Intelligent scheduling method and system for quadruped robot

Abstract

The invention relates to the field of robot scheduling management, in particular to a quadruped robot intelligent scheduling method and system, and the method comprises the steps: constructing and training a graph neural network; nodes of the graph neural network represent loading and unloading point coordinates, cargo weight and initial electric quantity of the robot; edges represent routes between nodes, and an adjacent matrix and a node feature matrix are input; the output is a power consumption prediction value and a task completion time prediction value; acquiring a task route, and selecting the robot at the node directly associated with the node through which the task route passes and the robot at the node through which the task route passes as a target robot; calculating the value of the target robot; and obtaining a priority coefficient of each task, taking the priority coefficient and the output value as an edge weight for matching the task and the robot, taking a task allocation scheme with the maximum edge weight accumulation sum as a matching scheme, and carrying out global matching on a plurality of robots and a plurality of tasks. The method has the effect of improving the task allocation flexibility.

Description

Technical Field

[0001] The present application relates to the field of robot scheduling management, and in particular to a method and system for intelligent scheduling of a quadruped robot. Background Art

[0002] In a factory or warehouse environment, when multiple quadruped robots perform tasks at multiple loading and unloading points, the robot's task allocation and scheduling management are important links. For example, in the Chinese patent application document with announcement number CN112967002B, the pickup task allocation method and its cargo sorting system disclosed, the pickup task allocation method includes: determining the priority of the pickup task according to the target cargo position of the pickup task; assigning at least one pickup task to the robot according to the current position of the robot and / or the priority of the pickup task; calculating the total proportion of the completed pickup task indicators to the total pickup task indicators; when the total proportion is greater than the preset first completion threshold, assigning the highest priority pickup task to the robot.

[0003] In multi-robot scenarios, existing solutions do not fully consider the collaborative work between robots. For example, when multiple robots approach high-priority tasks at the same time, task conflicts or resource waste may occur, and the flexibility of task allocation is insufficient. Summary of the invention

[0004] In order to solve the above-mentioned technical problem of low flexibility in task classification, the present application provides a quadruped robot intelligent scheduling method and system.

[0005] In the first aspect, the present application provides a method and system for intelligent scheduling of a quadruped robot, which adopts the following technical solutions: A method for intelligent scheduling of quadruped robots, comprising the steps of: constructing and training a graph neural network; the nodes of the graph neural network represent the coordinates of loading and unloading points, the weight of the cargo, and the initial power of the robot; the edges of the graph neural network represent the routes between the nodes, and the input of the graph neural network is an adjacency matrix and a node feature matrix; the output is a power consumption prediction value and a task completion time prediction value; obtaining a task route, selecting a robot at a node directly associated with a node passed by the task route and a robot at a node passed by the task route as a target robot; calculating the output value of the target robot; obtaining a priority coefficient of each task, and assigning the priority coefficient to the target robot; The offer value is used as the edge weight for matching tasks and robots, and the task allocation plan with the largest cumulative sum of edge weights is used as the matching plan. Multiple robots and multiple tasks are globally matched according to the KM matching algorithm. The offer value is calculated as follows: according to the power consumption prediction value, completion time prediction value and preset constraint function of each target robot under the same task, the offer value of each target robot for the same task is calculated, the power consumption prediction value is inversely proportional to the offer value, and the completion time prediction value is inversely proportional to the offer value. The constraint function expresses that when the power consumption prediction value is greater than the initial power of the target robot, the offer value of the robot is set to 0.

