Carry-scraper and global path optimization method, device and system thereof, and program product

Abstract

The invention relates to a carry-scraper and a global path optimization method, device and system thereof, and a program product. The method comprises the following steps: converting an original discrete grid path into a continuous path conforming to kinematics constraints of the carry-scraper; according to the continuous path, a total objective function of the continuous path is constructed, and the total objective function comprises a smooth item, a curvature item and a data item; and performing iterative optimization solution on the total objective function by adopting a predetermined algorithm to obtain a global optimization path of the carry-scraper. According to the method, the path curvature constraint is strictly ensured, sharp turning is avoided, the curvature penalty term is introduced into the optimization objective function, penalty can be applied to the over-limit region, and the optimization path can be forced to meet the kinematics constraint.

Description

Technical Field

[0001] The present disclosure relates to the technical field of engineering machinery automation, and in particular to a scraper and a global path optimization method, device, system, and program product thereof. Background Art

[0002] As the core transportation equipment in mining, the path planning optimization of underground loaders is a key technical issue to improve operational efficiency and ensure safe operation. Summary of the Invention

[0003] Through research, the inventors found that the underground mining environment has the characteristics of complex spatial structure, many dynamic obstacles, and low visibility. Although the related technology's path planning method based on grid maps (such as the A* algorithm) can generate an initial path, it often ignores the vehicle's kinematic constraints and path smoothness requirements, resulting in problems such as sharp turns and discontinuous curvature in the planning results.

[0004] In view of at least one of the above technical problems, the present disclosure provides a scraper and its global path optimization method, device and system, and program product, which strictly ensure the path curvature constraint, avoid sharp turns, introduce a curvature penalty term in the optimization objective function, impose penalties on the excess area, and force the optimized path to meet the kinematic constraints.

[0005] According to one aspect of the present disclosure, a method for optimizing a global path of a scraper is provided, comprising:

[0006] The original discrete grid path is converted into a continuous path that meets the kinematic constraints of the scraper;

[0007] Constructing a total objective function of the continuous path according to the continuous path, wherein the total objective function includes a smoothing term, a curvature term, and a data term;

[0008] The total objective function is iteratively optimized and solved using a predetermined algorithm to obtain a global optimization path for the scraper.

[0009] In some embodiments of the present disclosure, converting the original discrete grid path into a continuous path that complies with the kinematic constraints of the scraper includes:

[0010] The original discrete grid path is thinned by cubic spline interpolation to generate a continuous path that meets the kinematic constraints of the scraper.

[0011] In some embodiments of the present disclosure, performing a thinning process on the original discrete grid path by using cubic spline interpolation includes:

[0012] The original discrete grid path is thinned by cubic spline interpolation using half the number of original path points as the number of interpolation points.

[0013] In some embodiments of the present disclosure, constructing an overall objective function of the continuous path according to the continuous path includes:

[0014] Constructing a smoothing term, wherein the smoothing term is used to minimize the rate of change of path curvature through a second-order difference matrix so that the path points are evenly distributed;

[0015] Obtaining a predetermined curvature threshold value according to a minimum turning radius of the scraper;

[0016] Constructing a curvature term, wherein the curvature term is used to impose a penalty on a path segment whose path point curvature is greater than a predetermined curvature threshold through a piecewise penalty function;

[0017] Constructing a data item, wherein the data item is used to constrain the degree of deviation between the continuous path and the original discrete grid path;

[0018] The overall objective function is determined based on the smoothness term, the curvature term and the data term.

[0019] In some embodiments of the present disclosure, determining the overall objective function according to the smoothness term, the curvature term, and the data term includes:

[0020] The overall objective function is determined based on the path smoothing coefficient, the smoothing term, the curvature constraint coefficient, the curvature term, the data fitting coefficient and the data term.

[0021] In some embodiments of the present disclosure, the scraper global path optimization method further includes:

[0022] Determine the first coordinate and the second coordinate of the path point of the scraper moving in the plane coordinate system;

[0023] Obtaining the first-order derivative and the second-order derivative of the first coordinate with respect to the second coordinate, and the first-order derivative and the second-order derivative of the second coordinate with respect to the first coordinate;

[0024] The curvature of the waypoint is determined based on the first and second derivatives of the first coordinate with respect to the second coordinate, and the first and second derivatives of the second coordinate with respect to the first coordinate.

[0025] In some embodiments of the present disclosure, using a predetermined algorithm to iteratively optimize and solve the overall objective function includes:

[0026] Calculate the total gradient of the total objective function with respect to the path point coordinates to guide the optimization direction;

[0027] Iteratively optimize the path point coordinates, minimize the objective function, and generate the optimized path point set based on the input initial path point, total objective function, and total gradient.

[0028] In some embodiments of the present disclosure, iteratively optimizing the path point coordinates, minimizing the objective function, and generating an optimized path point set based on the input initial path points, the total objective function, and the total gradient include:

[0029] A local quadratic model is constructed in each iteration, limiting the step size to within the trust region radius;

[0030] Use trust region approximation to solve the optimal step size for quadratic models;

[0031] The trust region radius is dynamically adjusted based on the drop ratio between the model prediction and the actual function.

[0032] In some embodiments of the present disclosure, calculating the total gradient of the total objective function with respect to the path point coordinates includes:

[0033] Calculate the gradient of the smoothing term, the gradient of the curvature term, and the gradient of the data term;

[0034] The total gradient is determined according to the path smoothing coefficient, the smoothing term gradient, the curvature constraint coefficient, the curvature term gradient, the data fitting coefficient and the data term gradient.

[0035] In some embodiments of the present disclosure, calculating the smoothing term gradient includes:

[0036] The smoothing term gradient of each path point is determined based on the coordinates of each path point and the second-order difference matrix.

