i ,y i ,z i ); finally, the points are connected in order to form a complete trajectory.

[0088] (3) Step S3, constructing an objective function and determining a constraint condition.

[0089] (1) Constructing an objective function;

[0090] Since the planned path is composed of a series of three-dimensional coordinate points, the total spatial distance between adjacent path points is used as the path length cost, i.e., the path travel length cost function F L is defined as:

[0091]

[0092] wherein n represents the number of path segments in the path, and l iIt represents the spatial distance between the two end points of the i-th path, which can be expressed as:

[0093]

[0094] The total travel time of each path segment is used as the path time cost, that is, the path travel time cost function F T Defined as:

[0095]

[0096] In the formula, n represents the number of path points contained in a path, t i It represents the travel time between two adjacent path points in the i-th segment of the path, which can be expressed as:

[0097]

[0098] Where, l i represents the spatial distance between two adjacent path points in the i-th path, v i and v i-1 Indicates the safe speed value calculated by the dynamic agent model at two path points.

[0099] The core of the path smoothing objective function is to ensure smooth vehicle travel for easy path tracking while avoiding accidents caused by excessive steering angles. Since a motion path consists of a series of spatial coordinate points, with the line connecting each two points forming a straight line segment, to ensure smooth driving, it is necessary to maintain smooth transitions between straight lines. Therefore, the path smoothness cost F can be measured by evaluating the change in angle between adjacent straight lines on each path. S :

[0100]

[0101] In the formula, n represents the number of path points contained in a path, and For q i-2 ,q i-1 ,q i The two adjacent vectors formed by these three consecutive path points are:

[0102]

[0103] Where x i is the x coordinate of the i-th path, y i is the y coordinate of the i-th path, z i is the z coordinate of the i-th path, x i-1 is the x coordinate of the i-1th path, y i-1 is the y coordinate of the i-1th path, z i-1z-coordinate of the i-1th segment path, x i-2 x-coordinate of the i-2th segment path, y i-2 y-coordinate of the i-2th segment path, z i-2 z-coordinate of the i-2th segment path

[0104] (2) determining the constraint condition

[0105] In view of the characteristics of the non-structural environment and the characteristics of the engineering vehicle, the maximum articulation angle, the maximum pitch angle and the maximum roll angle constraint are incorporated into the constraint condition. The articulation angle constraint is to ensure that the turning angle of the loader is strictly controlled within a safe range during turning, so as to avoid the risk caused by excessive turning, and the present application is set to ±30°. In addition, the pitch angle constraint reflects the maximum angle that the loader can reach when overcoming obstacles, which is a key indicator for evaluating the off-road performance of the vehicle, and the present application is selected as ±30°. It is worth noting that the obstacle crossing critical roll angle of different loaders under different speed, pitch angle and articulation angle conditions also differs, and the present application is set to ±20°.

[0106] Suppose the vehicle crosses the obstacle at the path point p i , the right front wheel first contacts the obstacle, causing the front vehicle body to tilt to the left as shown in (a) of Figure 3 . At this time, the posture of the rear vehicle body is shown in (b) of Figure 3 , and the whole vehicle body tilts the rear axle without tilting. Therefore, in the constraint condition of the present application, the roll angle mentioned refers to the tilting angle of the front vehicle body, which is represented by φ as shown in (c) of Figure 3 , in addition, the ground contact point of the path point q i (x i , y i , z i ) is q il (x il , y il , z il ), and the obstacle crossing point is q ir (x ir , y ir , z ir ), then the roll angle φ i is:

[0107]

[0108] The maximum roll angle constraint is:

[0109] φ i ≤φ max

[0110] In the formula, φ max is the maximum roll angle, and L is the width of the articulated loader body.

[0111] Assume the path point q i The angle between the path segment where q il lies and the x-axis in the x-y plane is β, then q ir The angle between the path segment where q i lies and the x-axis in the x-y plane is β, then q il (x i -Δx,y i -Δy) and q ir (x i +Δx,y i +Δy), where the calculation formula of Δx and Δy is:

[0112]

[0113]

[0114] On a certain planning path segment, the starting point q i-1 and the ending point q i constitute a vector The angle of the vector in space represents the pitch angle θ i of the path segment:

[0115]

[0116] The maximum pitch angle constraint is:

[0117] θ i ≤θ max

[0118] In the formula, θ max is the maximum pitch angle;

[0119] The articulation angle of the i-th path point q i is defined as ψ i For three adjacent path points q i-2 , q i-1 , q i , the articulation angle at q i is described by the projection of the corresponding vectors of the two adjacent path segments on the horizontal plane.

[0120]

[0121] The maximum articulation angle constraint is:

[0122] ψ i ≤ψ max

[0123] In the formula, ψ max is the maximum articulation angle;

[0124] Based on the constructed objective function and constraint conditions, a mathematical model of path planning is obtained:

[0125] minF(x)=[F L (x),F T (x),F S (x)] T

[0126]

[0127] In the formula, n represents the number of path points contained in a path, F(X) represents a multi-objective function, X represents a set of decision variables, and T represents an inversion operation.

