Quadruped robot task allocation method, device and equipment and medium

Abstract

The invention discloses a quadruped robot task allocation method, device and equipment and a medium, and relates to the technical field of task allocation, the method aims at the terrain adaptability defect in quadruped robot application and aims at solving the problem of terrain element deficiency during multi-machine cooperative task planning in an unstructured environment, and a terrain assessment matrix (TAS) is constructed, so that the task allocation efficiency is improved. And fusing the terrain constraint into the task allocation model. On the basis of a traditional unmanned aerial vehicle model, terrain evaluation parameters are added, the adaptability and operation efficiency of the system in an actual scene are improved, and a prediction result is closer to actual application. Meanwhile, a mixed heuristic strategy is adopted, the global search capability of the genetic algorithm and the local optimization advantage of tabu search are combined, repeated search and invalid area exploration are effectively avoided through a double-tabu table mechanism, the search efficiency is improved, and stable solving efficiency can still be kept when the task scale is expanded.

Description

Technical Field

[0001] The present invention relates to the technical field of task allocation, and in particular to a quadruped robot task allocation method, device, equipment and medium combined with terrain assessment. Background Art

[0002] As the scale of infrastructure in the energy, transportation, and industrial sectors continues to expand, the demand for autonomous, high-precision, and all-weather inspections in scenarios such as power equipment inspections, oil and gas pipeline monitoring, and disaster site investigations is becoming increasingly urgent. Traditional manual inspections have problems such as high costs and high risks (such as working in hot / cold areas, toxic environments, and confined spaces). In recent years, quadruped robots have demonstrated unique advantages in scenarios such as substation equipment inspections, oil and gas pipeline leak detection, and earthquake disaster site investigations, thanks to their excellent terrain adaptability. However, it is difficult for a single robot to simultaneously meet the requirements of large-scale coverage and multi-tasking parallelism. Therefore, multi-robot collaborative systems have begun to attract attention, and the design of task allocation algorithms is the core link in improving system performance.

[0003] At present, the research results on cluster task allocation are mainly concentrated on drone swarms, unmanned vehicle swarms, etc., and most of them adopt task allocation frameworks based on heuristic rules, market auction mechanisms or swarm intelligence optimization. Certain progress has been made in top-level design, theoretical research, project demonstration, and key technology research.

[0004] Most of the existing technical solutions are based on the field of drone swarms. They quantify and integrate key parameters such as path length, time cost, completion benefit, and damage probability to build a multi-dimensional weighted evaluation system. At the same time, they incorporate constraints such as drone endurance and payload capacity to establish a model for multi-drone task allocation problems.

[0005] Based on the above model, the discrete particle swarm optimization (DPSO) algorithm is then used to solve the problem. The particle encoding rules are used to represent the execution plan and execution order of the tasks. The particle swarm algorithm is used for iterative optimization, and the evaluation function is used to evaluate the quality of the searched solutions.

[0006] As can be seen, the UAV-based swarm task allocation model constructs an evaluation system using common parameters such as path length and time cost. However, this is constructed in an ideal airspace with no terrain or obstacles. In a quadruped robot operation scenario, it is necessary to consider the rigid constraints imposed by the spatial distribution of obstacles on the ground environment and the influence of terrain complexity. This results in deviations between the operation process of quadruped robot swarms in actual terrain and the ideal prediction. Summary of the Invention

[0007] In view of the above problems, the present invention provides a quadruped robot task allocation method, device, equipment and medium for overcoming the above problems or at least partially solving the above problems.

[0008] The present invention provides the following solutions:

[0009] A quadruped robot task allocation method, comprising:

[0010] By constructing a grid map from discretized geographic information, complex terrain is abstracted into a matrix model that can be quantified and analyzed; a terrain assessment matrix is ​​constructed based on the matrix model; the terrain assessment matrix includes a slope matrix, a surface friction coefficient matrix, and a stair height matrix;

[0011] Based on the quadruped robot body design constraints and task scenario requirements, a task allocation optimization model is constructed; the task allocation optimization model uses the terrain assessment matrix as an environmental map input for the task scenario and performs path planning based on the environmental map and the coordinates of each task using an A* algorithm combined with the terrain assessment matrix;

[0012] The encoding method based on the real number encoding of a single chromosome is used to encode the problem so that the solution to the problem can be represented as a chromosome in the genetic algorithm;

[0013] A hybrid heuristic optimization strategy is adopted to solve the task allocation optimization model by combining a genetic algorithm and a taboo search algorithm. The optimal solution obtained under the maximum number of iterations is the optimal task allocation scheme for the quadruped robot. The genetic algorithm is used for global search, and the taboo search algorithm is used as a local optimizer to correct the genetic algorithm from falling into the local optimal solution through a dynamic taboo table and a neighborhood perturbation mechanism.