[0037] In some embodiments of the present disclosure, calculating the curvature term gradient includes:

[0038] The curvature term gradient of each path point is determined according to the curvature of each path point, the derivative of the curvature with respect to the coordinates of the path point, the sign function of the curvature, and a predetermined curvature threshold.

[0039] In some embodiments of the present disclosure, calculating the data item gradient includes:

[0040] The data item gradient of each path point is determined according to the path point coordinates of the continuous path and the path point coordinates of the original discrete grid path.

[0041] In some embodiments of the present disclosure, the scraper global path optimization method further includes:

[0042] Searching for the shortest feasible path from a starting point to an end point in a grid map, wherein the shortest feasible path is an initial path composed of a plurality of grid nodes;

[0043] The expansion area of ​​the obstacle in the global map is increased based on the shortest feasible path to form a discrete grid path that conforms to the original one.

[0044] According to another aspect of the present disclosure, a global path optimization device for a scraper is provided, comprising:

[0045] a path continuation module configured to convert an original discrete raster path into a continuous path that complies with kinematic constraints of the scraper;

[0046] an objective function construction module, configured to construct a total objective function of the continuous path according to the continuous path, wherein the total objective function includes a smoothing term, a curvature term, and a data term;

[0047] The path optimization module is configured to use a predetermined algorithm to iteratively optimize and solve the overall objective function to obtain a global optimized path of the scraper.

[0048] According to another aspect of the present disclosure, a global path optimization device for a scraper is provided, comprising:

[0049] Memory; and

[0050] A processor coupled to the memory is configured to execute the scraper global path optimization method according to any one of the above embodiments based on instructions stored in the memory.

[0051] According to another aspect of the present disclosure, a global path optimization system for a scraper is provided, comprising a data acquisition device and the global path optimization device for a scraper as described in any one of the above embodiments.

[0052] According to another aspect of the present disclosure, a scraper is provided, comprising the scraper global path optimization system according to any one of the above embodiments.

[0053] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the method for global path optimization of a scraper as described in any of the above embodiments is implemented.

[0054] According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the method for global path optimization of a scraper according to any one of the above embodiments is implemented.

[0055] The present invention strictly ensures the path curvature constraint, avoids sharp turns, introduces a curvature penalty term in the optimization objective function, can impose penalties on the excess area, and can force the optimization path to meet the kinematic constraints. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0057] Figure 1 Schematic diagram of some embodiments of the global path optimization method for a scraper disclosed in the present invention.

[0058] Figure 2 Schematic diagrams of other embodiments of the global path optimization method for a scraper disclosed in the present invention.

[0059] Figure 3 Schematic diagram of the structure of some embodiments of the global path optimization device for scraper disclosed in the present invention.

[0060] Figure 4 Schematic diagrams of other embodiments of the global path optimization device for scraper disclosed in the present invention.

[0061] Figure 5 Schematic diagram of some embodiments of the global path optimization system for scrapers disclosed herein. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is in no way intended to limit the present disclosure and its application or use. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present disclosure.

[0063] Unless specifically stated otherwise, the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure.

[0064] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0065] Technologies, methods and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, such technologies, methods and equipment should be considered part of the authorization specification.

[0066] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0067] It should be noted that like reference numerals and letters refer to like items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0068] Through research, the inventors found that in related technologies, in narrow alleys or intersections, the right-angle turning path can easily cause the curvature of the scraper to exceed the mechanical limit when turning, causing increased tire wear, trajectory tracking deviation and even rollover risk.

[0069] The technical difficulty in related research lies in efficiently solving nonlinear constraints. Traditional methods use gradient descent or sequential quadratic programming, but their ability to handle the strong non-convexity of curvature constraints is limited.

[0070] A related technology is underground unmanned mining vehicle path optimization based on a vehicle body sweep model. In this related technology, Bezier curve and polynomial fitting require a large number of sampling points and high-order parametric equations, which may lead to high consumption of computing resources. The objective function requires that the left and right boundaries are equidistant from obstacles, which may have no solution when the obstacle distribution is asymmetric (such as dense obstacles on one side).

[0071] Another related technique relies heavily on the quality of the initial path generated by algorithm A. If the initial path closely adheres to obstacles, trust region constraints may limit the optimization space, preventing a true time-optimal solution. When using a trapezoidal integral discretization model, the increase in path points leads to a sharp increase in the dimensionality of the optimization variables. With N path points, the number of constraints is O(MN) (where M is the number of obstacles). Solving convex problems in large-scale environments can exceed the computational time of the drone's onboard computing power.

[0072] Another related technology uses a quadratic programming solver to process constraints in the inner loop. When the number of path points increases, the scale of variables and constraints grows exponentially, resulting in a decrease in real-time performance.

[0073] In view of at least one of the above technical problems, the present disclosure provides a scraper and a global path optimization method, device, system, and program product thereof. The present disclosure is described below through specific embodiments.

[0074] Figure 1 Schematic diagram of some embodiments of the global path optimization method for a scraper disclosed in the present invention. Figure 2 Schematic diagram of some other embodiments of the global path optimization method for scraper disclosed in the present invention. Preferably, Figure 1 or Figure 2The embodiment can be executed by the scraper global path optimization device of the present disclosure, the scraper global path optimization system of the present disclosure, or the scraper of the present disclosure. Figure 1 The global path optimization method for scraper shown may include at least one of steps 100 to 300. Figure 2 The global path optimization method for scraper shown in the figure can include Figure 1 In addition to at least one step from step 100 to step 300 , the embodiment may further include at least one step from step 90 and step 250 .

[0075] In step 90, an original discrete grid path is generated.

