Air-ground heterogeneous robot collaborative planning method
Through the method of building passable maps and ground robots to plan paths, the problem of continuous tasks of open-ground heterogeneous robots in multiple regions is solved, and efficient task execution and reliable target recognition are achieved.
Patent Information
- Application Number
- CN202410685048.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-30
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-05-30
AI Technical Summary
It is difficult for the prior art to effectively design a collaborative planning scheme for air-ground heterogeneous robots in large-area scenarios, especially when the environment and task target information are unknown, how to enable air-ground heterogeneous robots to continuously perform tasks in multiple areas.
The flying robot uses a planning method based on boundary guidance to explore the target area and build a passable map. The ground robot uses a sampling-based planning method to plan the path. The flying robot obtains the target location of the task within the target area. The ground robot plans it. The optimal path completes the task, cycled back and forth until all tasks are completed.
The continuous task cooperation of open-ground heterogeneous robots in multiple regions is realized, ensuring the connectivity of ground robot motion planning, and improving task execution efficiency and reliability of target recognition.
Smart Images

Figure CN118915714A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of multi-robot collaborative planning, and in particular to an air-ground heterogeneous robot collaborative planning method. Background Art
[0002] In the current era of rapid technological progress, intelligent robot technology has penetrated deeply into many fields, providing efficient and precise solutions for tasks such as space exploration, environmental monitoring, emergency rescue and logistics distribution. However, with the increasing complexity of robot tasks, the strategy of relying solely on flying robots or ground robots can no longer fully meet actual needs. Therefore, the air-ground heterogeneous robot system, an innovative platform that combines the advantages of air and ground robots, is increasingly highlighting its significant research value and broad application prospects with its unique functional complementary characteristics, task execution flexibility and excellent environmental adaptability.
[0003] Bloesch et al. used the monocular camera system on the flying robot to build a visual feature map for simultaneous localization and mapping, and densely represented the environment as a height map to provide detailed environmental information for ground robots to plan paths. Although this strategy shows effective support for the global optimal path planning of ground robots in limited-scale environments, the execution efficiency of this method is significantly reduced in large-area scenes. Professor Li Jianqiang's team refined the ground image data collected by the flying robot to build a high-precision ground map, and used a hybrid path planning algorithm to plan the path of the ground robot. When applied to mission scenarios with complex and changing terrain structures, this method is prone to problems such as unclear maps and reduced target recognition reliability. Mueggler et al. proposed an exploration strategy using an autonomous visual system on a flying robot, aiming to accurately identify and guide ground robots to act in traversable areas through an aerial perspective. This method is suitable for scenarios where the target location is known. However, in actual tasks, the target location is often unknown and requires real-time detection and identification.
[0004] Previous research has pioneered a new direction in the study of collaborative planning methods for air-ground heterogeneous robots. In the face of unknown environments and mission target information, how to effectively design a collaborative planning scheme that enables air-ground heterogeneous robots to continuously perform tasks in multiple areas has become an important issue that needs to be explored and solved in this field. Summary of the invention
[0005] In view of this, an object of the present invention is to provide an air-ground heterogeneous robot collaborative planning method, in which the air-ground heterogeneous robots collaborate well, can complete tasks quickly, and can effectively meet the task requirements of the air-ground heterogeneous robots.
[0006] To achieve the above object, the present invention adopts the following technical solution: a space-ground heterogeneous robot collaborative planning method, comprising the following steps:
[0007] Step S1, the flying robot UAV uses a boundary-guided planning method to explore the target area until it reaches the target area, and uses the data obtained by the onboard sensors to build a traversable map;
[0008] Step S2: the ground robot UGV plans a feasible path to the target area using a sampling-based planning method SBP according to the traversable map constructed by the flying robot UAV;
[0009] Step S3: The UAV uses a regional exploration planning method to explore the target area, and while building a traversable map in the area, it also obtains the exact location of the mission target in the area.
[0010] Step S4: After the ground robot UGV reaches the target area, it plans the optimal path to the mission target using the sampling-based planning method SBP according to the traversable map constructed by the flying robot UAV, and completes the mission in the target area;
[0011] Step S5: If there is an unfinished task, the air-ground heterogeneous robot will go to the next task target area and return to step S1 to continue;
[0012] Step S6: If all tasks have been completed, the flying robot UAV moves to the end point using a boundary-guided planning method while building a traversable map;
[0013] Step S7: The ground robot UGV uses a sampling-based planning method to move to the end area x according to the map constructed by the flying robot UAV. end move;
[0014] Step S8: When both the flying robot UAV and the ground robot UGV reach the end area x end , the continuous mission in multiple target areas is completed.
