Ground robot path and detection method based on information detection function

Abstract

The present invention provides a ground robot path and detection method based on information detection function, which belongs to the field of robot technology. The method includes: collecting task scene characterization information; constructing the exploration function of node information based on the task scene characterization information; solving the exploration function of the node information to obtain the total detection degree of all nodes; taking the node detection order and the detection point position as decision variables, and maximizing the weighted value of the total detection degree and the task completion time as the goal, constructing a ground robot path and detection optimization model; using a combination of a large-scale neighborhood search algorithm and a gradient descent algorithm to solve the ground robot's optimal path and detection optimization model, and obtain the ground robot's optimal path and detection plan. By taking the robot's detection capability and detection function into consideration, the detection time and information detection benefits are comprehensively optimized, providing further guidance for the application of ground robots in the field of perception coverage.

Description

Technical Field

[0001] The present invention belongs to the field of robot technology, and more specifically, relates to a ground robot path and detection method based on an information detection function. Background Art

[0002] In the robotic perception coverage problem, a mobile robot must choose its motion mode so that its sensors can detect multiple targets in a bounded environment. Robotic perception coverage applications include minesweeping, crop sampling, intruder detection, and sewage system inspection. They also include aerial inspection of industrial facilities, aerial search, seabed mapping, and hull inspection. When the detection capability of ground robots is limited, the information detection degree of nodes is usually related to the distance. At this time, in the scenario of using ground robots to detect information of multiple nodes in a large area, how to decide the detection order and the location of the detection points is an unresolved problem. Summary of the invention

[0003] The main purpose of this application is to provide a ground robot path and detection method based on information detection function, which can solve the optimal path and detection plan of the ground robot.

[0004] In order to achieve the above objectives, this application proposes a ground robot path and detection method based on an information detection function, including:

[0005] Collect mission scenario representation information;

[0006] Based on the task scenario representation information, construct an exploration function of the node information;

[0007] Solve the exploration function of the node information to obtain the total detection degree of all nodes;

[0008] Based on the total detection degree, a ground robot path and detection optimization model is constructed with node detection sequence and detection point positions as decision variables;

[0009] The optimal path and detection optimization model of the ground robot is solved by combining the large-scale neighborhood search algorithm and the gradient descent algorithm, and the optimal path and detection plan of the ground robot is obtained.

[0010] Furthermore, constructing an exploration function of node information based on the task scenario representation information includes:

[0011] Based on the detection capability of the ground robot, an exploration function related to the node and distance is constructed:

[0012]

[0013] Where d is the distance between the ground robot and the node during detection, i is the node number, and m is the maximum node number.

[0014] Furthermore, solving the exploration function of the node information to obtain the total detection degree of all nodes includes the following process:

[0015] The number of the node detected by the ground robot for the i-th time is: ;

[0016] The position of the detection point of the ground robot's i-th detection is: ;

[0017] When the ground robot detects for the i-th time, the distance between it and the detection node is:

[0018] ;

[0019] The information detection degree of the ground robot for the i-th detection is:

[0020] ;

[0021] The total detection degree of all nodes is:

[0022] ;

[0023] In the above formula, represents the total detection degree of all nodes, represents the information detection degree of the ground robot's i-th detection, It represents the distance between the ground robot and the detection node during the i-th detection, i is the node number, and m is the maximum node number.

[0024] Furthermore, based on the total detection degree, a ground robot path and detection optimization model is constructed with the node detection sequence and the detection point position as decision variables, including:

[0025] Taking the node detection order and detection point position as constraints, combined with the driving information of the ground robot, and taking the weighted value maximization of the total detection degree and the task completion time as the goal, a ground robot path and detection optimization model is constructed:

[0026]

[0027] In the formula, is the weighted value of the total detection degree and the time taken to complete the task, and are the weights of total detection degree and task completion time, respectively. Time consuming for task completion.

[0028] Furthermore, the method of solving the optimal path and detection optimization model of the ground robot by combining the large-scale neighborhood search algorithm and the gradient descent algorithm to obtain the optimal path and detection solution of the ground robot includes:

[0029] The large-scale neighborhood search algorithm is aimed at optimizing the decision variables of the detection node numbers, and the gradient descent algorithm is aimed at optimizing the decision variables of the detection point positions.