[0076] In some embodiments of the present disclosure, the original discrete grid path may be the original grid path P_ind.

[0077] In some embodiments of the present disclosure, step 90 may include at least one of step 91 and step 92 .

[0078] In step 91 , the shortest feasible path from the starting point to the end point is searched in the grid map, wherein the shortest feasible path is an initial path composed of a plurality of grid nodes.

[0079] In some embodiments of the present disclosure, the shortest feasible path is an initial path consisting of a series of grid nodes.

[0080] In step 92, the expansion area of ​​the obstacle in the global map is increased based on the shortest feasible path to form a discrete grid path that conforms to the original one.

[0081] In some embodiments of the present disclosure, step 92 may include: improving the A* algorithm by adding an expansion area to the path boundary, thereby ensuring the safety of the vehicle in the initial path and increasing safety redundancy for further curve optimization at turns.

[0082] In some embodiments of the present disclosure, step 92 may include: based on the A* algorithm, according to formula (1), using the Manhattan distance heuristic function h(n), to construct a discrete grid path that conforms to the original.

[0083] f(n)=g(n)+h(n) (1)

[0084] The above-mentioned embodiment of the present disclosure increases the expansion area of ​​obstacles in the global map, performs turning redundancy for further curve optimization, and prevents collisions.

[0085] The inventors discovered through research that the original grid path generated based on the improved A* algorithm is usually composed of discrete nodes, and has a curvature discontinuity problem caused by broken line connections.

[0086] In step 100 , the original discrete grid path is converted into a continuous path that complies with the kinematic constraints of the scraper.

[0087] In some embodiments of the present disclosure, step 100 may include: path continuation processing.

[0088] In some embodiments of the present disclosure, step 100 may include: path interpolation.

[0089] In some embodiments of the present disclosure, step 100 may include: performing a thinning process on the original discrete grid path by cubic spline interpolation to generate a continuous path that meets the kinematic constraints of the scraper.

[0090] In some embodiments of the present disclosure, step 100 may include: performing a sparse process on the path by cubic spline interpolation to construct a continuously differentiable parameterized path point set, which can be mathematically represented as a discrete node sequence {x i ,y i} Perform spline fitting to generate high-order continuous functions x(t) and y(t), where the parameter t represents the arc length of the path, providing an operational continuous path description for subsequent gradient optimization.

[0091] In some embodiments of the present disclosure, step 100 may include: performing a thinning process on the original discrete grid path by using cubic spline interpolation, using half of the number of original path points as the number of interpolation points.

[0092] In some embodiments of the present disclosure, step 100 may include: calculating the length of the raster path, taking half of the length of the raster path, and re-evaluating the path, thereby reducing the density of path points, reducing the constraints on data items, avoiding the curvature explosion problem that may occur at right-angle turning points, and realizing comprehensive constraints on smoothness items, curvature items, and data items.

[0093] In some embodiments of the present disclosure, step 100 may include: determining the number of interpolation points according to formula (2), with half of the original path points as the interpolation points, thereby balancing path smoothness and curvature constraints and improving computational efficiency.

[0094] N=0.5*size(P ind ,1) (2)

[0095] In some embodiments of the present disclosure, step 100 may further include: merging the interpolation coordinates into a column vector, as shown in formula (3), size(P ind ,1) is the matrix P ind number of rows.

[0096]

[0097] In formula (3), x interp and y interp are the horizontal and vertical coordinates after interpolation, and X0 is the merged column vector.

[0098] In step 200, a total objective function of the continuous path is constructed according to the continuous path, wherein the total objective function includes a smoothness term, a curvature term, and a data term.

[0099] In some embodiments of the present disclosure, step 200 may include: defining a mathematical goal of optimization, balancing smoothness, curvature constraints, and path conformity.

[0100] In some embodiments of the present disclosure, step 200 may include: constructing a multi-objective optimization function to balance path smoothness, curvature constraints, and global optimality.

[0101] In some embodiments of the present disclosure, step 200 may include at least one of steps 201 to 205 .

[0102] In step 201 , a smoothing term is constructed, wherein the smoothing term is used to minimize the rate of change of path curvature through a second-order difference matrix D so that the path points are evenly distributed.

[0103] In some embodiments of the present disclosure, the smoothing term minimizes the rate of change of the path curvature through the second-order difference matrix D, forcing the path points to be evenly distributed. The mathematical form of the smoothing term is shown in Formula (4) or Formula (5).

[0104] In some embodiments of the present disclosure, step 201 may include: determining a smoothing term according to formula (4).

[0105]

[0106] In some embodiments of the present disclosure, the discrete form of formula (4) is shown as formula (5).

[0107] J smooth =‖D x ‖ 2 +||D y || 2 (5)

[0108] In step 202, a predetermined curvature threshold k is obtained according to the minimum turning radius of the scraper. max .

[0109] In step 203, a curvature term is constructed, wherein the curvature term J curevature , which is used to impose penalties on path segments whose path point curvature is greater than a predetermined curvature threshold through a piecewise penalty function.

[0110] In some embodiments of the present disclosure, the curvature term may be a curvature constraint term.

[0111] In some embodiments of the present disclosure, step 203 may include: determining a curvature term according to formula (6).

[0112]

[0113] In formula (6), k max It is calculated based on the minimum turning radius of the vehicle.

[0114] In some embodiments of the present disclosure, the curvature k i The following steps can be used to determine the first coordinate x and the second coordinate y of the path point of the scraper moving in the plane coordinate system; obtain the first-order derivative of the second coordinate y with respect to the first coordinate x and the second-order derivative and the first derivative of the first coordinate x with respect to the second coordinate y and the second-order derivative The first derivative of the second coordinate y with respect to the first coordinate x and the second-order derivative and the first derivative of the first coordinate x with respect to the second coordinate y and the second-order derivative Determine the path point curvature k i .