[0014] Preferably, the terrain assessment matrix is ​​calculated as follows:

[0015]

[0016] Among them, α, β, and γ are slope weight, ground friction coefficient weight, and step height weight, respectively. s(m,n) represents the terrain slope within the unit. max represents the maximum slope that the quadruped robot can pass, f(m,n) represents the standard friction coefficient of the terrain material, h(m,n) represents the step height, h max Indicates the maximum stair height that a quadruped robot can pass through.

[0017] Preferably, the slope matrix s(m,n) is calculated as follows:

[0018] Calculate the surface curvature by spatial difference based on the digital elevation model;

[0019] The elevation matrix is ​​convolved using the Sobel operator to obtain horizontal and vertical gradients, and the average slope angle of each grid cell is calculated.

[0020] The friction coefficient matrix f(m,n) is calculated as follows:

[0021] Preset the basic friction value μ0 of different materials in a dry state, and correct the friction coefficient based on the motion data of the quadruped robot during movement;

[0022] The calculation method of the step height matrix h(m,n) includes:

[0023] Based on the prior terrain information, the stair height matrix of the environment map is established.

[0024] Preferably: the A* algorithm combined with TAS parameters is expressed as:

[0025] f(n)=g(n)×(1+δ·TAS)+h(n)

[0026] Where: f(n) is the estimate of each possible trial point, g(n) represents the cost from the starting search point to the current point, h(n) represents the estimate from the current node to the target node, and δ represents the terrain influence parameter.

[0027] Preferably, the allocation scheme is optimized by the following objective function during the solution process:

[0028]

[0029] Where: f s represents the path distance cost, f t Indicates the time cost, represents the distance standard deviation cost, represents the standard deviation cost of damage probability, f TAS Represents the terrain evaluation cost.

[0030] Preferably, the constraints in the task allocation optimization model include endurance constraints and load capacity constraints.

[0031] Preferably, when the taboo algorithm is combined with the genetic algorithm, optimizing the search process through a dual taboo table mechanism includes:

[0032] The legal solution taboo table dynamically records the effective solutions that have been explored and is updated according to the first-in-first-out principle;

[0033] The illegal solution taboo table permanently stores all illegal solutions, forming a global interception;

[0034] After a new solution is generated, it is first compared with the two tables. If the new solution is already in the illegal table or the legal table, it is eliminated and regenerated. Only the new solution that is not in the taboo table is subjected to constraint checking and fitness calculation.

[0035] When the termination condition is met, the algorithm terminates and outputs the solution with the best fitness, which is decoded as the optimal task allocation solution.

[0036] A quadruped robot task allocation device, used to execute the above quadruped robot task allocation method, the device comprising:

[0037] A terrain assessment matrix acquisition unit is used to construct a raster map by discretizing geographic information, abstracting complex terrain into a matrix model that can be quantified and analyzed; and constructing a terrain assessment matrix based on the matrix model; the terrain assessment matrix includes a slope matrix, a surface friction coefficient matrix, and a stair height matrix;

[0038] A task allocation optimization model construction unit is used to construct a task allocation optimization model based on the quadruped robot body design constraints and task scenario requirements; the task allocation optimization model uses the terrain assessment matrix as an environmental map input of the task scenario and performs path planning based on the environmental map and the coordinates of each task through an A* algorithm combined with the terrain assessment matrix;

[0039] an encoding unit, configured to perform encoding using a real number encoding method based on a single chromosome, so as to represent a solution to a problem as a chromosome in a genetic algorithm;

[0040] A hybrid heuristic optimization unit is used to adopt a hybrid heuristic optimization strategy, combining a genetic algorithm and a taboo search algorithm to solve a task allocation optimization model. The optimal solution obtained under the maximum number of iterations is the optimal task allocation plan of the quadruped robot; the genetic algorithm is used for global search, and the taboo search algorithm is used as a local optimizer to correct the genetic algorithm from falling into a local optimal solution through a dynamic taboo table and a neighborhood perturbation mechanism.