[0030] Furthermore, the method further comprises the following steps:

[0031] S501, randomly generating an initial solution of node detection sequence as the current solution of node detection sequence;

[0032] S502, for the current solution of the node detection sequence, taking the location of the node as the initial solution of the detection point location, and taking the initial solution of the detection point location as the current solution of the detection point location;

[0033] S503, for the current solution of the node detection sequence and the current solution of the detection point position, calculate the gradient of the objective function with respect to the detection point position;

[0034] S504, using the learning rate to update the detection point position;

[0035] S505: for the current solution of the node detection sequence, when the relative difference between the objective function values ​​of two consecutive updates of the detection point position is less than a first preset value, the objective function value at this time is used as the objective function value of the current solution of the node detection sequence;

[0036] S506, use the number of removals The random removal operator is used as the destruction operator to detect the current solution of the node detection order. nodes to be removed; among them, is an integer;

[0037] S507, the number of repairs used is The best repair operator is used to repair the node, and a new solution of the node detection order is generated; wherein, when repairing, the calculation steps of the objective function value under the given node detection order are S502 to S505;

[0038] S508, calculating the difference between the objective function value under the current solution of the node detection sequence and the new solution of the node detection sequence;

[0039] S509. If the difference in the objective function values ​​is greater than or equal to the second preset value, the new solution to the node detection sequence is accepted as the current solution to the node detection sequence, and the process returns to step S506. Otherwise, the node detection sequence and detection point positions under the new solution to the node detection sequence are output as the optimal path and detection plan for the ground robot.

[0040] Furthermore, the acquisition of task scenario representation information includes:

[0041] The scenario of using ground robots to detect multi-node information in a large area is characterized.

[0042] Furthermore, the task scenario characterization information includes: the number of the node, the position of the node, the number of the starting point of the ground robot, the position of the starting point of the ground robot, the number of the return point of the ground robot, the position of the return point of the ground robot, the driving speed of the ground robot, the detection range of the ground robot, and the time required for a single detection by the ground robot.

[0043] Compared with the prior art, the present invention has the following beneficial effects: the present invention aims to maximize the weighted value of the total detection degree and the task completion time, and adopts the node detection order and the detection point position as the decision variables to optimize the path and detection of the ground robot. The present invention utilizes the combination of the large-scale neighborhood search algorithm and the gradient descent algorithm to solve the optimal path and detection scheme of the ground robot. By taking the robot's detection capability and the detection degree function into consideration, the detection time and information detection benefits are comprehensively optimized, which provides further guidance for the application of ground robots in the field of perception coverage. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of the path and detection method of the ground robot based on the information detection function of the present invention;

[0045] Figure 2 Schematic diagram of the implementation scenario of the ground robot path and detection method based on the information detection function of the present invention;

[0046] Figure 3 A schematic diagram of a task scenario of a ground robot path and detection method based on an information detection function of the present invention;

[0047] Figure 4 Schematic diagram of the optimal path and detection solution for a ground robot provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0049] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0050] In the robot perception coverage problem, a mobile robot must choose its motion mode so that its sensors can detect multiple targets in a bounded environment. Robot perception coverage applications include minesweeping, crop sampling, intruder detection, and sewage system inspection. Other applications include aerial inspection of industrial facilities, aerial search, seabed mapping, and hull inspection. The purpose of this invention is to provide a ground robot path and detection method based on information detection function to solve the following technical problems:

[0051] When the detection capability of a ground robot is limited, the information detection degree of a node is usually related to the distance; at this time, how to decide the robot's detection order and detection point location.

[0052] Figure 1 The following is a flow chart of a ground robot path and detection method based on an information detection function provided in an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0053] S1. Collect mission scenario representation information.

[0054] Among them, the mission scenario characterization information can be used to characterize the scenario of using a ground robot to detect multi-node information in a large area, including the following processes:

[0055] After numbering the nodes, they form a node number set:

[0056] ,

[0057] In the formula, represents a set of m node numbers. In this embodiment, , node use Figure 3 The square in the figure indicates that the node number is marked inside the square.

[0058] The position of the node numbered i is: In this embodiment, Figure 3 The position of the square in the middle represents the two-dimensional projection position of the corresponding node, and the height coordinates of all nodes are 0.

[0059] The starting point of the ground robot is numbered 0. In this embodiment, the starting point is Figure 3 The triangle in the figure indicates the starting point number.

[0060] The starting point of the ground robot is In this embodiment, Figure 3 The position of the middle triangle represents the two-dimensional projection position of the corresponding starting point, and the height of the starting point is 0.

[0061] The return point of the ground robot is numbered In this embodiment, the return point is Figure 3 The triangle in the figure indicates the number of the return point.

[0062] The return point of the ground robot is In this embodiment, Figure 3 The position of the middle triangle represents the two-dimensional projection position of the corresponding return point, and the height of the return point is 0.

[0063] The travel speed of the ground robot is v. In this embodiment, .

[0064] The detection range of the ground robot is the inside of a circle with the current position as the center and r as the radius. ,exist Figure 3 In the figure, the dotted circle with the center of the square as the center represents the range of the corresponding node that can be detected.