[0115] In some embodiments of the present disclosure, the curvature k is determined according to formula (7): i .

[0116]

[0117] In some embodiments of the present disclosure, the curvature term is used to penalize the curvature k that exceeds the maximum allowed curvature. max path segments to ensure that the path meets the vehicle kinematic constraints and avoids turning radius that is too small.

[0118] The curvature constraint term in the above embodiment of the present disclosure adopts a piecewise penalty function J curvature , when the curvature k i Exceeding the threshold k max A secondary penalty is imposed when , ensuring that the path meets the vehicle's minimum turning radius constraint.

[0119] In step 204, a data item is constructed, wherein the data item is used to constrain the degree of deviation between the continuous path and the original discrete grid path.

[0120] In some embodiments of the present disclosure, the data constraint term J is determined according to formula (8): data .

[0121] Jdata =‖x-x0‖ 2 -‖y-y0‖ 2 (8)

[0122] In some embodiments of the present disclosure, the data constraint item J data , which is used to keep the optimized path (x, y) close to the original path (x0, y0) and prevent excessive deviation from the initial feasible path generated by algorithms such as A*.

[0123] In some embodiments of the present disclosure, the data item constrains the deviation degree of the optimized path {x, y} from the initial path {x0, y0} through formula (8) to avoid obstacle avoidance failure caused by excessive smoothing.

[0124] In step 205, the overall objective function is determined based on the smoothness term, the curvature term, and the data term.

[0125] In some embodiments of the present disclosure, step 205 may include: determining the overall objective function according to the path smoothing coefficient, the smoothing term, the curvature constraint coefficient, the curvature term, the data fitting coefficient, and the data term.

[0126] In some embodiments of the present disclosure, step 205 may include: setting the overall objective function according to formula (9).

[0127] J=w smooth ·J smooth +w curvature ·J curvature +w data ·J data (9)

[0128] In formula (9), J represents the total cost based on the smoothness term, curvature constraint and path fit. smooth is the path smoothing coefficient, w curvature is the curvature constraint coefficient, w data is the data fitting coefficient. J smooth is the smoothing term, J curvature is the curvature constraint term, J data is a data item.

[0129] In some embodiments of the present disclosure, the weight coefficient can be dynamically adjusted according to the scenario.

[0130] In the above embodiment of the present disclosure, the smoothness term, the curvature term and the data term are weightedly superimposed to form the overall objective function.

[0131] In step 250, the total gradient of the total objective function with respect to the path point coordinates is calculated to guide the optimization direction.

[0132] In some embodiments of the present disclosure, step 250 may include step 251 and at least one of step 251 .

[0133] In step 251 , the smoothing term gradient, the curvature term gradient, and the data term gradient are calculated.

[0134] In some embodiments of the present disclosure, in step 251, calculating the smoothing term gradient may include: determining the smoothing term gradient of each path point according to the coordinates of each path point and a second-order difference matrix.

[0135] In some embodiments of the present disclosure, in step 251, calculating the smoothing term gradient may include: calculating the smoothing term gradient according to formula (10)

[0136]

[0137] In formula (10), D is the second-order difference matrix, which penalizes the second-order derivative of the path. It is the smoothing gradient, which promotes the uniform distribution of path points, reduces the sudden change of curvature, and suppresses the "sawtooth" fluctuation in the path.

[0138] In some embodiments of the present disclosure, in step 251, calculating the curvature term gradient may include: determining the curvature term gradient of each path point based on the curvature of each path point, the derivative of the curvature with respect to the path point coordinates, the sign function of the curvature, and a predetermined curvature threshold.

[0139] In some embodiments of the present disclosure, in step 251, calculating the curvature term gradient may include: calculating the curvature term gradient according to formula (11)

[0140]

[0141] In formula (11), k max The maximum curvature that the vehicle can safely perform is determined by the minimum turning radius. i Represents the curvature value at each path point i. The curvature gradient at each path point i can be calculated by the chain rule sign(k i ) is a symbolic function, in k i When it is greater than or equal to 0, sign(k i ) is equal to 1; in k i When it is less than 0, sign(k i ) is equal to -1.

[0142] In some embodiments of the present disclosure, the gradient is non-zero only when the curvature exceeds a threshold; in the curvature-exceeding region, the gradient points in the direction of reducing the curvature; at a right-angle turn, the gradient adjusts the path point position to reduce the curvature.

[0143] In some embodiments of the present disclosure, in step 251, calculating the data item gradient may include: determining the data item gradient of each path point according to the path point coordinates of the continuous path and the path point coordinates of the original discrete grid path.

[0144] In some embodiments of the present disclosure, in step 251, calculating the data item gradient may include: calculating the data item gradient according to formula (12)

[0145]

[0146] In formula (12), It is the data item gradient, which is used to pull the optimization path toward the initial path and prevent excessive smoothing from deviating from the global optimum.

[0147] In some embodiments of the present disclosure, w data When the weight is low, local adjustment is allowed, w data Strictly follow the initial path when the weight is high.

[0148] In step 252, the total gradient is determined based on the path smoothing coefficient, the smoothing term gradient, the curvature constraint coefficient, the curvature term gradient, the data fitting coefficient, and the data term gradient.

[0149] In some embodiments of the present disclosure, step 252 may include: calculating the total gradient function according to formula (13):

[0150]

[0151] In formula (13), the total gradient It is the weighted sum of the gradient of the smoothing term, the gradient of the curvature term, and the gradient of the data term.

[0152] In some embodiments of the present disclosure, each gradient component reflects the sensitivity of the corresponding target item to the position of the waypoint.