[0041] A quadruped robot task allocation device, the device comprising a processor and a memory:

[0042] The memory is used to store program code and transmit the program code to the processor;

[0043] The processor is used to execute the above-mentioned quadruped robot task allocation method according to the instructions in the program code.

[0044] A computer-readable storage medium is used to store program codes, and the program codes are used to execute the above-mentioned quadruped robot task allocation method.

[0045] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0046] The embodiments of the present application provide a method, apparatus, device, and medium for assigning tasks to a quadruped robot. This method incorporates terrain constraints into the task assignment model by constructing a terrain assessment matrix (TAS). Based on the traditional drone model, terrain assessment parameters are added to improve the system's adaptability and operating efficiency in actual scenarios, making the prediction results more relevant to practical applications. At the same time, a hybrid heuristic strategy is adopted, combining the global search capabilities of the genetic algorithm with the local optimization advantages of the taboo search. The dual taboo table mechanism effectively avoids repeated searches and invalid area exploration, thereby improving search efficiency and maintaining stable solution efficiency as the task scale expands.

[0047] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be derived from these drawings without inventive effort.

[0049] Figure 1 is a flow chart of a quadruped robot task allocation method provided by an embodiment of the present invention;

[0050] Figure 2 is a schematic diagram of terrain information provided by an embodiment of the present invention;

[0051] Figure 3 is a terrain assessment thermal map provided by an embodiment of the present invention;

[0052] Figure 4 Schematic diagram of real number encoding based on a single chromosome provided by an embodiment of the present invention;

[0053] Figure 5 This is a flow chart of a hybrid optimization strategy algorithm provided by an embodiment of the present invention;

[0054] Figure 6 is a schematic diagram of a quadruped robot task allocation device provided by an embodiment of the present invention;

[0055] Figure 7 Schematic diagram of a quadruped robot task allocation device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of the present invention.

[0057] See also Figure 1 , is a quadruped robot task allocation method provided by an embodiment of the present invention, such as Figure 1 As shown, the method may include:

[0058] S101: constructing a grid map by discretizing geographic information, abstracting complex terrain into a matrix model that can be quantified and analyzed; constructing a terrain assessment matrix based on the matrix model; the terrain assessment matrix includes a slope matrix, a surface friction coefficient matrix, and a stair height matrix;

[0059] In specific implementation, the embodiment of the present application can provide the following calculation method for the terrain assessment matrix:

[0060]

[0061] Among them, α, β, and γ are slope weight, ground friction coefficient weight, and step height weight, respectively. s(m,n) represents the terrain slope within the unit. max represents the maximum slope that the quadruped robot can pass, f(m,n) represents the standard friction coefficient of the terrain material, h(m,n) represents the step height, h max Indicates the maximum stair height that a quadruped robot can pass through.

[0062] Furthermore, the slope matrix s(m,n) is calculated as follows:

[0063] Calculate the surface curvature by spatial difference based on the digital elevation model;

[0064] The elevation matrix is ​​convolved using the Sobel operator to obtain horizontal and vertical gradients, and the slope angle of each grid cell is calculated.

[0065] The friction coefficient matrix f(m,n) is calculated as follows:

[0066] Preset the basic friction value μ0 of different materials in a dry state, and correct the friction coefficient based on the motion data of the quadruped robot during movement;

[0067] The calculation method of the step height matrix h(m,n) includes:

[0068] Based on the prior terrain information, the stair height matrix of the environment map is established.

[0069] S102: Based on the quadruped robot body design constraints and task scenario requirements, construct a task allocation optimization model; the task allocation optimization model uses the terrain assessment matrix as the environment map input of the task scenario and performs path planning based on the environment map and the coordinates of each task using the A* algorithm combined with TAS;

[0070] In specific implementation, the embodiment of the present application can provide the A* path planning method combined with TAS parameters as follows:

[0071] f(n)=g(n)×(1+δ·TAS)+h(n)

[0072] Where: f(n) is the estimate of each possible trial point, g(n) represents the cost from the starting search point to the current point, h(n) represents the estimate from the current node to the target node, and δ represents the terrain influence parameter.