[0128] Step S4, outputting an optimal path by using a planner: extracting non-inferior solutions in the first 10 non-dominated sorting after iteration of the multi-objective genetic algorithm, and selecting an optimal solution as the optimal path output by the algorithm after weighted comparison and comprehensive consideration;

[0129] Referring to Figure 4 , a multi-objective genetic algorithm (NSGA-II) is used to combine the objective function and the constraint condition to perform iterative operation on the algorithm; after the iteration is completed, the first 10 non-inferior solutions are extracted from the obtained non-dominated sorting results; then, the difference between the same objective function values in different non-inferior solutions is calculated, and a weight is assigned to each difference; by comprehensively considering the difference between the objective function values and the weight, an optimal solution is selected as the optimal path output by the algorithm; finally, the path point coordinates and speed values of the optimal path are stored.

[0130] Step S5, verifying whether the output optimal path is reasonable by joint simulation; if the path is reasonable, the path is output; if the path is not reasonable, the step S4 is returned to further optimize.

[0131] In order to test the performance of the method proposed in the application, a joint simulation platform based on MATLAB / Simulink and RecurDyn software is built.

[0132] Firstly, a terrain map of a three-dimensional unstructured environment is drawn and generated by using EarthSculptor, and then it is imported into MATLAB in the form of point cloud for simulation, as shown in Figure 5 .

[0133] Then, joint simulation is performed by using Simulink and RecurDyn to dynamically simulate and verify the planned path, as shown in Figure 6 .

[0134] In a relatively flat road surface environment, as shown in Figure 7As shown in (a), the generated trajectory always follows the road with less undulations, and when approaching an obstacle, the off-road characteristics of the articulated loader are fully utilized to successfully cross the obstacle and reach the end point. Figure 7 As shown in (b) to (d), the path length, path time, and path smoothness gradually converged with the increase in the number of iterations, and convergence was completed before iteration 40. Specifically, the path length converged to within 131.129 meters, with the minimum value converged to 114.735 meters; the path time converged to within 26.090 seconds, with the minimum value converged to 23.008 seconds; and the path smoothness converged to within 18 degrees, with the minimum value converged to 4.483 degrees.

[0135] This result shows that the proposed method can effectively adapt to complex terrain conditions and plan a driving path for the articulated loader that meets its characteristics. As the number of iterations increases, the convergence of path length, time, and smoothness demonstrates the stability and reliability of the algorithm.

[0136] Path planning is performed for vehicle driving conditions where the starting point is far away from conventional obstacles and the end point is surrounded by positive and negative obstacles, such as Figure 8 As shown in (a). As the number of iterations increases, Figure 8 As shown in (b) to (d), the path length, path time, and path smoothness gradually converged, and convergence was completed before the number of iterations reached 50. Specifically, the path length converged to within 118.254 meters, with the minimum value converged to 112.999 meters; the path time converged to within 38.033 seconds, with the minimum value converged to 29.495 seconds; and the path smoothness converged to within 11 degrees, with the minimum value converged to 4.142 degrees.

[0137] This result demonstrates that the proposed method can effectively implement path planning and optimization when considering complex driving scenarios, with good convergence performance. The proposed method demonstrates good stability and reliability when handling complex driving situations. Convergence analysis of metrics such as path length, time, and smoothness further validates the algorithm's effectiveness and practicality, providing a feasible solution for the practical application of articulated loaders.

[0138] Path planning is performed for vehicles whose starting point is far from an extreme obstacle and whose end point is near the obstacle. Figure 9 As shown in (a), the path planning successfully avoids extreme obstacles and reaches the end point smoothly after crossing regular obstacles. Figure 9As shown in (b)-(d), the path length, path time and path smoothness gradually converge, and the convergence is completed before the iteration number is 30. Specifically, the path length converges within 107.448 m, and the minimum value converges to 105.787 m; the path time converges within 27.207 s, and the minimum value converges to 23.078 s; the path smoothness converges within 12°, and the minimum value converges to 2.599°.

[0139] This result shows that the method successfully avoids extreme obstacles and selects a relatively shorter, shorter time and higher smoothness path. The improvement of path smoothness helps to reduce the bump and shake in vehicle driving, and improves the comfort and stability of driving. At the same time, the effective control of path length and path time also reduces the energy consumption in the driving process, and improves the driving efficiency. These results further verify the feasibility and practicability of the method in complex terrain.

[0140] The above is only the preferred embodiment of the present application, it should be noted that for those skilled in the art, without departing from the concept of the present application, can make a number of deformation and improvement, these should be considered as the protection scope of the present application, these will not affect the effect and practicality of the patent of the present application.