[0015] In a preferred embodiment, the specific method of constructing the traversable map in step S1 is: the flying robot UAV calculates the height value of the position detected by the sensor to obtain the traversable value of the newly discovered position within the current range, and constructs the traversable map. The traversable value calculation formula is:
[0016]
[0017] Among them, G(x) is the pass value of the exploration position x; L is the height of the flying robot UAV exploration position x, L se is the set flying robot UAV height threshold, when L ≥ Lse When L se When >L>0, the pass value G(x)∈(0,1), and only the flying robot UAV can pass through this position; when L=0, the pass value G(x)=0, and both the flying robot UAV and the ground robot UGV can pass through this position; for all unexplored positions, the pass value is -1.
[0018] In a preferred embodiment, the boundary-guided planning method of step S6 specifically includes the following steps:
[0019] Step S61: The flying robot UAV detects the boundary and sets the boundary of the current position of the flying robot UAV With the boundary set B t-1 The intersection of the boundary set B t-1 Remove it and then set the current position of the flying robot UAV to x UAV Deposit into collection V q If it is the first execution, the boundary set No intersection processing is performed, and the current position x of the flying robot UAV is directly UAV Store the bounds search set V q middle;
[0020] Step S62: From the set V q Remove and delete the first position, perform a breadth-first search to explore the positions around it, and add the positions on the border to V new , at non-boundary locations, join the set V q , iterate until the set Get all the boundary positions that the flying robot UAV can reach;
[0021] Step S63: Use the Boundary_Line function to convert adjacent boundary position combinations into boundary set B t , then use the Boundary_Center function to calculate the centroid of all boundaries, and add the boundary position closest to the centroid to the set of points to be explored V p middle;
[0022] Step S64: Treat the exploration point set V p If the judgment value is greater than the threshold, the exploration point with the largest evaluation value is calculated and the exploration point with the largest evaluation value is selected as the current optimal exploration point x. best , the flying robot UAV moves to the exploration point x best , the judgment function is defined as:
[0023]
[0024] in, is the number of locations around position x that the ground robot UGV can pass through, is the number of locations around position x that are inaccessible to the ground robot UGV. Through the judgment function, the traversability judgment can be made, so that the traversable map constructed by the flying robot UAV can ensure the connectivity of the ground robot UGV planning;
[0025] Step S65: If the flying robot UAV reaches the designated location, the planning ends; otherwise, return to step S21 to continue.
[0026] In a preferred embodiment: in the step S63, a set of points to be explored is obtained. When the length of a boundary exceeds a given value, the boundary is evenly divided into short boundaries, and the centroid of each short boundary is calculated, thereby determining the points to be explored on each short boundary.
[0027] In a preferred embodiment: the evaluation function of step S64 is designed, and the evaluation function is composed of a distance evaluation value S, a rotation evaluation value R and an information gain I. The distance evaluation value is the sum of the actual distance from the current position of the flying robot UAV to the exploration point and the estimated distance from the exploration point to the target position, and the formula is:
[0028]
[0029] The rotation evaluation value is the evaluation value of the change in exploration direction of the current position of the flying robot UAV, and the formula is:
[0030]
[0031] Among them, x UAV 、x s ,x,x g They are the current position of the flying robot UAV, the position of the previous exploration point of the current position, the exploration point position and the target position;
[0032] Information gain is the total number of unknown positions x that can be explored by the exploration point. n The number of, the formula is:
[0033]
[0034] The distance evaluation value, rotation evaluation value and information gain are normalized to avoid a certain parameter from affecting the accuracy of the evaluation value. The formula is:
[0035]
[0036] Among them, N is the set of points to be explored V p The number of points to be explored in, i is the exploration point;
[0037] The evaluation function formula is:
[0038] ES i =λ1·S i +λ2·R i +λ3·I i (7)
[0039] Among them, λ1, λ2, and λ3 are cost weight parameters.