[0073] S103: Encoding is performed using a real number encoding method based on a single chromosome, so as to represent the solution to the problem as a chromosome in the genetic algorithm;

[0074] S104: A hybrid heuristic optimization strategy is adopted, combining the genetic algorithm and the taboo search algorithm to solve the task allocation optimization model. The optimal solution obtained under the maximum number of iterations is the optimal task allocation plan of the quadruped robot; the genetic algorithm is used for global search, and the taboo search algorithm is used as a local optimizer, so as to correct the genetic algorithm from falling into the local optimal solution through the dynamic taboo table and the neighborhood perturbation mechanism.

[0075] During the solution process, the allocation scheme is optimized through the following objective function:

[0076]

[0077] Where: f s represents the path distance cost, f t Indicates the time cost, represents the distance standard deviation cost, represents the standard deviation cost of damage probability, f TAS Represents the terrain evaluation cost.

[0078] When the tabu algorithm is combined with the genetic algorithm, the search process is optimized through the dual tabu table mechanism, including:

[0079] The legal solution taboo table dynamically records the effective solutions that have been explored and is updated according to the first-in-first-out principle;

[0080] The illegal solution taboo table permanently stores all illegal solutions, forming a global interception;

[0081] After a new solution is generated, it is first compared with the two tables. If the new solution is already in the illegal table or the legal table, it is eliminated and regenerated. Only the new solution that is not in the taboo table is subjected to constraint checking and fitness calculation.

[0082] When the termination condition is met, the algorithm terminates and outputs the solution with the best fitness, which is decoded as the optimal task allocation solution.

[0083] The quadruped robot task allocation method provided in the embodiments of this application addresses the terrain adaptability deficiencies of existing cluster task allocation technologies in quadruped robot applications, and is dedicated to resolving the issue of missing terrain elements when planning multi-machine collaborative tasks in unstructured environments. By incorporating terrain assessment parameters into the traditional evaluation system framework, incorporating environmental factors such as the slope matrix, the surface friction coefficient matrix, and the stair height matrix, and employing a hybrid heuristic optimization strategy combining a genetic algorithm and a tabu search algorithm, the traditional cluster task allocation model is adapted to the unique terrain adaptability requirements of quadruped robots, making the model more adaptable to specific scenarios.

[0084] Building on the traditional UAV model, terrain assessment parameters were added to improve the system's adaptability and efficiency in real-world scenarios, making predictions more relevant to practical applications. A hybrid heuristic strategy was also employed, combining the global search capabilities of a genetic algorithm with the local optimization advantages of tabu search. A dual tabu table mechanism effectively avoided repeated searches and exploration of invalid areas, improving search efficiency and maintaining stable solution efficiency as the task scale expanded.

[0085] The following is a detailed introduction to the quadruped robot task allocation method provided in the embodiments of the present application.

[0086] This method incorporates terrain assessment parameters, including a slope matrix, a surface friction matrix, and a stair height matrix, to construct a task allocation model for quadruped robots adapted to terrestrial scenarios. Through quantitative integration and optimization of terrain elements, this method enables efficient collaborative task planning for quadruped robots in complex terrain environments, improving the system's adaptability and operational efficiency in real-world scenarios.

[0087] At the same time, a hybrid heuristic optimization strategy is adopted, combining genetic algorithm and taboo search algorithm. Genetic algorithm is used for global search, and taboo search is introduced as a local optimizer: dynamic taboo table and neighborhood perturbation mechanism are used to correct the genetic algorithm from falling into the local optimal solution, thereby improving the convergence accuracy.

[0088] 1. Terrain Assessment

[0089] By constructing a raster map from discretized geographic information, complex terrain can be abstracted into a matrix model that can be quantitatively analyzed. Figure 2As shown in the figure, based on elevation data and surface materials, the system first identifies non-passable areas such as rivers, trees, buildings, and areas where parameters exceed safety thresholds, marks them as obstacle units, and generates a basic binary matrix.

[0090] On this basis, the Terrain Assessment Matrix (TAS) integrates the following multi-dimensional physical features:

[0091] (1) Slope matrix s(m,n).

[0092] Based on the digital elevation model (DEM), the surface curvature is calculated through spatial differencing. The elevation matrix is ​​convolved with the Sobel operator to obtain the horizontal and vertical gradients, and then the average slope angle of each grid cell is calculated.

[0093] (2) Friction coefficient matrix f(m,n).

