An active mapping method for harvesting robots based on greedy strategy

Abstract

The present invention discloses an active mapping method for a harvesting robot based on a greedy strategy, which includes a SLAM mapping method, the selection of active exploration points based on the greedy strategy, and the design of a local planner. The core point is the selection of active exploration points. The present invention comprehensively considers the distance between the frontier and the robot, and the information entropy of the frontier set is used to determine the exploration point, and the simulated annealing algorithm is used to optimize the exploration path according to the exploration point, thereby reducing the exploration path length of the robot and shortening the mapping time.

Description

Technical Field

[0001] The present invention belongs to the technical field of agricultural production, and in particular relates to a method and device for active mapping of a harvesting robot based on a greedy strategy. Background Art

[0002] With the continuous deepening of research on computer vision technology, navigation technology, and multi-sensor fusion technology, robotics technology has begun to be widely used in unmanned driving, unmanned agricultural machinery, fruit picking robots, drone rescue and other fields. Among them, as the core of navigation, simultaneous localization and mapping technology (SLAM) has attracted widespread attention from researchers. In the mainstream SLAM technology process, robots are usually guided by personnel with expert knowledge to ensure full coverage of the environment and complete reconstruction of the environment. Therefore, SLAM technology is a passive mapping technology, and the quality of the established map depends on the operator's experience.

[0003] The operator controls the robot's trajectory and termination criteria, and the precision and completeness of the map are directly determined by the operator. If the operator is inexperienced, it will greatly affect the quality of the generated map, resulting in the failure of subsequent navigation and path planning tasks.

[0004] Therefore, it is necessary to establish a mapping method that does not rely on the operator's subjective experience to ensure the integrity of the map and reduce the exploration time. Summary of the invention

[0005] In view of the above defects and improvement needs of the prior art, the purpose of the present invention is to propose an active mapping method for a picking robot based on a greedy strategy applied to an orchard picking robot.

[0006] The objective of the present invention is achieved through the following technical solutions:

[0007] A method for actively building a map for a harvesting robot based on a greedy strategy is provided, comprising the following steps:

[0008] S1, initialize robot parameters;

[0009] S2, obtains the measurement values ​​of the robot's surrounding environment and its own position in the world coordinate system;

[0010] S3, based on the measurement value of the robot's surrounding environment in the world coordinate system in step S2, using the SLAM algorithm to estimate the robot's own position in the local map coordinate system, and construct a local map;

[0011] S4, based on the measured value of the robot's own position in the world coordinate system in step S2, construct a conversion relationship between the world coordinate system and the local map coordinate system, read the local map, use the robot position as the origin, use the Dijkstra algorithm to perform frontier detection, and use the DBscan algorithm to construct a candidate friontiers set, select the optimal exploration point in the candidate frontiers set according to the greedy strategy, and use the simulated annealing algorithm to minimize the exploration distance;

[0012] S5, based on the local map and the minimum exploration distance, plans a safe trajectory, calculates control commands through the robot kinematic model, and constructs a local augmented map;

[0013] S6, compare the local augmented map with the actual map area, calculate the percentage of the mapped area, and if the percentage is greater than the termination threshold for mapping, the mapping is completed, otherwise return to step S2.

[0014] Furthermore, the candidate frontiers set is constructed in step S4, which specifically includes the following operations:

[0015] Use Dijkstra algorithm for frontier detection, use DBscan algorithm for clustering, calculate the number of frontiers in the frontiers set, and if the minimum number of boundary points n is met k , it is determined to be a candidate frontiers set, otherwise the frontiers set is discarded.