[0040] In a preferred embodiment, the specific content of the regional exploration planning method in step S3 is: according to the area of the target area and the exploration range of the flying robot UAV, a set of pre-exploration points V in each target area is solved. area When the flying robot UAV arrives near the target area, the flying robot UAV first ensures that the selected path is passable for both air-ground heterogeneous robots, and then navigates to the current nearest and unexplored pre-exploration point; after arriving at the point, the system updates the status information of all positions within the current exploration range; the reward function is used to quantitatively evaluate the value of the unexplored adjacent pre-exploration points, and the point with the highest reward value is selected as the target of subsequent exploration until all pre-exploration points are processed.
[0041] Compared with the prior art, the present invention has the following beneficial effects:
[0042] (1) This invention proposes a collaborative planning method for air-ground heterogeneous robots to solve the problem of multi-region continuous tasks. The collaborative planning method uses the characteristics of flying robots and ground robots to perform different planning tasks, and comprehensively considers the differences in the traffic capabilities of heterogeneous robots to ensure the connectivity of ground robot motion planning;
[0043] (2) The present invention proposes a flying robot exploration planning method based on boundary guidance, which can quickly guide the flying robot to the target position and use sensor information to construct a traversability map for ground robots to plan. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] Attached Figure 1 It is a flow chart of a preferred embodiment of the present invention;
[0045] Attached Figure 2 Schematic diagram of an air-ground heterogeneous robot heading to a core area in a preferred embodiment of the present invention, wherein (a) is the UAV exploration planning heading area, and (b) is the UGV motion planning driving to the area;
[0046] Attached Figure 3Schematic diagram of an air-ground heterogeneous robot coverage exploration execution task in a preferred embodiment of the present invention, wherein (a) is the UAV coverage exploration area, and (b) is the UGV motion planning to reach the target and leave the area;
[0047] Attached Figure 4 Schematic diagram of the return of an air-ground heterogeneous robot to a safe area in a preferred embodiment of the present invention, wherein (a) is the UAV exploration plan to return to the termination area, and (b) is the UGV motion plan to arrive at the termination area;
[0048] Attached Figure 5 Schematic diagram of flying robot exploration based on boundary detection according to a preferred embodiment of the present invention, wherein (a) is the initial state, (b) is the convenient exploration, (c) is the boundary processing, and (d) is the next exploration point selection;
[0049] Attached Figure 6 The flying robot autonomous exploration area of the preferred embodiment of the present invention Figure 1 Schematic diagram of simulation results;
[0050] Attached Figure 7 The flying robot autonomous exploration area of the preferred embodiment of the present invention Figure 2 Schematic diagram of simulation results;
[0051] Attached Figure 8 Schematic diagram of simulation results of air-ground heterogeneous robot collaborative planning according to a preferred embodiment of the present invention, wherein (a) is the overall process of a UAV multi-area continuous task, and (b) is the overall process of a UGV multi-area continuous task. DETAILED DESCRIPTION
[0052] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0053] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.
[0054] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application; as used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations.
[0055] A collaborative planning method for heterogeneous robots in space and ground, reference Figure 1-8 , including the following steps:
[0056] Step S1: The UAV (Unmanned Aerial Vehicle) uses a boundary-guided planning method to explore the target area until it reaches the target area and uses the data obtained by the sensors to build a traversable map, such as Figure 2 As shown in (a);
[0057] Step S2: The unmanned ground vehicle (UGV) plans a feasible path to the target area using a sampling-based planning method (SBP) based on the traversable map constructed by the UAV, such as Figure 2 as shown in (b);
[0058] Step S3: The UAV uses the regional exploration planning method to explore the target area, and while building a traversable map in the area, it obtains the exact location of the mission target in the area, such as Figure 3 As shown in (a);
[0059] Step S4: After the UGV reaches the target area, it uses the SBP method to plan the optimal path to the mission target based on the traversable map constructed by the UAV to complete the mission in the target area, such as Figure 3 as shown in (b);
[0060] Step S5: If there is an unfinished task, the air-ground heterogeneous robot will go to the next task target area and return to step S1 to continue;
[0061] Step S6: If all tasks have been completed, the UAV moves to the end point using a boundary-guided planning method and constructs a traversable map, such as Figure 4 As shown in (a);
[0062] Step S7: UGV uses the SBP method to move to the end area x according to the map constructed by the UAV. end Move, such as Figure 4 as shown in (b);
[0063] Step S8: When both the UAV and the UGV reach the end area x end , the continuous mission in multiple target areas is completed.