[0094] The basic friction value μ0 of different materials (such as asphalt, grass, and sand) in a dry state is preset, and the friction coefficient is corrected based on the motion data of the robot during movement.

[0095] (3) Stair height matrix h(m,n).

[0096] Based on the prior terrain information, the stair height matrix of the environment map is established.

[0097] The above parameters are weighted and fused to form the terrain assessment matrix TAS(m,n), which is calculated as follows:

[0098]

[0099] Among them, α, β, and γ are the slope weight, friction coefficient weight, and stair height weight, respectively. This matrix can represent an ideal passageway that is flat, non-slip, and has continuous footholds.

[0100] like Figure 3 As shown in the figure, the TAS matrix heat map can intuitively display the degree of terrain structure, presented as a color gradient. Ideal paths (L1) represent flat and safe areas, poor paths (L2) are dangerous areas that require avoidance or special measures, and black (L3) indicates impassable areas. Embedding the terrain assessment matrix into the task allocation model can significantly improve the efficiency and safety of task execution.

[0101] 2. Model system construction

[0102] The research problem is described as follows: n quadruped robots need to collaborate to complete m inspection tasks within an area. To efficiently complete these tasks, tasks must be rationally assigned to the quadruped robots while satisfying constraints, ensuring timely execution, optimal resource utilization, and improved overall execution efficiency.

[0103] (1) Variable design

[0104] Define the set of quadruped robots as D = {D1, D2..., D n}, the task set is T={T1,T2...,T m}. The i-th task T i Represented by a five-element vector group in is the number of task i, is the location of task i, is the estimated time required for task i, is the resource consumption of task i, represents the damage probability of the quadruped robot performing this task.

[0105] The j-th quadruped robot D j Represented by a six-element vector group in is the number of the quadruped robot j, is the coordinate of the quadruped robot j, is the moving speed of quadruped robot j, is the flight time of quadruped robot j, is the load capacity of quadruped robot j, is the task sequence of quadruped robot j.

[0106] The terrain assessment matrix TAS(m,n) is used as the environment map input for the mission scenario. Based on the environment map and the coordinates of each task, the path planning is performed by combining the A* algorithm with TAS. The influence of the terrain assessment value is included in the calculation of the distance cost. δ is the terrain influence parameter, as shown in the following formula:

[0107] f(n)=g(n)×(1+δ·TAS)+h(n)

[0108] Where f(n) is the estimated value for each possible trial point, g(n) represents the cost from the starting search point to the current point (usually expressed as the depth of a node in the search tree), and h(n) represents the estimated value from the current node to the target node. The actual distance costs and paths between tasks are obtained, and this process is repeated to obtain the distance cost matrix S(m,n) and the path matrix P(m,n).

[0109] (2) Model design

[0110] During the solution process, the allocation scheme is optimized through the objective function, which consists of five parts:

[0111]

[0112] The following formulas are used to obtain (where is the parameter of the corresponding item):

[0113] A. Path distance cost f s Calculated by the following formula.

[0114]

[0115] Where, S(t i,j ,t i+1,j ) is the path distance cost.

[0116] B. Time cost f t Calculated by the following formula.

[0117]

[0118] Where, f t The longest time taken by the quadruped robot group was obtained. To implement the moving speed of the quadruped robot, is the time taken for task i.

[0119] C. Distance standard deviation cost Calculated by the following formula.

[0120]

[0121] Where, is the standard deviation of the task path distances of each group, The path distances of the four-legged robots can be made as balanced as possible.

[0122] D. Damage probability standard deviation cost Calculated by the following formula.

[0123]

[0124] Where k i is the damage probability of executing task i, is the average damage probability of all tasks, The damage probability of each quadruped robot can be balanced, and there will be no situation where the damage probability of a certain group is particularly high.

[0125] E. Terrain assessment cost f TAS Calculated by the following formula.

[0126]

[0127] The constraints defined in the model are as follows:

[0128] Endurance constraint: the total time of each group of tasks cannot exceed the endurance of the quadruped robot, which is expressed as:

[0129]

[0130] Load capacity constraint: the total resource consumption of each group of tasks must not exceed the load capacity of the quadruped robot, which is expressed as:

[0131]

[0132] 3. Hybrid heuristic optimization strategy

[0133] (1) Encoding method

[0134] Encoding is the process of representing the solution of a problem as a chromosome in a genetic algorithm. Using a real number encoding method based on a single chromosome, any task allocation solution can be represented by a 1-dimensional sequence consisting of n-1 0s and integer sequences from 1 to m, where 0 represents the dividing point and the integer sequence from 1 to m represents the task and its order, such as Figure 4 shown.