[0016] Furthermore, the construction of the conversion relationship between the world coordinate system and the local map coordinate system in step S4 specifically includes the following operations:

[0017] S41, integrating the robot wheel encoder, inertial measurement unit and global positioning system information, calculating the robot odometer and the position in the world coordinate system, wherein the coordinate P of the robot in the world coordinate system is calculated. robot as follows:

[0018] P robot =[x w y w 1] T

[0019] Among them, x w is the x-coordinate of the robot in the world coordinate system, y w is the y-coordinate of the robot in the world coordinate system;

[0020] S42, obtain the robot's own position p in the local map coordinate system robot as follows:

[0021] p robot =[x m y m 1] T

[0022] Among them, x m is the x-coordinate of the robot in the local map coordinate system, y m is the y-coordinate of the robot in the local map coordinate system, p robot is the coordinate of the robot in the local map coordinate system

[0023] S43, based on the robot's own position p in the local map coordinate system in step S3 robot The measured value P of the robot's own position in the world coordinate system in step S2 robot The mapping relationship between the local map and the world coordinate system is calculated. mw as follows:

[0024] P robot =T mw ×P robot

[0025]

[0026] Among them, r is the map resolution, b x is the offset between the local map and the world coordinate system in the x direction, b y is the offset between the local map and the world coordinate system in the y direction, T mw It is the mapping relationship between the local map and the world coordinate system.

[0027] Furthermore, the greedy strategy in step S4 comprehensively considers three metrics: the distance between the robot and the frontier, the information entropy brought by the frontier, and the number of points in a circle with a radius R and the frontier as the center.

[0028] Furthermore, in step S3, the best exploration point in the candidate frontiers set is selected according to the greedy strategy, which specifically includes the following operations:

[0029] S44, calculate the distance between the robot's current position and the i-th point in the frontiers set The calculation formula is as follows:

[0030]

[0031] Among them, x r is the x-coordinate of the robot in the local map coordinate system, y r is the y-coordinate of the robot in the local map coordinate system, xf is the x-coordinate of the i-th point in the frontiers set in the local map coordinate system, and y f is the y-coordinate of the i-th point in the frontiers set in the local map coordinate system;

[0032] S45, obtain the number of points contained in the neighborhood radius R of the i-th point in the current candidate frontiers set

[0033] S46, obtaining the information entropy of the current candidate frontiers set. The information entropy calculation formula is as follows:

[0034]

[0035] in, represents the information entropy of the eight neighborhoods of the i-th point in the current frontiers set, p(x i ,y i ) represents the i-th point (x i ,y i ) is the probability of a point in the unexplored region, where x i is the x-coordinate of the i-th point in the frontiers set in the local map coordinate system, and y i is the y-coordinate of the i-th point in the frontiers set in the local map coordinate system;

[0036] S47, construct a cost utility function based on the greedy strategy, where the greedy strategy comprehensively considers the distance between the robot and the frontier, the information entropy brought by the frontier, and the number of points in the circle with the frontier as the center and the field radius R. The cost utility function is as follows:

[0037]

[0038] The optimal exploration point in the current frontiers set is:

[0039] i * = arg max Cost i , i∈{1...n}

[0040] Among them, i * Indicates that the i-th point in the current frontiers set is the optimal exploration point.

[0041] Furthermore, in step S4, the simulated annealing algorithm is used to minimize the exploration distance, which specifically includes the following operations:

[0042] S48, initialize temperature parameter T = T0, n = 0, termination condition ΔE ≤ ∈ | n ≤ N max ∣T≤η, the optimal solution is X * Indicates; where N max is the maximum number of iterations, represents the minimum energy difference, and η represents the minimum temperature parameter;

[0043] S49, with the robot's current position p robot As the extreme point, the exploration point set consisting of the best exploration point in the candidate frontiers set is sorted by extreme angle as the initial solution X0 = [f1, f2, ..., f n ], X * =X0, and calculate the energy function Among them, f i represents the i-th exploration point, (x i ,y i ) is f i The coordinates in the local map coordinate system, (x i+1 ,y i+1 ) is f i+1 Coordinates in the local map coordinate system, x i The x-coordinate and y-coordinate in the local map coordinate system i The coordinate in the y direction in the local map coordinate system;