[0064] The specific method of constructing the traversable map in step S1 is: the UAV calculates the altitude value of the position detected by the sensor to obtain the traversable value of the newly discovered position within the current range, and constructs the traversable map. The calculation formula of the traversable value is:
[0065]
[0066] Where G(x) is the pass value of position x. L is the height of the UAV exploring position x, L se is the set UAV height threshold, when L ≥ L se When L se When >L>0, the pass value G(x)∈(0,1), and only UAV can pass through this location. When L=0, the pass value G(x)=0, and both UAV and UGV can pass through this location. For all unexplored locations, the pass value is -1.
[0067] The boundary-guided planning method of steps S2 and S6 specifically includes the following steps:
[0068] Step S21: The UAV detects the boundary and sets the boundary of the UAV's current position With the boundary set B t-1 The intersection of (the boundary set of the previous iteration) from boundary set B t-1 Remove it and then set the UAV current position x UAV Deposit into collection V q If it is the first execution, No intersection processing is performed, and the current position x of the UAV is directly UAV Save to search set V q In Figure 5 As shown in (a);
[0069] Step S22: From V q Remove and delete the first position, perform a breadth-first search to explore the positions around it, and add the positions on the border to V new , added to V at non-boundary locations q , and iterate until Get all the boundary positions that the UAV can reach, such as Figure 5 as shown in (b);
[0070] Step S23: Use the Boundary_Line function to convert adjacent boundary position combinations into boundary set B t , then use the Boundary_Center function to calculate the centroid of all boundaries, and add the boundary position closest to the centroid to the set of points to be explored V p middle;
[0071] Step S24: Treat the exploration point set V p If the judgment value is greater than the threshold, the exploration point with the largest evaluation value is calculated and the exploration point with the largest evaluation value is selected as the current optimal exploration point x. best , the UAV moves to the exploration point xbest , the judgment function is defined as:
[0072]
[0073] in, is the number of locations around position x that the UGV can navigate, is the number of locations around position x that are inaccessible to UGV. By using the judgment function to determine the accessibility, the accessible map constructed by UAV can ensure the connectivity of UGV planning, such as Figure 5 As shown in (d);
[0074] Step S25: If the UAV reaches the designated location, the planning ends; otherwise, the process returns to step S21 to continue.
[0075] In step S23, when the length of a boundary exceeds a given value, the boundary is evenly divided into short boundaries, and the centroid of each short boundary is calculated to determine the points to be explored on each short boundary. This can avoid the problem that an overly long boundary causes only one low-quality exploration point on the boundary. Figure 5 as shown in (c).
[0076] The evaluation function of step S24 is designed, and the evaluation function is composed of a distance evaluation value (S), a rotation evaluation value (R) and an information gain (I). The distance evaluation value is the sum of the actual distance from the current position of the UAV to the exploration point and the estimated distance from the exploration point to the target position, and the formula is:
[0077]
[0078] The rotation evaluation value is the evaluation value of the change in exploration direction of the UAV's current position, and the formula is:
[0079]
[0080] Among them, x UAV 、x s ,x,x g They are the current position of the UAV, the position of the previous exploration point before the current position, the exploration point position, and the target position;
[0081] Information gain is the total number of unknown positions x that can be explored by the exploration point. n The number of, the formula is:
[0082]
[0083] The distance evaluation value, rotation evaluation value and information gain are normalized to avoid a certain parameter from affecting the accuracy of the evaluation value. The formula is:
[0084]
[0085] Among them, N is the set of points to be explored V p The number of points to be explored in, i is the exploration point;
[0086] The evaluation function formula is:
[0087] ES i =λ1·S i +λ2·R i +λ3·I i (7)
[0088] Among them, λ1, λ2, and λ3 are cost weight parameters.
[0089] The specific content of the regional exploration planning method in step S3 is: according to the area of the target area and the exploration range of the UAV, a set of pre-exploration points V in each target area is solved. area When the UAV reaches the vicinity of the target area, the UAV first ensures that the selected path is passable for both air-ground heterogeneous robots, and then navigates to the nearest and unexplored pre-exploration point. After arriving at this point, the system updates the status information of all positions within the current exploration range. Furthermore, the reward function is used to quantitatively evaluate the value of the unexplored adjacent pre-exploration points, and the point with the highest reward value is selected as the target of subsequent exploration until all pre-exploration points are processed.