[0006] Each robot bids for each task, and the bid value reflects the robot's completion cost (power consumption to complete the task) and completion quality (time to complete the task). The lower the completion cost, the higher the completion quality, and the higher the bid value. At this time, the robot is more likely to match the task. At the same time, the value calculated by combining the constraint function can effectively avoid the task being assigned to a robot that cannot complete the task, making the bid value of a robot with insufficient power 0. The bid value and task priority are used as the edge weights for matching tasks with robots. Combined with the graph neural network to predict power consumption and completion time, global matching of multiple tasks and multiple robots is performed according to the matching algorithm. Task allocation can be dynamically adjusted according to real-time data, and task allocation is highly flexible; Through graph neural networks, robots and tasks can be modeled as nodes and edges in the graph. Node features and edge features are used to assign tasks, which can effectively handle the complex relationships in multi-robot task assignment and improve the efficiency and accuracy of task assignment. Graph neural networks can dynamically adjust task assignment strategies. For example, when the location of a task point or the state of a robot changes, tasks can be quickly reallocated. The global information processing capabilities of graph neural networks can achieve global optimization of multi-robot task assignment. The purpose of directly associating the nodes and robots that the task route passes through and selecting the target robot is to screen the robots and quickly locate the robots that are directly related to the task. When there are a large number of robots, this method can significantly reduce the screening time and the amount of calculation when matching subsequent robots with tasks, reduce the waste of computing resources, and improve computing efficiency.

[0007] Optionally, the expression for the bid value is: ;in, Indicates target robot for the task The output value, Indicates The exponential function with base , Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The predicted power consumption value is Indicates the preset time adjustment parameters. Indicates the preset power consumption adjustment parameters; represents the constraint function; Indicates The initial power of the target robot.

[0008] For a task, the smaller the robot's corresponding completion time prediction value and power consumption prediction value, the higher the robot's bid value for the task, and the higher the possibility that the robot matches the task. When the robot's power consumption prediction value exceeds its initial power, the bid value is directly reset to zero, ensuring that the robot will not be assigned to tasks beyond its capabilities. This not only avoids task failures due to insufficient power, but also improves the reliability of task assignment; setting multiple adjustment parameters can better adapt to different actual working conditions.

[0009] Optionally, the expression for the bid value is: ;in, Indicates target robot for the task The output value, Indicates The exponential function with base , Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The predicted power consumption value is Indicates the preset time adjustment parameters. Indicates the preset power consumption adjustment parameters; represents the constraint function; Indicates The initial power of the target robot; Indicates the preset safe power level; Indicates the preset power regulation parameters.

[0010] In actual working conditions, factory loading and unloading tasks usually require robots to complete tasks efficiently and return safely for charging to avoid shutdown due to depletion of power. Safety power is added to the calculation of constraints to reserve enough power for the robot to automatically return to the charging position.

[0011] Optionally, the expression for the sum of edge weights is: ; represents the cumulative sum of edge weights, Indicates target robot task The output value; Representation Task The priority coefficient of is the total number of target robots; Representation Task The total number of .

[0012] The priority coefficient of each task is multiplied by the corresponding robot's bid value, ensuring that high-priority tasks contribute more to the overall edge weight. If a task has a high priority, the robot's bid value assigned to the task will have a greater impact on the overall edge weight. The importance of the task and the robot's execution capability are considered comprehensively. This ensures that the task allocation plan not only considers the urgency of the task, but also the actual capabilities of the robot. The calculation method of the cumulative sum of edge weights allows the system to evaluate the quality of the task allocation plan from a global perspective. By maximizing the cumulative sum of edge weights, the optimal task allocation plan can be found to ensure the highest overall task execution efficiency.

[0013] Optionally, the graph neural network includes a shared feature extraction layer, a first output layer and a second output layer that are independent of each other, the shared feature extraction layer is connected to the first output layer and the second output layer respectively, the first output layer is used to output the power consumption prediction value, and the second output layer is used to output the completion time prediction value; the shared feature extraction layer includes: a graph convolution layer, which is used to extract local features of nodes; a graph attention layer, which assigns different weights to different neighboring nodes through an attention mechanism, and a gated graph neural network layer, which processes dynamic transmission between nodes through a gating mechanism; the output of the graph convolution layer serves as the input of the graph attention layer, and the output of the graph attention layer serves as the input of the gated graph neural network layer.