[0135] (2) Hybrid Optimization Strategy

[0136] A hybrid heuristic optimization strategy is adopted, combining genetic algorithm and tabu search algorithm. Genetic algorithm is used for global search, and tabu search is introduced as a local optimizer. Dynamic tabu table and neighborhood perturbation mechanism are used to correct the genetic algorithm from falling into the local optimal solution, thereby improving the convergence accuracy.

[0137] A genetic algorithm (GA) is a search heuristic algorithm that mimics the principles of natural selection and genetics. It iteratively searches for the optimal solution through selection, crossover, and mutation, and is used to solve optimization and search problems. In a genetic algorithm, each candidate solution is an individual, and a group of candidate solutions is called a population. The process of continuously generating new individuals is the process of searching for the optimal solution.

[0138] When the taboo algorithm is combined with the genetic algorithm, the search process is optimized through a dual taboo table mechanism: the legal solution taboo table dynamically records the effective solutions that have been explored and is updated according to the first-in-first-out principle to avoid repeated searches; the illegal solution taboo table permanently stores all illegal solutions to form a global interception. After a new solution is generated, the dual tables are compared first. If the new solution is already in the illegal table or the legal table, the new solution is eliminated and regenerated. Only the new solution that is not in the taboo table is subject to constraint checking and fitness calculation. When the termination condition is met (the maximum number of iterations is reached or the fitness of the optimal individual in the population does not change after multiple consecutive iterations, that is, the population stops evolving), the algorithm terminates and outputs the solution with the best fitness, which is decoded in a manner opposite to the encoding as a task allocation plan. This strategy can significantly reduce repeated searches in complex task allocation, maintain population diversity, avoid invalid areas, and achieve a dual improvement in efficiency and solution quality. The algorithm process is as follows Figure 5 shown.

[0139] In summary, the quadruped robot task allocation method proposed in this application integrates terrain constraints into the task allocation model by constructing a terrain assessment matrix (TAS). Compared with traditional drone models, this improves the model's terrain adaptability and enhances mission safety. A hybrid heuristic strategy combines the global search capabilities of a genetic algorithm with the local optimization advantages of tabu search. Through a dual tabu table mechanism, it effectively avoids repeated searches and exploration of invalid areas, improving search efficiency and maintaining stable solution efficiency as the task scale increases.

[0140] See also Figure 6 , the embodiment of the present application can also provide a quadruped robot task allocation device, such as Figure 6 As shown, for executing the above-mentioned quadruped robot task allocation method, the device may include:

[0141] The terrain assessment matrix acquisition unit 601 is configured to construct a grid map by discretizing geographic information, abstracting the complex terrain into a matrix model that can be quantified and analyzed; and constructing a terrain assessment matrix based on the matrix model; the terrain assessment matrix includes a slope matrix, a surface friction coefficient matrix, and a stair height matrix.

[0142] The task allocation optimization model construction unit 602 is used to construct a task allocation optimization model based on the quadruped robot body design constraints and the task scenario requirements; the task allocation optimization model uses the terrain assessment matrix as the environment map input of the task scenario and performs path planning based on the environment map and the coordinates of each task using the A* algorithm combined with TAS;

[0143] An encoding unit 603 is configured to encode using a real number encoding method based on a single chromosome, so as to represent a solution to the problem as a chromosome in a genetic algorithm;

[0144] The hybrid heuristic optimization unit 604 is used to adopt a hybrid heuristic optimization strategy, combining the genetic algorithm and the taboo search algorithm to solve the task allocation optimization model. The optimal solution obtained under the maximum number of iterations is the optimal task allocation plan of the quadruped robot; the genetic algorithm is used to perform global search, and the taboo search algorithm is used as a local optimizer to correct the genetic algorithm from falling into the local optimal solution through a dynamic taboo table and a neighborhood perturbation mechanism.

[0145] The embodiment of the present application may further provide a quadruped robot task allocation device, the device comprising a processor and a memory:

[0146] The memory is used to store program code and transmit the program code to the processor;

[0147] The processor is used to execute the steps of the above-mentioned quadruped robot task allocation method according to the instructions in the program code.