Claims

1. A method for planning a driving path for an unmanned loader, characterized in that: The following steps are involved: Step S1: constructing a dynamic adaptive model to provide dynamic characteristic data of the articulated loader for subsequent path planning; Step S2, establishing a working environment: simplifying the vehicle working environment into a three-dimensional map to determine the initial position and target position of the path; Step S3: constructing the objective function and determining the constraints; Step S4: Output the optimal path using the planner: extract the top 10 non-inferior solutions from the non-dominated sorting after the iteration of the multi-objective genetic algorithm, perform weighted comparison, and select the best solution after comprehensive consideration as the optimal path output by the algorithm; Step S5: Check whether the output optimal path is reasonable through joint simulation; if it is reasonable, output the path; if it is not reasonable, return to step S4 for further optimization; In step S3, the specific process of constructing the objective function is as follows: The sum of the spatial distances between adjacent path points is used as the path length cost, that is, the path travel length cost function Defined as: ； Where, n Indicates the number of path segments in the path. Indicates the i The spatial distance between the path points at both ends of the segment path is specifically expressed as: ； The total travel time of each path segment is used as the path time cost, that is, the path travel time cost function Defined as: ； Where, n Indicates the number of path points contained in a path. Indicates the path i The travel time between two adjacent waypoints in a segment is expressed as: ； Where, Indicates the i The spatial distance between two adjacent path points in a segment path, and Represents the safe speed value calculated by the dynamic agent model at two path points; The smoothness cost of a path is measured by evaluating the change in angle between adjacent straight line segments on each path. : ； Where, n Indicates the number of path points contained in a path. and for the reason 、 、 The two adjacent vectors formed by these three consecutive path points are: ； ； Where, For the i Segment path coordinate, For the i Segment path coordinate, For the i Segment path coordinate, For the Segment path coordinate, For the Segment path coordinate, For the Segment path coordinate, For the Segment path coordinate, For the Segment path coordinate, For the Segment path coordinate; The mathematical model of path planning is obtained: ； ； Where, n Indicates the number of path points contained in a path. represents a multi-objective function, represents the set of decision variables, T Indicates an inversion operation. is the roll angle, is the maximum roll angle, is the pitch angle of the path segment, is the maximum pitch angle, For the i Waypoints The hinge angle, is the maximum articulation angle.

2. The method for planning a driving path for an unmanned loader according to claim 1, wherein: The specific process of step S1 is as follows: Based on the Kriging model and Monte Carlo simulation, the learning function is introduced to construct the adaptive Kriging Monte Carlo dynamics agent model; in the dynamics agent model, the learning function index is used to With lower confidence boundary function Optimize the association, Defined as: ； Where, represents the confidence lower limit function, represents the model prediction mean, Represents the model prediction bias; is 2, calculate and minimize is 0, at this time and The values ​​are equal, corresponding to the probability of model error ,After training, the dynamic agent model provides the ,dynamic characteristic data of the articulated loader for path planning.

3. The method for planning a driving path for an unmanned loader according to claim 2, wherein: The specific process of step S2 is as follows: Simplify 3D environment data into coordinate values 、 、 , and store the coordinate values ​​in the map environment matrix; The design variables are the coordinates of each path point , set the starting point The coordinates are ,end The coordinates are ; Place the map in The plane is evenly divided into grids, and the remaining trajectory path points between the starting point and the end point are discretely distributed at the grid vertices, expressed as ; Finally, connect the points in sequence to form a complete trajectory.

4. The method for planning a driving path for an unmanned loader according to claim 3, wherein: In step S3, the specific process of determining the constraint conditions is as follows: The constraints include the maximum articulation angle, maximum pitch angle, and maximum roll angle constraints. The articulation angle constraint is ±30°, the pitch angle constraint is ±30°, and the roll angle constraint is ±20°. Waypoints The ground contact point is , the obstacle crossing point is , then the roll angle for: ； The maximum roll angle constraint is: ； Where, is the maximum roll angle, is the body width of the articulated loader; Hypothetical waypoints The path segment is in In-plane and The angle between the axes is ,but and exist In-plane Calculated and expressed as and ,in and The calculation formula is: ； ； On a planned path segment, the starting point and end point Composition vector , the pitch angle of the path segment is represented by the angle of the vector in space : ； The maximum pitch angle constraint is: ； Where, is the maximum pitch angle; The first i Waypoints The hinge angle is defined as , for three adjacent path points 、 、 , the hinge angle at the path point is described by the projection angle of the corresponding vectors of the adjacent two end path segments on the horizontal plane; ； The maximum hinge angle constraint is: ； Where, is the maximum articulation angle.

5. The method for planning a driving path for an unmanned loader according to claim 4, wherein: The process of step S4 is as follows: A multi-objective genetic algorithm is used to iterate the algorithm, combining the objective function and constraints. After the iteration is completed, the top 10 non-inferior solutions are extracted from the obtained non-dominated sorting results. The differences between the same objective function values ​​in different non-inferior solutions are then calculated, and a weight is assigned to each difference. By comprehensively considering the differences and weights of the objective function values, an optimal solution is selected as the optimal path output by the algorithm. Finally, the path point coordinates and speed of the optimal path are stored.

Images