[0044] S410, let T = kT, k∈[0,1], k is the temperature drop rate;

[0045] S411, randomly exchange the order of two exploration points as the new solution X1=[f2,f1,…,f n ], and calculate the energy function E(X1), ΔE=E(X1)-E(X0);

[0046] S412, based on the Metropolis criterion, determine whether to accept the new solution. If accepted, X * =X1, otherwise proceed to the next step;

[0047]

[0048] Where P represents the probability of accepting a new solution. If the energy value of the current new solution is less than the optimal solution, the new solution is accepted and the optimal solution is updated. The probability of accepting the new solution and updating the optimal solution is The probability of rejecting the new solution is , if the new solution is rejected, jump to S413;

[0049] S413, iteration number n=n+1, repeat the cycle [3-5] until the termination condition is met.

[0050] The beneficial effects of the present invention are: an active mapping algorithm based on a greedy strategy applied to an orchard picking robot improves the robot's autonomy and reduces labor costs in large-scale scenarios such as orchards and greenhouses, and has significant economic advantages. It integrates multiple sensor data, including but not limited to wheel encoders (Encoders), inertial measurement units (IMUs), and global positioning systems (GPS) to calculate the position of the robot in the world coordinate system, reducing the uncertainty of single sensor positioning. By using the Dijkstra algorithm, the exploration boundary is accurately located to ensure the completeness of the map, and it can be flexibly expanded to a 3D map to achieve a "plug and play" function. Taking into account the size, distance, and information entropy of the boundary, a cost-utility function is designed. Based on a greedy strategy, the selection of exploration points is completed according to the principle of the highest utility. Finally, the robot's exploration path is optimized based on a simulated annealing algorithm, the exploration path is shortened, and the energy consumption of the system is reduced. By comparing the established map with the actual map area to determine whether the termination condition is met, the limited number of algorithm iterations is guaranteed, the influence of human factors is eliminated, the accuracy and completeness of the established map is guaranteed, and the intelligence of the robot is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0052] Figure 1 It is a flow chart of a method for actively building a map for a harvesting robot based on a greedy strategy provided in one embodiment of the present invention;

[0053] Figure 2 A flowchart of a greedy strategy algorithm in one embodiment of the present invention;

[0054] Figure 3 This is a test image of an active mapping method for a harvesting robot based on a greedy strategy in one embodiment of the present invention;

[0055] Figure 4 This is the expected effect of the active mapping method for a harvesting robot based on a greedy strategy in one embodiment of the present invention; DETAILED DESCRIPTION

[0056] In order to better understand the technical solution of the present application, the embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0057] It should be clear that the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0058] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "an", "the" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0059] The present invention provides a method for actively building a map for a harvesting robot based on a greedy strategy. Figure 1 As shown in the figure, the active mapping method of the picking robot based on the greedy strategy provides an active exploration framework for the passive SLAM mapping method, reduces human interference factors, and improves the initiative of mapping. The sensor model observes the real environment, obtains odometer information and local environment observations, including but not limited to scan lidar data, RGB image data, etc., uses the SLAM mapping algorithm to build a local map, passes the constructed part of the map into the active exploration framework, obtains the optimized exploration path, and converts it into control instructions through the robot model. The robot continuously interacts with the real environment until the map is built.

[0060] like Figure 2 As shown, the active mapping method of the picking robot based on the greedy strategy includes the following steps:

[0061] S1, initialize robot parameters; specifically, during the active mapping process of the picking robot, initialize the map area S, map resolution r, termination threshold ∈, neighborhood radius R, robot speed limit Acceleration Limit Minimum number of boundary points n k , cost utility function parameters α, β, γ, where are the minimum and maximum speeds of the robot in the x direction, are the minimum and maximum speed of the robot in the y direction, a x 、a y are the maximum linear acceleration of the robot in the x and y directions, is the maximum angular acceleration of the robot;

[0062] In one embodiment, during the active mapping process of the picking robot, the area of ​​the initial map is S = 48m 2 ,Map resolution r = 0.2m, termination threshold ∈ = 0.96, neighborhood radius R = 5, robot speed limit Acceleration Limit Minimum number of boundary points n k =5, cost-utility function parameters α=3.00, β=1.25, γ=1.00.