[0148] like Figure 7 As shown, a quadruped robot task allocation device provided by an embodiment of the present application may include: a processor 10, a memory 11, a communication interface 12 and a communication bus 13. The processor 10, the memory 11, and the communication interface 12 all communicate with each other through the communication bus 13.

[0149] In the embodiment of the present application, the processor 10 may be a central processing unit (CPU), an application-specific integrated circuit, a digital signal processor, a field programmable gate array, or other programmable logic devices.

[0150] The processor 10 may call a program stored in the memory 11. Specifically, the processor 10 may execute the operations in the embodiment of the quadruped robot task allocation method.

[0151] The memory 11 is used to store one or more programs. The program may include program code, and the program code includes computer operating instructions. In the embodiment of the present application, the memory 11 stores at least a program for implementing the following functions:

[0152] By constructing a grid map from discretized geographic information, complex terrain is abstracted into a matrix model that can be quantified and analyzed; a terrain assessment matrix is ​​constructed based on the matrix model; the terrain assessment matrix includes a slope matrix, a surface friction coefficient matrix, and a stair height matrix;

[0153] Based on the quadruped robot body design constraints and task scenario requirements, a task allocation optimization model is constructed; the task allocation optimization model uses the terrain assessment matrix as the environment map input of the task scenario and performs path planning based on the environment map and the coordinates of each task through the A* algorithm combined with TAS;

[0154] The encoding method based on the real number encoding of a single chromosome is used to encode the problem so that the solution to the problem can be represented as a chromosome in the genetic algorithm;

[0155] A hybrid heuristic optimization strategy is adopted to solve the task allocation optimization model by combining a genetic algorithm and a taboo search algorithm. The optimal solution obtained under the maximum number of iterations is the optimal task allocation scheme for the quadruped robot. The genetic algorithm is used for global search, and the taboo search algorithm is used as a local optimizer to correct the genetic algorithm from falling into the local optimal solution through a dynamic taboo table and a neighborhood perturbation mechanism.

[0156] In addition, the memory 11 may include a high-speed random access memory and may also include a non-volatile memory, such as at least one disk storage device or other volatile solid-state storage device.

[0157] The communication interface 12 may be an interface of a communication module, and is used to connect to other devices or systems.

[0158] Of course, it needs to be explained that Figure 7 The structure shown does not constitute a limitation on the quadruped robot task allocation device in the embodiment of the present application. In actual application, the quadruped robot task allocation device may include Figure 7 More or fewer components than shown, or combinations of certain components.

[0159] An embodiment of the present application may also provide a computer-readable storage medium, wherein the computer-readable storage medium is used to store program code, and the program code is used to execute the steps of the above-mentioned quadruped robot task allocation method.

[0160] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0161] Through the description of the above embodiments, it can be seen that those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments of the present application.

[0162] Each embodiment in this specification is described in a progressive manner. The same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for system or system embodiments, since they are basically similar to method embodiments, the description is relatively simple. For relevant parts, refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without expending creative work.

[0163] The above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention are included in the scope of protection of the present invention.

Claims

1. A quadruped robot task allocation method, characterized in that: include: By discretizing geographic information to construct a raster map, complex terrain is abstracted into a matrix model that can be quantified and analyzed; Constructing and obtaining a terrain assessment matrix based on the matrix model; The terrain assessment matrix includes a slope matrix, a surface friction coefficient matrix and a stair height matrix; Based on the quadruped robot body design constraints and task scenario requirements, a task allocation optimization model is constructed; the task allocation optimization model uses the terrain assessment matrix as an environmental map input for the task scenario and performs path planning based on the environmental map and the coordinates of each task using an A* algorithm combined with the terrain assessment matrix; The encoding method based on the real number encoding of a single chromosome is used to encode the problem so that the solution to the problem can be represented as a chromosome in the genetic algorithm; A hybrid heuristic optimization strategy is adopted to solve the task allocation optimization model by combining a genetic algorithm and a taboo search algorithm. The optimal solution obtained under the maximum number of iterations is the optimal task allocation scheme for the quadruped robot. The genetic algorithm is used for global search, and the taboo search algorithm is used as a local optimizer to correct the genetic algorithm from falling into the local optimal solution through a dynamic taboo table and a neighborhood perturbation mechanism.