[0153] In step 300, a predetermined algorithm is used to iteratively optimize and solve the overall objective function to obtain a global optimization path for the scraper.

[0154] In some embodiments of the present disclosure, step 300 may include: iteratively optimizing the path point coordinates, minimizing the objective function, and generating an optimized path point set based on the input initial path points, the total objective function, and the total gradient.

[0155] In some embodiments of the present disclosure, step 300 may include: constructing a local quadratic model in each iteration, limiting the step size to within the trust region radius; using the trust region approximation to solve the optimal step size of the quadratic model; and dynamically adjusting the trust region radius based on the descent ratio between the model prediction and the actual function.

[0156] In some embodiments of the present disclosure, step 300 may include: performing iterative optimization to solve the objective function based on a trust region algorithm. The algorithm approximates the objective function through a local quadratic model and solves the optimal step size within the trust region radius, combining explicit gradients (such as curvature gradients) with the objective function. With smooth gradient ) to achieve efficient convergence.

[0157] The above-mentioned embodiments of the present disclosure can efficiently process non-convex objective functions and ensure that the curvature constraints are strictly satisfied.

[0158] In some embodiments of the present disclosure, Figure 2 As shown, step 300 may include at least one of steps 310 to 330 .

[0159] In step 310, the objective function is iteratively optimized and solved based on the trust region algorithm.

[0160] In step 320 , it is determined whether the objective function has converged. If the objective function has converged, step 330 is executed; otherwise, if the objective function has not converged, step 330 is continued.

[0161] In step 330 , the optimized path (x_opt, y_opt) is output.

[0162] The core advantage of the above-described embodiments of the present disclosure lies in its rigorous handling of non-convex curvature constraints, preventing traditional gradient descent from falling into local optima. At each iteration, the algorithm updates the pathpoint coordinates and dynamically adjusts the trust region radius until the rate of change of the objective function value or the gradient norm falls below a preset threshold. The resulting output path strikes a balance between curvature continuity, smoothness, and global feasibility, meeting the practical motion requirements of equipment such as underground loaders.

[0163] The above-mentioned embodiment of the present disclosure adopts a multi-stage path planning architecture, based on the initial modified A* algorithm grid path (obstacles on both sides are expanded to achieve safety redundancy settings), using cubic spline interpolation, selecting an appropriate number of interpolations (N / 2), and converting the discrete grid path into a continuously differentiable initial path, and then performing final optimization.

[0164] The above-mentioned embodiment of the present disclosure establishes an objective function based on the smoothness term, curvature term and data term and establishes corresponding gradient terms respectively. Finally, the sensitivity of each gradient component to the position of the path point is represented by the weighted sum to achieve the adjustment of the path point. Finally, the trust region algorithm is used to iteratively fit the optimized path.

[0165] The above-mentioned embodiments of the present disclosure provide a global path planning and optimization method for a scraper loader suitable for underground mining operation environments. The method solves the comprehensive optimization problems of path curvature constraints, kinematic feasibility and safety in complex environments by combining initial path planning based on improved A* grid search with path smoothing technology based on nonlinear optimization.

[0166] The global path optimization of the above-mentioned embodiment of the present disclosure aims to convert a discrete grid path into a smooth continuous trajectory that conforms to the vehicle kinematic constraints. Its core process can be divided into three steps: path continuity processing (step 100), multi-objective optimization modeling (step 300) and nonlinear constraint solving (step 300).

[0167] The above-mentioned embodiments of the present disclosure, based on the optimization framework of the trust region algorithm, iteratively approximate the objective function through a local quadratic model, combined with explicit gradient calculation, to achieve rapid convergence while maintaining curvature constraints. Furthermore, the above-mentioned embodiments of the present disclosure address the numerical instability caused by the denominator approaching zero in the curvature calculation by introducing a regularization term design, thereby enhancing the robustness of the algorithm.

[0168] The above-mentioned embodiments of the present disclosure provide a path planning optimization method based on an underground scraper in a tunnel, which has the following technical effects.

[0169] The above-mentioned embodiments of the present disclosure strictly ensure the path curvature constraint, avoid sharp turns, and introduce a curvature penalty term into the optimization objective function, thereby imposing a penalty on the excess area and forcing the optimized path to satisfy the kinematic constraints.

[0170] The above-mentioned embodiments of the present disclosure improve path smoothness, reduce energy consumption and mechanical wear, and can reduce curvature mutations by penalizing the second-order derivative (curvature change rate) of the path.

[0171] The above embodiments of the present disclosure can efficiently process nonlinear optimization problems, thereby avoiding local optimal traps. The above embodiments of the present disclosure can still converge to a globally feasible solution in complex terrains (such as narrow alleys).

[0172] Figure 3 Schematic diagram of some embodiments of the global path optimization device for scraper disclosed in the present invention. Figure 3 As shown, the global path optimization device for a scraper disclosed herein may include a path continuity module 31 , an objective function construction module 32 and a path optimization module 33 .

[0173] The path continuation module 31 is configured to convert the original discrete grid path into a continuous path that complies with the kinematic constraints of the scraper.

[0174] In some embodiments of the present disclosure, the path continuation module 31 may be configured to perform a thinning process on the original discrete grid path through cubic spline interpolation to generate a continuous path that complies with the kinematic constraints of the scraper.

[0175] In some embodiments of the present disclosure, the path continuation module 31 may be configured to perform a thinning process on the original discrete grid path by using cubic spline interpolation and using half of the number of original path points as interpolation points.

[0176] The objective function construction module 32 is configured to construct a total objective function of the continuous path according to the continuous path, wherein the total objective function includes a smoothness term, a curvature term and a data term.