[0065] The time taken for a single detection by a ground robot is In this embodiment, .

[0066] S2. Constructing an exploration function of node information based on the task scenario representation information.

[0067] Summarizing the above mission scenario representation information, based on the limited detection capabilities of ground robots, we construct an exploration function related to nodes and distances:

[0068] . Where d is the distance between the ground robot and the node during detection. In this embodiment, all The expressions are .

[0069] S3, solving the exploration function of the node information to obtain the total exploration degree of all nodes, including the following process:

[0070] The number of the node detected by the ground robot for the i-th time is: ;

[0071] The position of the detection point of the ground robot's i-th detection is: ;

[0072] When the ground robot detects for the i-th time, the distance between it and the detection node is:

[0073] ,

[0074] In the formula, represents the distance between the ground robot and the detection node during the i-th detection;

[0075] The information detection degree of the ground robot for the i-th detection is:

[0076] ,

[0077] In the formula, represents the information detection degree of the ground robot's i-th detection;

[0078] The total detection degree of all nodes is:

[0079] ,

[0080] In the formula, Represents the total detection degree of all nodes

[0081] S4. Based on the total detection degree, a ground robot path and detection optimization model is constructed with node detection sequence and detection point positions as decision variables.

[0082] Specifically, taking the node detection order and the detection point position as constraints, combined with the driving information of the ground robot, and taking the weighted value maximization of the total detection degree and the task completion time as the goal, a ground robot path and detection optimization model is constructed, including the following processes:

[0083] Ground Robot i The number of the node to be detected is determined by the range of decision variables:

[0084] ;

[0085] For each node, the ground robot only detects the constraint once:

[0086] ;

[0087] Ground Robot i The range of value constraints of the detection point location decision variables for the second detection:

[0088] ;

[0089] Ground Robot i The distance traveled is:

[0090] ,

[0091] In the formula, Represents the ground robot i Distance travelled;

[0092] Ground Robot i The time of the trip is:

[0093] ,

[0094] In the formula, represents the time of the ground robot's ith travel;

[0095] The time taken to complete the task is:

[0096] ,

[0097] In the formula, Indicates the time taken to complete the task;

[0098] The objective function is set to maximize the weighted value of the total detection degree and the time taken to complete the task:

[0099] ,

[0100] In the formula, Represents the weighted value of the total detection degree and the time taken to complete the task. and Respectively represent the weights of the total detection degree and the time taken to complete the task. In this embodiment, , ;

[0101] In summary, the following specific ground robot path and detection model is obtained:

[0102] .

[0103] S5. Use a combination of a large-scale neighborhood search algorithm and a gradient descent algorithm to solve the optimal path and detection optimization model of the ground robot to obtain the optimal path and detection plan of the ground robot.

[0104] Among them, the large-scale neighborhood search algorithm is used to optimize the decision variables of the detection node number, and the gradient descent algorithm is used to optimize the decision variables of the detection point location. Specifically, it includes the following processes:

[0105] Step S501, randomly generate an initial solution of node detection sequence as the current solution of node detection sequence;

[0106] Step S502: for the current solution of the node detection sequence, the node location is used as the initial solution of the detection point location, and the initial solution of the detection point location is used as the current solution of the detection point location;

[0107] Step S503: For the current solution of the node detection sequence and the current solution of the detection point position, calculate the gradient of the objective function with respect to the detection point position. , , ;

[0108] Step S504: According to the learning rate Update the position of the detection point, that is , , In this embodiment, ;

[0109] Step S505: For the current solution of the node detection sequence, when the relative difference between the objective function values ​​of two consecutive updates of the detection point position is less than When , the objective function value at this time is used as the objective function value of the current solution of the node detection sequence. In this embodiment, ;

[0110] Step S506: Use the number of removals to The random removal operator is used as the destruction operator to detect the current solution of the node detection order. Nodes are removed. In this embodiment, ;

[0111] Step S507: Use the number of repairs to The best repair operator is used to repair the node, and a new solution of the node detection order is generated. When repairing, the calculation steps of the objective function value under the given node detection order are steps S502 to S505;

[0112] Step S508: Calculate the difference between the objective function value under the current solution of the node detection order and the new solution of the node detection order ;

[0113] Step S509: If , then accept the new solution of the node detection sequence as the current solution of the node detection sequence, and return to step 506, otherwise, output the node detection sequence and detection point positions under the new solution of the node detection sequence as the optimal path and detection plan of the ground robot, and end the entire algorithm. In this embodiment, , the maximum value of the objective function can be obtained as 1.85, and the following can be obtained: Figure 4 The optimal path and detection scheme of the ground robot shown in FIG. Among them, the solid arrow indicates the driving direction of the ground robot, the dotted arrow indicates the detection point of the ground robot, and the solid circle indicates the location of the detection point.