[0014] The role of the shared feature extraction layer is to extract common features of all nodes, which can be shared and utilized by subsequent different tasks (such as power consumption prediction and completion time prediction). The graph convolution layer extracts local features of nodes and can capture the attribute information of the nodes themselves; the graph attention layer assigns different weights to different neighboring nodes through the attention mechanism, so that the model can pay more attention to neighboring nodes that are more important to the current node, thereby better utilizing the relationship information between nodes. The gated graph neural network layer handles the dynamic transmission between nodes through the gating mechanism, which can better handle the long-term dependencies in the graph and avoid the loss of information during the transmission process. The combination of the shared feature extraction layer and the independent output layer enables the model to output power consumption prediction values ​​and completion time prediction values ​​at the same time. These two output layers correspond to different task requirements, which improves the versatility and adaptability of the model.

[0015] Optionally, the loss function for graph neural network training is: + ; In the formula, represents the loss function, Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The actual value of the completion time; Indicates the total number of target robots; Indicates Target robots complete the task The predicted power consumption value is Indicates The actual power consumption of the target robot.

[0016] This loss function takes into account the difference between the predicted completion time and the actual value, as well as the difference between the predicted power consumption and the actual value. In actual working conditions, accurately predicting the time and power consumption of the robot to complete the task is crucial for task allocation. By minimizing this loss function, the graph neural network can more accurately predict these two key indicators, thereby improving the reliability of task allocation.

[0017] Optionally, the task route is a node sequence, which represents the loading and unloading points that the robot needs to visit in sequence.

[0018] Representing the mission route as a sequence of nodes clearly defines the order in which the robot needs to visit.

[0019] Optionally, a path planning algorithm is used to calculate the shortest path between nodes as the route between the nodes.

[0020] By calculating the shortest path through the path planning algorithm, the robot can be ensured to take the shortest route when completing the task. In actual working conditions, the factory environment is complex and there are various obstacles and restrictions. The shortest path planning can reduce the robot's moving time and energy consumption during the task execution process and improve the efficiency of task execution.

[0021] Optionally, the path planning algorithm is an A-star algorithm or a Dijkstra algorithm.

[0022] In the second aspect, the present application provides a quadruped robot intelligent scheduling system, which adopts the following technical solutions: A quadruped robot intelligent scheduling system comprises: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the quadruped robot intelligent scheduling method described above is implemented.

[0023] The above-mentioned quadruped robot intelligent scheduling method is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a system based on the memory and the processor for easy use.

[0024] This application has the following technical effects: 1. Each robot bids for each task. The bid value reflects the robot's completion cost (power consumption to complete the task) and completion quality (time to complete the task). The lower the completion cost, the higher the completion quality, and the higher the bid value. At this time, the robot is more likely to match the task. At the same time, the value calculated by combining the constraint function can effectively avoid the task being assigned to a robot that cannot complete the task, and the robot with insufficient power has a bid value of 0. The bid value and task priority are used as the edge weights for matching tasks with robots. The power consumption and completion time are predicted by the graph neural network. The global matching of multiple tasks and multiple robots is performed according to the matching algorithm. The task allocation can be dynamically adjusted according to real-time data, and the task allocation is highly flexible.

[0025] 2. Through graph neural networks, robots and tasks can be modeled as nodes and edges in the graph. Task assignment can be performed using node features and edge features. This can effectively handle the complex relationships in multi-robot task assignment and improve the efficiency and accuracy of task assignment. Graph neural networks can dynamically adjust task assignment strategies. For example, when the task point location or robot status changes, tasks can be quickly reallocated. The global information processing capabilities of graph neural networks can achieve global optimization of multi-robot task assignment.

[0026] 3. Directly associate the nodes and robots that the task route passes through. The purpose of selecting the target robot is to screen the robots and quickly locate the robots directly related to the task. When there are a large number of robots, this method can significantly reduce the screening time and the amount of calculation when matching subsequent robots with tasks, reduce the waste of computing resources, and improve computing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a method flow chart of a method for intelligent scheduling of a quadruped robot according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] The present application embodiment discloses a method for intelligent scheduling of a quadruped robot, referring to Figure 1 , including steps S1 to S4, which are specifically as follows: S1: Construct and train a graph neural network. The nodes of the graph neural network represent the coordinates of the loading and unloading points, the weight of the cargo, and the initial power of the robot. The edges of the graph neural network represent the routes between the nodes. The input of the graph neural network is the adjacency matrix and the node feature matrix. The output is the predicted value of power consumption and the predicted value of task completion time.