[0063] S2, obtain the measurement values ​​of the robot's surrounding environment and its own position in the world coordinate system; specifically, start the robot's sensors, including but not limited to Lidar, Inertial Measurement Unit (IMU) and Positioning System (GPS), obtain the measurement values ​​of the robot's surrounding environment and its own position, start the motion controller, and ensure that the control command [v x ,v y ,ω] can effectively act on the picking robot;

[0064] S3, based on the measurement value of the robot's surrounding environment in the world coordinate system in step S2, using the SLAM algorithm to estimate the robot's own position in the local map coordinate system, and construct a local map;

[0065] S4, based on the measured value of the robot's own position in the world coordinate system in step S2, construct a conversion relationship between the world coordinate system and the local map coordinate system, read the local map, use the robot position as the origin, use the Dijkstra algorithm to perform frontier detection, and use the DBscan algorithm to construct a candidate friontiers set, select the optimal exploration point in the candidate frontiers set according to the greedy strategy, and use the simulated annealing algorithm to minimize the exploration distance;

[0066] S5, based on the local map and the minimum exploration distance, plans a safe trajectory, calculates control commands through the robot kinematic model, and constructs a local augmented map;

[0067] S6, compare the local augmented map with the actual map area, calculate the percentage of the mapped area, and if the percentage is greater than the termination threshold for mapping, the mapping is completed, otherwise return to step S2.

[0068] In one embodiment, constructing a candidate frontiers set in step S4 specifically includes the following operations:

[0069] Use Dijkstra algorithm for frontier detection, use DBscan algorithm for clustering, calculate the number of frontiers in the frontiers set, and if the minimum number of boundary points n is met k , it is determined to be a candidate frontiers set, otherwise the frontiers set is discarded.

[0070] In one embodiment, the conversion relationship between the world coordinate system and the local map coordinate system is constructed in step S4, which specifically includes the following operations:

[0071] S41, integrating the robot wheel encoder, inertial measurement unit and global positioning system information, calculating the robot odometer and the position in the world coordinate system, wherein the coordinate P of the robot in the world coordinate system is calculated. robot as follows:

[0072] P robot =[x w y w 1] T

[0073] Among them, x w is the x-coordinate of the robot in the world coordinate system, y w is the y-coordinate of the robot in the world coordinate system;

[0074] S42, obtain the robot's own position p in the local map coordinate system robot as follows:

[0075] p robot =[x m y m 1] T

[0076] Among them, x m is the x-coordinate of the robot in the local map coordinate system, y m is the y-coordinate of the robot in the local map coordinate system, p robot is the coordinate of the robot in the local map coordinate system;

[0077] S43, based on the robot's own position p in the local map coordinate system in step S3 robot The measured value P of the robot's own position in the world coordinate system in step S2 robot The mapping relationship between the local map and the world coordinate system is calculated. mw as follows:

[0078] p robot =T mw ×P robot

[0079]

[0080] Among them, r is the map resolution, b x is the offset between the local map and the world coordinate system in the x direction, b y is the offset between the local map and the world coordinate system in the y direction, T mw It is the mapping relationship between the local map and the world coordinate system.

[0081] In one embodiment, the greedy strategy in step S4 comprehensively considers three metrics: the distance between the robot and the frontier, the information entropy brought by the frontier, and the number of points in a circle with a radius R and the frontier as the center.