2. The quadruped robot task allocation method according to claim 1, characterized in that: The terrain assessment matrix is ​​calculated as follows: Among them, α, β, and γ are slope weight, ground friction coefficient weight, and step height weight, respectively. s(m,n) represents the terrain slope within the unit. max represents the maximum slope that the quadruped robot can pass, f(m,n) represents the standard friction coefficient of the terrain material, h(m,n) represents the step height, h max Indicates the maximum stair height that a quadruped robot can pass through.

3. The quadruped robot task allocation method according to claim 2, characterized in that: The calculation method of the slope matrix s(m,n) includes: Calculate the surface curvature by spatial difference based on the digital elevation model; The elevation matrix is ​​convolved using the Sobel operator to obtain horizontal and vertical gradients, and the average slope angle of each grid cell is calculated. The friction coefficient matrix f(m,n) is calculated as follows: Preset the basic friction value μ0 of different materials in a dry state, and correct the friction coefficient based on the motion data of the quadruped robot during movement; The calculation method of the step height matrix h(m,n) includes: Based on the prior terrain information, the stair height matrix of the environment map is established.

4. The quadruped robot path planning method according to claim 1, wherein the A* algorithm combined with TAS parameters is expressed as: f(n)=g(n)×(1+δ·TAS)+h(n) Where: f(n) is the estimate of each possible trial point, g(n) represents the cost from the starting search point to the current point, h(n) represents the estimate from the current node to the target node, and δ represents the terrain influence parameter.

5. The quadruped robot task allocation method according to claim 1, characterized in that: During the solution process, the allocation scheme is optimized through the following objective function: Where: f s represents the path distance cost, f t Indicates the time cost, represents the distance standard deviation cost, represents the standard deviation cost of damage probability, f TAS Represents the terrain evaluation cost.

6. The quadruped robot task allocation method according to claim 1, characterized in that: The constraints in the task allocation optimization model include endurance constraints and load capacity constraints.

7. The quadruped robot task allocation method according to claim 1, characterized in that: When the tabu algorithm is combined with the genetic algorithm, the search process is optimized through the dual tabu table mechanism, including: The legal solution taboo table dynamically records the effective solutions that have been explored and is updated according to the first-in-first-out principle; The illegal solution taboo table permanently stores all illegal solutions, forming a global interception; After a new solution is generated, it is first compared with the two tables. If the new solution is already in the illegal table or the legal table, it is eliminated and regenerated. Only the new solution that is not in the taboo table is subject to constraint checking and fitness calculation. When the termination condition is met, the algorithm terminates and outputs the solution with the best fitness, which is decoded as the optimal task allocation solution.

8. A quadruped robot task allocation device, characterized in that: The device is used to execute the quadruped robot task allocation method according to any one of claims 1 to 7, comprising: A terrain assessment matrix acquisition unit is used to construct a raster map by discretizing geographic information, abstracting complex terrain into a matrix model that can be quantified and analyzed; and constructing a terrain assessment matrix based on the matrix model; the terrain assessment matrix includes a slope matrix, a surface friction coefficient matrix, and a stair height matrix; A task allocation optimization model construction unit is used to construct a task allocation optimization model based on the quadruped robot body design constraints and task scenario requirements; the task allocation optimization model uses the terrain assessment matrix as an environmental map input of the task scenario and performs path planning based on the environmental map and the coordinates of each task through an A* algorithm combined with the terrain assessment matrix; an encoding unit, configured to perform encoding using a real number encoding method based on a single chromosome, so as to represent a solution to a problem as a chromosome in a genetic algorithm; A hybrid heuristic optimization unit is used to adopt a hybrid heuristic optimization strategy, combining a genetic algorithm and a taboo search algorithm to solve a task allocation optimization model. The optimal solution obtained under the maximum number of iterations is the optimal task allocation plan of the quadruped robot; the genetic algorithm is used for global search, and the taboo search algorithm is used as a local optimizer to correct the genetic algorithm from falling into a local optimal solution through a dynamic taboo table and a neighborhood perturbation mechanism.

9. A quadruped robot task allocation device, characterized in that: The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the quadruped robot task allocation method according to any one of claims 1 to 7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store program code, and the program code is used to execute the quadruped robot task allocation method according to any one of claims 1 to 7.

Images