[0029] In the graph neural network, each loading and unloading point is a node, and the edge represents the connection relationship between nodes. Specifically, it can represent the actual path of the robot from one loading and unloading point to another, and also represents the order of loading and unloading points that the robot needs to visit in sequence. The path planning algorithm is used to calculate the shortest path between nodes as the route between nodes. The path planning algorithm can be the A-star algorithm or the Dijkstra algorithm, and the prior art will not be repeated here.

[0030] The robots in this application are robots of the same model and parameters, the purpose is to make the robot parameters consistent in various aspects, such as load limit, total power, etc. The robot in this application is a quadruped robot that can move and transport goods.

[0031] In one embodiment, the input of the graph neural network is the adjacency matrix and the node feature matrix. The graph neural network adjacency matrix represents the connection relationship between nodes in the graph, which is used to describe which nodes have direct paths. The node feature matrix contains the characteristics of each node, such as the coordinates of the loading and unloading points, the weight of the cargo, the initial power of the robot, etc. The output is the power consumption prediction value and the task completion time prediction value.

[0032] The neural network includes a shared feature extraction layer, a first output layer and a second output layer which are independent of each other. The shared feature extraction layer is connected to the first output layer and the second output layer respectively. The first output layer is used to output a power consumption prediction value, and the second output layer is used to output a completion time prediction value.

[0033] The shared feature extraction layer includes: a graph convolution layer, which is used to extract local features of nodes; a graph attention layer, which assigns different weights to different neighboring nodes through an attention mechanism; and a gated graph neural network layer, which processes dynamic transmission between nodes through a gating mechanism; the output of the graph convolution layer serves as the input of the graph attention layer, and the output of the graph attention layer serves as the input of the gated graph neural network layer.

[0034] In the graph neural network training, the model training is stopped when the preset number of training times (such as 1000 times) is reached or the loss function is less than the preset loss function threshold (such as 0.01). The prior art will not be described here. The loss function of the graph neural network training is: + ; In the formula, represents the loss function, Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The actual value of the completion time; Indicates the total number of target robots; Indicates Target robots complete the task The predicted power consumption value is Indicates The actual power consumption of the target robot.

[0035] S2: Obtain the task route, and select the robot at the node directly associated with the node through which the task route passes and the robot at the node through which the task route passes as the target robot.

[0036] In graph neural networks, direct association refers to the existence of direct edge connections between nodes in the graph. In this case, information transmission is carried out through directly connected nodes. Indirect association means that the information transmission between nodes is not direct, but forwarded through intermediate nodes. The concepts of direct association and indirect association are existing technologies of graph neural networks and will not be repeated here.

[0037] The purpose of selecting the target robot is to screen the robots and quickly locate the robots directly related to the task. When there are a large number of robots, this method can significantly reduce the screening time and the amount of calculation when matching subsequent robots with tasks, reduce the waste of computing resources, and improve computing efficiency.

[0038] S3: Calculate the target robot’s bid value.

[0039] The calculation method of the offer value is as follows: according to the predicted power consumption value, predicted completion time value and preset constraint function of each target robot under the same task, the offer value of each target robot for the same task is calculated. The power consumption prediction value is inversely proportional to the offer value, and the completion time prediction value is inversely proportional to the offer value. The constraint function expresses that when the power consumption prediction value is greater than the initial power of the target robot, the offer value of the robot is set to 0.

[0040] In one embodiment, the expression of the bid value is: ;in, Indicates target robot for the task The output value, Indicates The exponential function with base , Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The predicted power consumption value is Indicates the preset time adjustment parameters. Indicates the preset power consumption adjustment parameters; represents the constraint function; Indicates The initial power of the target robot.