[0090] Example:
[0091] In this embodiment, the complete pseudo code of exploration planning based on boundary detection is as follows:
[0092]
[0093] The following is a detailed description of the implementation of the present invention using specific experiments. The present invention proposes an air-ground heterogeneous robot collaborative planning algorithm, which mainly verifies the effectiveness of the flying robot's exploration and planning to build a traversable map and the air-ground heterogeneous robot's completion of multi-region continuous tasks through simulation experiments. The specific experimental settings are as follows:
[0094] Flying robot exploration planning and construction of traversable map simulation experiment:
[0095] The simulation experiments were carried out in MATLAB R2020b software.
[0096] (1) Land Figure 1
[0097] The size of the simulation map is X=450m, Y=300m, all obstacles are static obstacles, and the UAV detection range is 50m×50m. The starting point (upper triangle) is (25, 25), and the target point (hexagonal star) is (360, 175). Figure 6 As shown, Figure 1 A simulation map of the map. The white areas in the map are accessible to both UAVs and UGVs, the dark gray areas are accessible only to UAVs, the black areas are inaccessible to both UAVs and UGVs, the light gray areas are unexplored areas, the black lines are UAV paths, and the white areas in the explored areas are connected, indicating that the UGV can go from the starting point to the target point.
[0098] (2) Land Figure 2
[0099] The size of the simulation map is X=450m, Y=300m, all obstacles are static obstacles, and the UAV detection range is 50m×50m. The starting point (upper triangle) is (25, 25), and the target point (hexagonal star) is (350, 150). Figure 7 As shown, Figure 2 A simulation map of the map. The white areas in the map are accessible to both UAVs and UGVs, the dark gray areas are accessible only to UAVs, the black areas are inaccessible to both UAVs and UGVs, the light gray areas are unexplored areas, the black lines are UAV paths, and the white areas in the explored areas are connected, indicating that the UGV can go from the starting point to the target point.
[0100] Simulation experiment of air-ground heterogeneous robots completing multi-region continuous tasks:
[0101] The simulation experiments were carried out in MATLAB R2020b software.
[0102] The size of the simulation map is X=Y=300m, all obstacles are static obstacles, and the UAV detection range is 16m×16m. The starting point (triangle) is (10, 10), and the target point (square) is (290, 290). Figure 8 As shown, Figure 8 (a) and Figure 8 In (b), the dot is the target location, the light gray area is the unexplored area, the white area is accessible to both UAV / UGV, the dark gray area is accessible only to UAV, the black area is inaccessible to both UAV / UGV, the red dotted box is the target area, and the black line is the UAV / UGV path.
[0103] The above description is only a preferred embodiment of the present invention. For ordinary technicians in this field, according to the teachings of the present invention, designing different forms of motion planning algorithms does not require creative labor. Without departing from the principles and spirit of the present invention, all equal changes, modifications, substitutions and variations made within the scope of the patent application of the present invention should fall within the scope of the present invention.
Claims
1. A method for collaborative planning of air-ground heterogeneous robots, characterized in that: The steps include: Step S1, the flying robot UAV uses a boundary-guided planning method to explore the target area until it reaches the target area, and uses the data obtained by the onboard sensors to build a traversable map; Step S2: the ground robot UGV plans a feasible path to the target area using a sampling-based planning method SBP according to the traversable map constructed by the flying robot UAV; Step S3: The UAV uses a regional exploration planning method to explore the target area, and while building a traversable map in the area, it also obtains the exact location of the mission target in the area. Step S4: After the ground robot UGV reaches the target area, it plans the optimal path to the mission target using the sampling-based planning method SBP according to the traversable map constructed by the flying robot UAV, and completes the mission in the target area; Step S5: If there is an unfinished task, the air-ground heterogeneous robot will go to the next task target area and return to step S1 to continue; Step S6: If all tasks have been completed, the flying robot UAV moves to the end point using a boundary-guided planning method while building a traversable map; Step S7: The ground robot UGV uses a sampling-based planning method to move to the end area x according to the map constructed by the flying robot UAV. end move; Step S8: When both the flying robot UAV and the ground robot UGV reach the end area x end , the continuous mission in multiple target areas is completed.
2. The method for collaborative planning of air-ground heterogeneous robots according to claim 1 is characterized in that: The specific method of constructing the traversable map in step S1 is: the flying robot UAV calculates the height value of the position detected by the sensor to obtain the traversable value of the newly discovered position within the current range, and constructs the traversable map. The traversable value calculation formula is: Among them, G(x) is the pass value of the exploration position x; L is the height of the flying robot UAV exploration position x, L se is the set flying robot UAV height threshold, when L ≥ L se When L se When >L>0, the pass value G(x)∈(0,1), and only the flying robot UAV can pass through this position; when L=0, the pass value G(x)=0, and both the flying robot UAV and the ground robot UGV can pass through this position; for all unexplored positions, the pass value is -1.