[0177] In some embodiments of the present disclosure, the objective function construction module 32 can be configured to construct a smoothing term, wherein the smoothing term is used to minimize the path curvature change rate through a second-order difference matrix so that the path points are evenly distributed; obtain a predetermined curvature threshold based on the minimum turning radius of the scraper; construct a curvature term, wherein the curvature term is used to impose a penalty on the path segment whose path point curvature is greater than the predetermined curvature threshold through a piecewise penalty function; construct a data term, wherein the data term is used to constrain the degree of deviation between the continuous path and the original discrete grid path; and determine the total objective function based on the smoothing term, the curvature term and the data term.

[0178] In some embodiments of the present disclosure, the objective function construction module 32 can be configured to determine the overall objective function based on the path smoothing coefficient, smoothing term, curvature constraint coefficient, curvature term, data fitting coefficient and data item when determining the overall objective function based on the smoothing term, curvature term and data item.

[0179] In some embodiments of the present disclosure, the objective function construction module 32 can also be configured to determine the first coordinate and the second coordinate of the path point of the shovel loader moving in the plane coordinate system; obtain the first-order derivative and the second-order derivative of the first coordinate with respect to the second coordinate, and the first-order derivative and the second-order derivative of the second coordinate with respect to the first coordinate; determine the curvature of the path point based on the first-order derivative and the second-order derivative of the first coordinate with respect to the second coordinate, and the first-order derivative and the second-order derivative of the second coordinate with respect to the first coordinate.

[0180] The path optimization module 33 is configured to use a predetermined algorithm to iteratively optimize and solve the overall objective function to obtain a global optimized path of the scraper.

[0181] In some embodiments of the present disclosure, the path optimization module 33 can be configured to calculate the total gradient of the total objective function with respect to the path point coordinates to guide the optimization direction; iteratively optimize the path point coordinates, minimize the objective function, and generate an optimized path point set based on the input initial path points, total objective function, and total gradient.

[0182] In some embodiments of the present disclosure, the path optimization module 33 can be configured to construct a local quadratic model in each iteration, limit the step size to the trust region radius, use the trust region approximation to solve the optimal step size of the quadratic model, and dynamically adjust the trust region radius according to the descent ratio between the model prediction and the actual function, when it iteratively optimizes the path point coordinates, minimizes the objective function, and generates an optimized path point set based on the input initial path points, the total objective function, and the total gradient.

[0183] In some embodiments of the present disclosure, the path optimization module 33 can be configured to calculate the smoothing term gradient, the curvature term gradient and the data term gradient when calculating the total gradient of the total objective function with respect to the coordinates of the path points; and determine the total gradient based on the path smoothing coefficient, the smoothing term gradient, the curvature constraint coefficient, the curvature term gradient, the data fitting coefficient and the data term gradient.

[0184] In some embodiments of the present disclosure, when calculating the smoothing term gradient, the path optimization module 33 may be configured to determine the smoothing term gradient of each path point based on the coordinates of each path point and the second-order difference matrix.

[0185] In some embodiments of the present disclosure, when calculating the gradient of the curvature term, the path optimization module 33 can be configured to determine the gradient of the curvature term for each path point based on the curvature of each path point, the derivative of the curvature with respect to the coordinates of the path point, the sign function of the curvature, and a predetermined curvature threshold.

[0186] In some embodiments of the present disclosure, when calculating the data item gradient, the path optimization module 33 may be configured to determine the data item gradient of each path point based on the path point coordinates of the continuous path and the path point coordinates of the original discrete grid path.

[0187] In some embodiments of the present disclosure, the global path optimization device for a scraper loader of the present disclosure can also be configured to search for the shortest feasible path from a starting point to an end point in a grid map, wherein the shortest feasible path is an initial path composed of a plurality of grid nodes; based on the shortest feasible path, the expansion area of ​​obstacles in the global map is added to form a discrete grid path that conforms to the original.

[0188] In some embodiments of the present disclosure, the scraper global path optimization device of the present disclosure may also be configured to execute the scraper global path optimization method described in any of the above embodiments of the present disclosure.

[0189] The above-mentioned embodiments of the present disclosure provide a scraper and a global path optimization method, device and system thereof.

[0190] The above-mentioned embodiments of the present disclosure solve the problem of path curvature discontinuity: the above-mentioned embodiments of the present disclosure construct a multi-objective optimization function by introducing a second-order differential smoothing term, a curvature constraint term and a data term, thereby forcing the path curvature to be continuous and strictly lower than the vehicle kinematic limit, thereby avoiding mechanical damage caused by sharp turns or path control failure caused by exceeding the turning limit of the engineering machinery vehicle itself.

[0191] The above-mentioned embodiments of the present disclosure achieve a balance between global optimality and local smoothness: the above-mentioned embodiments of the present disclosure design data item weight constraints to optimize the degree of deviation between the path and the initial improved A* grid path, thereby ensuring that the optimized path is close to the global optimal grid route and meets the smoothness and safety requirements.

[0192] The above embodiments of the present disclosure enhance the adaptability to dynamic environments: The above embodiments of the present disclosure improve the adaptability to dynamic environments by updating the curvature constraint weights (such as w curvature Dynamically adjusted with the path curvature), an exponential penalty can be imposed on high curvature areas, so that the above embodiments of the present disclosure can adapt to sudden working conditions such as tunnel collapse, temporary obstacles, etc.

[0193] Figure 4 Schematic diagram of the structure of some other embodiments of the global path optimization device for scraper disclosed in the present invention. Figure 4 As shown, the global path optimization device for a scraper includes a memory 41 and a processor 42 .