[0082] In one embodiment, in step S3, the optimal exploration point in the candidate frontiers set is selected according to the greedy strategy, which specifically includes the following operations:

[0083] S44, calculate the distance between the robot's current position and the i-th point in the frontiers set The calculation formula is as follows:

[0084]

[0085] Among them, x r is the x-coordinate of the robot in the local map coordinate system, y r is the y-coordinate of the robot in the local map coordinate system, x f is the x-coordinate of the i-th point in the frontiers set in the local map coordinate system, and y f is the y-coordinate of the i-th point in the frontiers set in the local map coordinate system;

[0086] S45, obtain the number of points contained in the neighborhood radius R of the i-th point in the current candidate frontiers set

[0087] S46, obtaining the information entropy of the current candidate frontiers set. The information entropy calculation formula is as follows:

[0088]

[0089] in, represents the information entropy of the eight neighborhoods of the i-th point in the current frontiers set, p(x i ,y i ) represents the i-th point (x i ,y i ) is the probability of a point in the unexplored region, where x i is the x-coordinate of the i-th point in the frontiers set in the local map coordinate system, and y i is the y-coordinate of the i-th point in the frontiers set in the local map coordinate system;

[0090] S47, construct a cost utility function based on the greedy strategy, where the greedy strategy comprehensively considers the distance between the robot and the frontier, the information entropy brought by the frontier, and the number of points in the circle with the frontier as the center and the field radius R. The cost utility function is as follows:

[0091]

[0092] The optimal exploration point in the current frontiers set is:

[0093] i * = arg max Cost i , i∈{1...n}

[0094] Among them, i * Indicates that the i-th point in the current frontiers set is the optimal exploration point.

[0095] In one embodiment, step S4 uses a simulated annealing algorithm to minimize the exploration distance, which specifically includes the following operations:

[0096] S48, initialize temperature parameter T = T0, n = 0, termination condition ΔE ≤ ∈ | n ≤ N max |T≤η,

[0097] The optimal solution is to use X * Indicates; where N max is the maximum number of iterations, represents the minimum energy difference, and η represents the minimum temperature parameter;

[0098] S49, with the robot's current position p robot As the extreme point, the exploration point set consisting of the best exploration point in the candidate frontiers set is sorted by extreme angle as the initial solution of the simulated annealing algorithm.

[0099] X0=[f1,f2,…,f n ], X * =X0, and calculate the energy function Among them, f i represents the i-th exploration point, (x i ,y i ) is fi i The coordinates in the local map coordinate system, (x i+1 ,y i+1 ) is f i+1 Coordinates in the local map coordinate system, x i The x-coordinate and y-coordinate in the local map coordinate system i The coordinate in the y direction in the local map coordinate system;

[0100] S410, let T = kT, k∈[0,1], k is the temperature drop rate;

[0101] S411, randomly exchange the order of two exploration points as the new solution X1=[f2,f1,…,f n ], and calculate the energy function E(X1), ΔE=E(X1)-E(X0);

[0102] S412, based on the Metropolis criterion, determine whether to accept the new solution. If accepted, X * =X1, otherwise proceed to the next step;

[0103]

[0104] Among them, P represents the probability of accepting a new solution. If the current new solution energy value is less than the optimal solution, the new solution is accepted and the optimal solution is updated; if the current new solution energy value is greater than the optimal solution, The probability of accepting the new solution and updating the optimal solution is If the new solution is rejected, the process jumps to step S413;

[0105] S413, iteration number n=n+1, repeat the cycle [3-5] until the termination condition is met.

[0106] like Figure 3 As shown in the figure, the effectiveness of the algorithm of the picking robot is simulated and tested based on the gazebo engine. a is the exploration trajectory of the robot, b is the exploration point calculated based on the greedy strategy, c is the boundary detected by the Dijkstra algorithm, and the gray area represents the unexplored area.

[0107] like Figure 4 As shown, the map established by the picking robot through active exploration. The gray area represents the unexplored area, and the effective coverage rate is 98%.

[0108] In summary, the present invention is an active mapping method applied to an orchard picking robot, which can effectively traverse the entire environment, shorten the exploration path, and effectively reduce labor costs.