[0041] Each robot bids for each task, and the bid value reflects the robot's completion cost (power consumption to complete the task) and completion quality (time to complete the task). The lower the completion cost, the higher the completion quality, and the higher the bid value, the higher the possibility of the robot matching the task. Conversely, the lower the bid value, the lower the possibility of the robot matching the task.

[0042] The constraint function can effectively avoid assigning tasks to robots that cannot complete the task, and make the output value of the robot with insufficient power equal to 0. Target robots complete the task The power consumption forecast value is less than or equal to When the initial power of the target robot is less than 1, it means that the power of the robot is not enough or just enough to support the robot to complete the task, the constraint function takes 0 and the bid value takes 0. The lower the bid value, the lower the probability that the robot will match the task in the subsequent matching. Otherwise, the constraint function takes 1.

[0043] The time adjustment parameter is used to adjust the influence of task completion time on the offer value. When the time adjustment parameter increases, the attenuation effect of time on the offer value increases, that is, the longer the completion time, the faster the offer value decreases; conversely, the attenuation effect of time on the offer value decreases. The power consumption adjustment parameter is used to adjust the influence of power consumption on the offer value. When the power consumption adjustment parameter increases, the attenuation effect of power consumption on the offer value increases, that is, the greater the power consumption, the faster the offer value decreases; conversely, the attenuation effect of power consumption on the offer value decreases. For example, , ,The values ​​of the time adjustment parameter and the power consumption adjustment parameter can be adjusted according to the actual ,application scenario.

[0044] In other embodiments, the constraint function may also be: ; represents the constraint function; Indicates The initial power of the target robot; Indicates the preset safe power level; Indicates the preset power regulation parameters.

[0045] In actual working conditions, factory loading and unloading tasks usually require robots to complete tasks efficiently and return safely to recharge to avoid downtime due to power exhaustion. Safety power is added to the calculation of constraints to reserve enough power for the robot to automatically return to the charging position. Power adjustment parameters It is used to adjust the weight of the safe power, so as to control the power consumption range allowed when the robot performs tasks. For example, , The value can be adjusted according to the actual application scenario.

[0046] S4: Obtain the priority coefficient of each task, use the priority coefficient and the bid value as the edge weight for matching the task with the robot, use the task allocation scheme with the largest cumulative edge weight as the matching scheme, and perform global matching of multiple robots and multiple tasks according to the KM matching algorithm.

[0047] The KM matching algorithm is an algorithm for solving the maximum weight matching problem in a weighted bipartite graph. In the KM matching algorithm, edge weight refers to the weight of each edge in the bipartite graph, which reflects the "value" or "importance" of the edge. The KM matching algorithm and edge weight are both existing technologies and will not be described in detail.

[0048] The expression of the cumulative sum of edge weights is: ; represents the cumulative sum of edge weights, Indicates target robot for the task The output value; Indicates the task The priority coefficient of is the total number of target robots; Indicates the task The total number of .

[0049] The tasks and the urgency of the tasks input by the user are obtained. Different priority coefficients can be set for different urgency levels. The larger the priority coefficient, the more urgent the task. For example, tasks are divided into two categories, urgent tasks and ordinary tasks. The priority coefficient of urgent tasks is 1.2, and the priority coefficient of ordinary tasks is 1.

[0050] The priority coefficient of each task is multiplied by the corresponding robot's bid value, ensuring that high-priority tasks contribute more to the overall edge weight. If a task has a high priority, the robot's bid value assigned to the task will have a greater impact on the overall edge weight.

[0051] Comprehensively consider the importance of the task and the execution capability of the robot. Ensure that the task allocation plan not only considers the urgency of the task, but also the actual capability of the robot. The calculation method of the edge weighted cumulative sum enables the system to evaluate the quality of the task allocation plan from a global perspective. By maximizing the edge weighted cumulative sum, the optimal task allocation plan can be found to ensure the highest overall task execution efficiency.

[0052] An embodiment of the present application also discloses a quadruped robot intelligent scheduling system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the quadruped robot intelligent scheduling method according to the present application is implemented.