[0194] The memory 41 is used to store instructions. The processor 42 is coupled to the memory 41. The processor 42 is configured to execute the scraper global path optimization method described in any embodiment of the present disclosure based on the instructions stored in the memory.

[0195] like Figure 4 As shown, the scraper global path optimization device further includes a communication interface 43 for exchanging information with other devices. Furthermore, the scraper global path optimization device further includes a bus 44 through which the processor 42, the communication interface 43, and the memory 41 communicate with each other.

[0196] Memory 41 may include high-speed RAM memory or non-volatile memory, such as at least one disk storage device. Memory 41 may also be a memory array. Memory 41 may also be divided into blocks, and the blocks may be combined into virtual volumes according to certain rules.

[0197] Furthermore, the processor 42 may be a central processing unit (CPU), or may be an application-specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the present disclosure.

[0198] The above-mentioned embodiments of the present disclosure utilize a multi-stage path planning architecture: The above-mentioned embodiments of the present disclosure first generate an initial grid path (modified A* algorithm), and then perform cubic spline interpolation at half the path length to convert it into a continuously differentiable path. In the above-mentioned embodiments of the present disclosure, interpolation with an excessively small path length can result in the optimized curve being unconstrained at the starting and ending points, while interpolation with an excessively large path length can increase the constraints on the optimized curve at turns, resulting in increased changes in the curve curvature.

[0199] The above-mentioned embodiments of the present disclosure utilize curvature-constrained mathematical modeling and gradient calculation: They construct an overall objective optimization function that integrates smoothing terms, curvature constraint terms, and data terms to jointly balance the curve optimization results. The above-mentioned embodiments of the present disclosure activate penalties when curvature exceeds a limit, thereby avoiding excessive restriction of path degrees of freedom. The above-mentioned embodiments of the present disclosure explicitly calculate the gradient of curvature with respect to path point coordinates through the chain rule, replacing the traditional numerical difference method in related technologies, thereby improving accuracy and efficiency.

[0200] Figure 5 Schematic diagram of some embodiments of the global path optimization system for scrapers disclosed in the present invention. Figure 5 As shown, the scraper global path optimization system disclosed in the present invention may include a data acquisition device 51, a data preprocessing device 52, a scraper global path optimization device 53, a control and execution device 54 and a communication and computing platform 55.

[0201] The data acquisition device 51 may include a laser radar (which acquires three-dimensional point cloud data of the environment through multi-beam laser scanning, which is used to construct a high-precision grid map and detect the position and shape of obstacles in real time), a visual sensor (camera), an inertial measurement unit (IMU) and GPS / RTK, an encoder, a wheel speed meter and other measurement equipment.

[0202] The data preprocessing device 52 is configured to implement functions such as point cloud denoising and registration, raster / semantic map construction (converting sensor data into raster maps for A* path planning and semantic maps to distinguish roads, obstacles, and passable areas).

[0203] The global path optimization device 53 of the scraper is configured to realize global path generation (searching for the initial obstacle avoidance path in the grid map based on the improved A* algorithm and outputting a discrete node sequence), path continuity processing (through cubic spline interpolation fitting and sparse processing, the discrete path is converted into a continuous parameterized curve), multi-objective optimization modeling (designing an objective function including smoothing terms, curvature constraint terms and data terms, and using the trust region algorithm to iteratively optimize the coordinates of the path points to ensure the curvature continuity and close to the initial path), real-time obstacle avoidance re-planning and other functions.

[0204] In some embodiments of the present disclosure, the scraper global path optimization device 53 may be a scraper global path optimization device as described in any of the above embodiments.

[0205] The control and execution device 54 may include a trajectory tracking controller (which converts the optimized path into speed and steering instructions, and uses model predictive control (MPC) or PID algorithm to achieve precise tracking), a drive and steering actuator, etc.

[0206] The communication and computing platform 55 may include an industrial computer (running path optimization algorithm and real-time control logic), a power management system, etc.

[0207] In the global path optimization process of the above-mentioned embodiments of the present disclosure, the systems and physical hardware involved can be divided into multiple levels, covering the entire chain from data collection, processing, decision-making to execution.

[0208] According to another aspect of the present disclosure, a scraper is provided, comprising the scraper global path optimization system according to any one of the above embodiments.

[0209] According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the method for global path optimization of a scraper according to any one of the above embodiments is implemented.

[0210] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the method for global path optimization of a scraper as described in any of the above embodiments is implemented.

[0211] In some embodiments of the present disclosure, the computer-readable storage medium may be a non-transitory computer-readable storage medium.

[0212] Those skilled in the art will appreciate that embodiments of the present disclosure may be provided as methods, apparatus, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product implemented on one or more computer-usable non-transient storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0213] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0214] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0215] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0216] The shovel loader global path optimization device, path continuation module, objective function construction module, path optimization module, data preprocessing device, control and execution device, communication and computing platform described above can be implemented as a general-purpose processor, programmable logic controller (PLC), digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component or any appropriate combination thereof for performing the functions described in this application.

[0217] The present disclosure has been described in detail. To avoid obscuring the concept of the present disclosure, some details known in the art have not been described. Based on the above description, those skilled in the art can fully understand how to implement the technical solutions disclosed herein.

[0218] Those skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or may be accomplished by instructing the relevant hardware through a program, and the program may be stored in a non-transitory computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0219] The description of the present disclosure is provided for purposes of illustration and description and is not intended to be exhaustive or to limit the disclosure to the disclosed form. Many modifications and variations will be apparent to those skilled in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present disclosure and to enable those skilled in the art to understand the present disclosure and design various embodiments with various modifications suitable for specific applications.