[0109] This active mapping algorithm based on a greedy strategy for orchard picking robots improves robot autonomy and reduces labor costs in large-scale scenarios such as orchards and greenhouses, and has significant economic advantages. It integrates multiple sensor data, including but not limited to wheel encoders (Encoders), inertial measurement units (IMUs), and global positioning systems (GPS) to calculate the position of the robot in the world coordinate system, reducing the uncertainty of single sensor positioning. By using the Dijkstra algorithm, the exploration boundary is accurately located to ensure the completeness of the map. At the same time, it can be flexibly expanded to 3D maps to achieve "plug and play" functions. Considering the size, distance, and information entropy of the boundary, a cost-utility function is designed. Based on the greedy strategy, the exploration point selection is completed according to the principle of the highest utility. Finally, the robot's exploration path is optimized based on the simulated annealing algorithm, shortening the exploration path and reducing the energy consumption of the system. By comparing the established map with the actual map area to determine whether the termination condition is met, the limited number of algorithm iterations is guaranteed, the influence of human factors is eliminated, the accuracy and completeness of the established map is guaranteed, and the intelligence of the robot is further improved.

[0110] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0111] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0112] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "a", "an", and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0113] It should be understood that although the terms first, second, third, etc. may be used to describe various information in one or more embodiments of this specification, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0114] The above are only preferred embodiments of one or more embodiments of this specification and are not intended to limit one or more embodiments of this specification. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of this specification should be included in the scope of protection of one or more embodiments of this specification.

Claims

1. A method for active mapping of harvesting robots based on a greedy strategy, characterized in that: The following steps are involved: S1, initialize robot parameters; S2, obtains the measurement values ​​of the robot's surrounding environment and its own position in the world coordinate system; S3, based on the measurement value of the robot's surrounding environment in the world coordinate system in step S2, using the SLAM algorithm to estimate the robot's own position in the local map coordinate system, and construct a local map; S4, based on the measured value of the robot's own position in the world coordinate system in step S2, construct a conversion relationship between the world coordinate system and the local map coordinate system, read the local map, use the robot position as the origin, use the Dijkstra algorithm to perform frontier detection, and use the DBscan algorithm to construct a candidate friontiers set, select the optimal exploration point in the candidate frontiers set according to the greedy strategy, and use the simulated annealing algorithm to minimize the exploration distance; The greedy strategy comprehensively considers the distance between the robot and the frontier, the information entropy brought by the frontier, and the number of points in the circle with the frontier as the center and the area radius R; Select the best exploration point in the candidate frontiers set according to the greedy strategy, which includes the following operations: S44, calculate the distance between the robot's current position and the i-th point in the frontiers set The calculation formula is as follows: Among them, x r is the x-coordinate of the robot in the local map coordinate system, y r is the y-coordinate of the robot in the local map coordinate system, x f is the x-coordinate of the i-th point in the frontiers set in the local map coordinate system, and y f is the y-coordinate of the i-th point in the frontiers set in the local map coordinate system; S45, obtain the number of points contained in the neighborhood radius R of the i-th point in the current candidate frontiers set S46, obtaining the information entropy of the current candidate frontiers set. The information entropy calculation formula is as follows: in, represents the information entropy of the eight neighborhoods of the i-th point in the current frontiers set, p(x i ,y i ) represents the i-th point (x i ,y i ) is the probability of a point in the unexplored region, where x i is the x-coordinate of the i-th point in the frontiers set in the local map coordinate system, and y i is the y-coordinate of the i-th point in the frontiers set in the local map coordinate system; S47, construct a cost utility function based on the greedy strategy, where the greedy strategy comprehensively considers the distance between the robot and the frontier, the information entropy brought by the frontier, and the number of points in the circle with the frontier as the center and the field radius R. The cost utility function is as follows: The optimal exploration point in the current frontiers set is: i * =arg max Cost i ,i∈{1…n} Among them, i * Indicates that the i-th point in the current frontiers set is the optimal exploration point; Use the simulated annealing algorithm to minimize the exploration distance, which includes the following operations: S48, initialize temperature parameter T = T0, n = 0, termination condition ΔE ≤ ∈ | n ≤ N max ∣T≤η, the optimal solution is X * Indicates; where N max is the maximum number of iterations, ∈ represents the minimum energy difference, and η represents the minimum temperature parameter; S49, with the robot's current position p robot As the extreme point, the exploration point set consisting of the best exploration point in the candidate frontiers set is sorted by extreme angle as the initial solution of the simulated annealing algorithm. X0=[f1,f2,…,f n ], X * =X0, and calculate the energy function Among them, f i represents the i-th exploration point, (x i ,y i ) is f i The coordinates in the local map coordinate system, (x i+1 ,y i+1 ) is f i+1 Coordinates in the local map coordinate system, x i The x-coordinate and y-coordinate in the local map coordinate system i The coordinate in the y direction in the local map coordinate system; S410, let T = kT, k∈[0,1], k is the temperature drop rate; S411, randomly exchange the order of two exploration points as the new solution X1=[f2,f1,…,f n ], and calculate the energy function E(X1), ΔE=E(X1)-E(X0); S412, based on the Metropolis criterion, determine whether to accept the new solution. If accepted, X * =X1, otherwise proceed to the next step; Among them, P represents the probability of accepting a new solution. If the energy value of the current new solution is less than the optimal solution, the new solution is accepted. And update the optimal solution; if the current new solution energy value is greater than the optimal solution, The probability of accepting the new solution and updating the optimal solution is If the new solution is rejected, the process jumps to step S413; S413, iteration number n=*+1, repeat the cycle [3-5] until the termination condition is met; S5, based on the local map and the minimum exploration distance, plans a safe trajectory, calculates control commands through the robot kinematic model, and constructs a local augmented map; S6, compare the local augmented map with the actual map area, calculate the percentage of the mapped area, and if the percentage is greater than the termination threshold for mapping, the mapping is completed, otherwise return to step S2.