[0114] In the description of the present invention, it is necessary to understand that the terms "middle", "length", "up", "down", "front", "back", "vertical", "horizontal", "inside", "outside", "radial", "circumferential" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0115] In the present invention, unless otherwise clearly specified and limited, the first feature "on" the second feature may be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. "Multiple" means at least two, such as two, three, etc., unless otherwise clearly and specifically limited.

[0116] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral one; it can be a mechanical connection, an electrical connection, or communication with each other; it can be a direct connection, or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0117] The above is only for explaining the implementation mode of the present invention and is not intended to limit the present invention. For those skilled in the art, any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention without creative work should be included in the protection scope of the present invention.

Claims

1. A ground robot path and detection method based on information detection function, characterized in that: include: Collecting mission scenario characterization information; wherein, characterizing scenarios in which a ground robot is used to detect information on multiple nodes in a large area, including the node number, the node location, the number of the ground robot's starting point, the location of the ground robot's starting point, the number of the ground robot's return point, the location of the ground robot's return point, the ground robot's travel speed, the ground robot's detection range, and the time taken by the ground robot for a single detection; Based on the task scenario representation information, an exploration function of the node information is constructed; wherein, based on the detection capability of the ground robot, an exploration function related to the node and the distance is constructed: In the formula, d is the distance between the ground robot and the node during detection, i is the node number, m is the maximum number of nodes; Solve the exploration function of the node information to obtain the total detection degree of all nodes; Based on the total detection degree, a ground robot path and detection optimization model is constructed with node detection sequence and detection point positions as decision variables; The optimal path and detection optimization model of the ground robot is solved by combining a large-scale neighborhood search algorithm and a gradient descent algorithm, and the optimal path and detection plan of the ground robot is obtained; wherein, the large-scale neighborhood search algorithm is used to optimize the decision variables of the detection node number, and the gradient descent algorithm is used to optimize the decision variables of the detection point position.

2. The ground robot path and detection method based on information detection function according to claim 1 is characterized in that: The step of solving the exploration function of the node information to obtain the total exploration degree of all nodes includes the following steps: Ground Robot i The number of the node detected this time is: ; Ground Robot i The detection point position of this detection is: ; In ground robot i During the first detection, the distance between it and the detection node is: ； The information detection degree of the ground robot for the i-th detection is: ； The total detection degree of all nodes is: ； In the above formula, represents the total detection degree of all nodes, Represents the ground robot i The information detection degree of the secondary detection, represents the distance between the ground robot and the detection node during the i-th detection. i is the node number, m The maximum number of nodes.

3. The ground robot path and detection method based on information detection function according to claim 2 is characterized in that: Based on the total detection degree, the node detection sequence and the detection point position are used as decision variables to construct a ground robot path and detection optimization model, including: Taking the node detection order and detection point position as constraints, combined with the driving information of the ground robot, and taking the weighted value maximization of the total detection degree and the task completion time as the goal, a ground robot path and detection optimization model is constructed: In the formula, is the weighted value of the total detection degree and the time taken to complete the task, and are the weights of total detection degree and task completion time, respectively. Time consuming for task completion.

4. The ground robot path and detection method based on information detection function according to claim 1 is characterized in that: The following steps are also included: S501, randomly generating an initial solution of node detection sequence as the current solution of node detection sequence; S502, for the current solution of the node detection sequence, taking the location of the node as the initial solution of the detection point location, and taking the initial solution of the detection point location as the current solution of the detection point location; S503, for the current solution of the node detection sequence and the current solution of the detection point position, calculate the gradient of the objective function with respect to the detection point position; S504, using the learning rate to update the detection point position; S505: for the current solution of the node detection sequence, when the relative difference between the objective function values ​​of two consecutive updates of the detection point position is less than a first preset value, the objective function value at this time is used as the objective function value of the current solution of the node detection sequence; S506, use the number of removals The random removal operator is used as the destruction operator to detect the current solution of the node detection order. nodes to be removed; among them, is an integer; S507, the number of repairs used is The best repair operator is used to repair the node, and a new solution of the node detection order is generated; wherein, when repairing, the calculation steps of the objective function value under the given node detection order are steps S502 to S505; S508, calculating the difference between the objective function value under the current solution of the node detection sequence and the new solution of the node detection sequence; S509. If the difference in the objective function values ​​is greater than or equal to the second preset value, the new solution to the node detection sequence is accepted as the current solution to the node detection sequence, and the process returns to step S506. Otherwise, the node detection sequence and detection point positions under the new solution to the node detection sequence are output as the optimal path and detection plan for the ground robot.

Images