Claims

1. A global path optimization method for a scraper, comprising: The original discrete grid path is converted into a continuous path that meets the kinematic constraints of the scraper; Constructing a total objective function of the continuous path according to the continuous path, wherein the total objective function includes a smoothing term, a curvature term, and a data term; The total objective function is iteratively optimized and solved using a predetermined algorithm to obtain a global optimization path for the scraper.

2. The global path optimization method for scraper according to claim 1, wherein: Converting the original discrete grid path to a continuous path that meets the kinematic constraints of the scraper involves: The original discrete grid path is thinned by cubic spline interpolation to generate a continuous path that meets the kinematic constraints of the scraper.

3. The global path optimization method for scraper according to claim 1, wherein: The thinning process of the original discrete grid path by cubic spline interpolation includes: The original discrete grid path is thinned by cubic spline interpolation using half the number of original path points as the number of interpolation points.

4. The global path optimization method for a scraper according to any one of claims 1 to 3, wherein: According to the continuous path, constructing the overall objective function of the continuous path includes: Constructing a smoothing term, wherein the smoothing term is used to minimize the rate of change of path curvature through a second-order difference matrix so that the path points are evenly distributed; Obtaining a predetermined curvature threshold value according to a minimum turning radius of the scraper; Constructing a curvature term, wherein the curvature term is used to impose a penalty on a path segment whose path point curvature is greater than a predetermined curvature threshold through a piecewise penalty function; Constructing a data item, wherein the data item is used to constrain the degree of deviation between the continuous path and the original discrete grid path; The overall objective function is determined based on the smoothness term, the curvature term and the data term.

5. The global path optimization method for scraper according to claim 4, wherein: Determining the overall objective function based on the smoothness term, the curvature term, and the data term includes: The overall objective function is determined based on the path smoothing coefficient, the smoothing term, the curvature constraint coefficient, the curvature term, the data fitting coefficient and the data term.

6. The global path optimization method for scraper according to claim 4, wherein: Also includes: Determine the first coordinate and the second coordinate of the path point of the scraper moving in the plane coordinate system; Obtaining the first-order derivative and the second-order derivative of the first coordinate with respect to the second coordinate, and the first-order derivative and the second-order derivative of the second coordinate with respect to the first coordinate; The curvature of the waypoint is determined based on the first and second derivatives of the first coordinate with respect to the second coordinate, and the first and second derivatives of the second coordinate with respect to the first coordinate.

7. The global path optimization method for a scraper according to any one of claims 1 to 3, wherein: The iterative optimization and solving of the overall objective function using a predetermined algorithm includes: Calculate the total gradient of the total objective function with respect to the path point coordinates to guide the optimization direction; Iteratively optimize the path point coordinates, minimize the objective function, and generate the optimized path point set based on the input initial path point, total objective function, and total gradient.

8. The global path optimization method for scraper according to claim 7, wherein: Iteratively optimize the path point coordinates to minimize the objective function. Based on the input initial path points, total objective function, and total gradient, the optimized path point set is generated, including: A local quadratic model is constructed in each iteration, limiting the step size to within the trust region radius; Use trust region approximation to solve the optimal step size for quadratic models; The trust region radius is dynamically adjusted based on the drop ratio between the model prediction and the actual function.

9. The global path optimization method for scraper according to claim 7, wherein: Calculating the total gradient of the total objective function with respect to the path point coordinates includes: Calculate the gradient of the smoothing term, the gradient of the curvature term, and the gradient of the data term; The total gradient is determined according to the path smoothing coefficient, the smoothing term gradient, the curvature constraint coefficient, the curvature term gradient, the data fitting coefficient and the data term gradient.

10. The global path optimization method for scraper according to claim 9, wherein: Calculating the gradient of the smoothing term involves: The smoothing term gradient of each path point is determined based on the coordinates of each path point and the second-order difference matrix.

11. The global path optimization method for scraper according to claim 9, wherein: Calculating the curvature gradient includes: The curvature term gradient of each path point is determined according to the curvature of each path point, the derivative of the curvature with respect to the coordinates of the path point, the sign function of the curvature, and a predetermined curvature threshold.

12. The global path optimization method for scraper according to claim 9, wherein: Calculating the gradient of a data item involves: The data item gradient of each path point is determined according to the path point coordinates of the continuous path and the path point coordinates of the original discrete grid path.

13. The global path optimization method for a scraper according to any one of claims 1 to 3, further comprising: Searching for the shortest feasible path from a starting point to an end point in a grid map, wherein the shortest feasible path is an initial path composed of a plurality of grid nodes; The expansion area of ​​the obstacle in the global map is increased based on the shortest feasible path to form a discrete grid path that conforms to the original one.

14. A global path optimization device for a scraper, comprising: a path continuation module configured to convert an original discrete raster path into a continuous path that complies with kinematic constraints of the scraper; an objective function construction module, configured to construct a total objective function of the continuous path according to the continuous path, wherein the total objective function includes a smoothing term, a curvature term, and a data term; The path optimization module is configured to use a predetermined algorithm to iteratively optimize and solve the overall objective function to obtain a global optimized path of the scraper.

15. A global path optimization device for a scraper, comprising: Memory; as well as A processor coupled to the memory, wherein the processor is configured to execute the scraper global path optimization method according to any one of claims 1 to 13 based on instructions stored in the memory.

16. A global path optimization system for a scraper, comprising a data acquisition device and the global path optimization device for a scraper according to claim 14 or 15.

17. A scraper, comprising the scraper global path optimization system according to claim 16.

18. A computer-readable storage medium, wherein: The computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the method for optimizing the global path of a scraper according to any one of claims 1 to 13 is implemented.

19. A computer program product comprising a computer program, wherein When the computer program is executed by a processor, the global path optimization method for a scraper according to any one of claims 1 to 13 is implemented.

Images