[0053] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0054] In the present application, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory, a dynamic random access memory, a static random access memory, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0055] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A quadruped robot intelligent scheduling method, characterized in that: Includes steps: Build and train a graph neural network; the nodes of the graph neural network represent the coordinates of the loading and unloading points, the weight of the cargo, and the initial power of the robot; the edges of the graph neural network represent the routes between the nodes, and the input of the graph neural network is the adjacency matrix and the node feature matrix; The output is the power consumption prediction value and the task completion time prediction value; the task route is obtained, and the robot at the node directly associated with the node passed by the task route and the robot at the node passed by the task route are selected as the target robot; Calculate the bid value of the target robot; obtain the priority coefficient of each task, use the priority coefficient and bid value as the edge weight for matching the task with the robot, use the task allocation scheme with the largest cumulative edge weight as the matching scheme, and perform global matching of multiple robots and multiple tasks according to the KM matching algorithm; The calculation method of the offer value is as follows: according to the predicted power consumption value, predicted completion time value and preset constraint function of each target robot under the same task, the offer value of each target robot for the same task is calculated. The power consumption prediction value is inversely proportional to the offer value, and the completion time prediction value is inversely proportional to the offer value. The constraint function expresses that when the power consumption prediction value is greater than the initial power of the target robot, the offer value of the robot is set to 0.

2. The intelligent scheduling method for a quadruped robot according to claim 1, characterized in that: The expression of the bid value is: ;in, Indicates target robot task The output value, Indicates The exponential function with base , Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The predicted power consumption value is Indicates the preset time adjustment parameters. Indicates the preset power consumption adjustment parameters; represents the constraint function; Indicates The initial power of the target robot.

3. The intelligent scheduling method for quadruped robots according to claim 1, characterized in that: The expression of the bid value is: ;in, Indicates target robot task The output value, Indicates The exponential function with base , Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The predicted power consumption value is Indicates the preset time adjustment parameters. Indicates the preset power consumption adjustment parameters; represents the constraint function; Indicates The initial power of the target robot; Indicates the preset safety power; Indicates the preset power regulation parameters.

4. The intelligent scheduling method for a quadruped robot according to claim 1, characterized in that: The expression of the cumulative sum of edge weights is: ; represents the cumulative sum of edge weights, Indicates target robot task The output value; Indicates the task The priority coefficient of is the total number of target robots; Indicates the task The total number of .

5. The intelligent scheduling method for a quadruped robot according to claim 1, characterized in that: The graph neural network includes a shared feature extraction layer, a first output layer and a second output layer that are independent of each other, the shared feature extraction layer is connected to the first output layer and the second output layer respectively, the first output layer is used to output a power consumption prediction value, and the second output layer is used to output a completion time prediction value; The shared feature extraction layer includes: a graph convolution layer, which is used to extract local features of nodes; a graph attention layer, which assigns different weights to different neighboring nodes through an attention mechanism; and a gated graph neural network layer, which processes dynamic transmission between nodes through a gating mechanism; the output of the graph convolution layer serves as the input of the graph attention layer, and the output of the graph attention layer serves as the input of the gated graph neural network layer.

6. The intelligent scheduling method for a quadruped robot according to claim 1, characterized in that: The loss function for graph neural network training is: + ; In the formula, represents the loss function, Indicates Target robots complete the task The predicted completion time of Indicates Target robots complete the task The actual value of the completion time; Indicates the total number of target robots; Indicates Target robots complete the task The predicted power consumption value is Indicates The actual power consumption of the target robot.

7. The intelligent scheduling method for a quadruped robot according to claim 1, characterized in that: The task route is a sequence of nodes, which represent the loading and unloading points that the robot needs to visit in sequence.

8. The intelligent scheduling method for quadruped robots according to claim 1, characterized in that: The shortest path between nodes is calculated using a path planning algorithm as the route between nodes.

9. The intelligent scheduling method for a quadruped robot according to claim 8, characterized in that: The path planning algorithm is A-star algorithm or Dijkstra algorithm.

10. A quadruped robot intelligent scheduling system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the intelligent scheduling method for a quadruped robot according to any one of claims 1 to 9 is implemented.

Images