2. The active mapping method for harvesting robots based on a greedy strategy according to claim 1 is characterized in that: The step S4 constructs a candidate frontiers set, which specifically includes the following operations: Use Dijkstra algorithm for frontier detection, use DBscan algorithm for clustering, calculate the number of frontiers in the frontiers set, and if the minimum number of boundary points n is met k , it is determined to be a candidate frontiers set, otherwise the frontiers set is discarded.

3. The active mapping method for harvesting robots based on a greedy strategy according to claim 1 is characterized in that: The step S4 constructs the conversion relationship between the world coordinate system and the local map coordinate system, specifically including the following operations: S41, integrating the robot wheel encoder, inertial measurement unit and global positioning system information, calculating the robot odometer and the position in the world coordinate system, wherein the coordinate P of the robot in the world coordinate system is calculated. robot as follows: P robot =[x w y w 1] T Among them, x w is the x-coordinate of the robot in the world coordinate system, y w is the y-coordinate of the robot in the world coordinate system; S42, obtain the robot's own position p in the local map coordinate system robot as follows: p robot =[x m y m 1] T Among them, x m is the x-coordinate of the robot in the local map coordinate system, y m is the y-coordinate of the robot in the local map coordinate system, p robot is the coordinate of the robot in the local map coordinate system; S43, based on the robot's own position p in the local map coordinate system in step S3 robot The measured value P of the robot's own position in the world coordinate system in step S2 robot The mapping relationship between the local map and the world coordinate system is calculated. mw as follows: p robot =T mw ×P robot Among them, r is the map resolution, b x is the offset between the local map and the world coordinate system in the x direction, b y is the offset between the local map and the world coordinate system in the y direction, T mw It is the mapping relationship between the local map and the world coordinate system.

Images