3. The method for collaborative planning of air-ground heterogeneous robots according to claim 1 is characterized in that: The boundary-guided planning method of step S6 specifically includes the following steps: Step S61: The flying robot UAV detects the boundary and sets the boundary of the current position of the flying robot UAV With the boundary set B t-1 The intersection of the boundary set B t-1 Remove it and then set the current position of the flying robot UAV to x UAV Deposit into collection V q If it is the first execution, the boundary set No intersection processing is performed, and the current position x of the flying robot UAV is directly UAV Store the bounds search set V q middle; Step S62: From the set V q Remove and delete the first position, perform a breadth-first search to explore the positions around it, and add the positions on the border to V new , at non-boundary locations, join the set V q , iterate until the set Get all the boundary positions that the flying robot UAV can reach; Step S63: Use the Boundary_Line function to convert adjacent boundary position combinations into boundary set B t , then use the Boundary_Center function to calculate the centroid of all boundaries, and add the boundary position closest to the centroid to the set of points to be explored V p middle; Step S64: Treat the exploration point set V p If the judgment value is greater than the threshold, the exploration point with the largest evaluation value is calculated and the exploration point with the largest evaluation value is selected as the current optimal exploration point x. best , the flying robot UAV moves to the exploration point x best , the judgment function is defined as: in, is the number of locations around position x that the ground robot UGV can pass through, is the number of locations around position x that are inaccessible to the ground robot UGV. Through the judgment function, the traversability judgment can be made, so that the traversable map constructed by the flying robot UAV can ensure the connectivity of the ground robot UGV planning; Step S65: If the flying robot UAV reaches the designated location, the planning ends; otherwise, return to step S21 to continue.
4. The method for collaborative planning of air-ground heterogeneous robots according to claim 3 is characterized by: In the step S63, a set of points to be explored is obtained. When the length of a boundary exceeds a given value, the boundary is evenly divided into short boundaries, and the centroid of each short boundary is calculated, so as to determine the points to be explored of each short boundary.
5. The method for collaborative planning of air-ground heterogeneous robots according to claim 3 is characterized by: The evaluation function of step S64 is designed. The evaluation function consists of a distance evaluation value S, a rotation evaluation value R and an information gain I. The distance evaluation value is the sum of the actual distance from the current position of the flying robot UAV to the exploration point and the estimated distance from the exploration point to the target position. The formula is: The rotation evaluation value is the evaluation value of the change in exploration direction of the current position of the flying robot UAV, and the formula is: Among them, x UAV 、x s ,x,x g They are the current position of the flying robot UAV, the position of the previous exploration point of the current position, the exploration point position and the target position; Information gain is the total number of unknown positions x that can be explored by the exploration point. n The number of, the formula is: The distance evaluation value, rotation evaluation value and information gain are normalized to avoid a certain parameter from affecting the accuracy of the evaluation value. The formula is: Among them, N is the set of points to be explored V p The number of points to be explored in, i is the exploration point; The evaluation function formula is: ES i =λ1·S i +λ2·R i +λ3·I i (7) Among them, λ1, λ2, and λ3 are cost weight parameters.
6. The method for collaborative planning of air-ground heterogeneous robots according to claim 1, characterized in that: The specific content of the regional exploration planning method in step S3 is: according to the area of the target area and the exploration range of the flying robot UAV, a set of pre-exploration points V in each target area is solved. area When the flying robot UAV arrives near the target area, the flying robot UAV first ensures that the selected path is passable for both air-ground heterogeneous robots, and then navigates to the current nearest and unexplored pre-exploration point; after arriving at the point, the system updates the status information of all positions within the current exploration range; the reward function is used to quantitatively evaluate the value of the unexplored adjacent pre-exploration points, and the point with the highest reward value is selected as the target of subsequent exploration until all pre-exploration points are processed.
Citation Information
Patent Citations
UAV-UGV cooperation method for visual navigation
CN116698028A
AGA-UAV path planning method based on improved constraint multi-objective optimization evolutionary algorithm
CN117032274A
Collaborative engagement for target identification and tracking
US20100017046A1
Symbiotic Unmanned Aerial Vehicle and Unmanned Surface Vehicle System
US20160018224A1
