CN114237226A - Navigation escaping method under complex environment area coverage of swarm robots - Google Patents
Navigation escaping method under complex environment area coverage of swarm robots Download PDFInfo
- Publication number
- CN114237226A CN114237226A CN202111384047.9A CN202111384047A CN114237226A CN 114237226 A CN114237226 A CN 114237226A CN 202111384047 A CN202111384047 A CN 202111384047A CN 114237226 A CN114237226 A CN 114237226A
- Authority
- CN
- China
- Prior art keywords
- robot
- grid
- navigation
- value
- coverage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 230000001413 cellular effect Effects 0.000 claims abstract description 20
- 230000000694 effects Effects 0.000 claims description 43
- 210000002569 neuron Anatomy 0.000 claims description 38
- 210000004027 cell Anatomy 0.000 claims description 20
- 230000000638 stimulation Effects 0.000 claims description 7
- 238000010586 diagram Methods 0.000 claims description 6
- 230000004888 barrier function Effects 0.000 claims description 4
- 238000003062 neural network model Methods 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 abstract description 7
- 230000008569 process Effects 0.000 abstract description 7
- 238000004364 calculation method Methods 0.000 abstract description 6
- 238000004891 communication Methods 0.000 abstract description 5
- 230000002457 bidirectional effect Effects 0.000 abstract description 3
- 238000004088 simulation Methods 0.000 description 9
- 238000001514 detection method Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 210000002457 barrier cell Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0221—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving a learning process
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Feedback Control In General (AREA)
Abstract
The invention relates to a navigation and trapped-free method under the coverage of a swarm robot complex environment area, and provides a navigation and trapped-free mechanism which is suitable for an unknown environment and has smaller calculation amount so that a robot escapes from a deadlock, in order to solve the problem of low coverage efficiency caused by the fact that the swarm robot is trapped into the deadlock when the swarm robot executes the area coverage in the unknown complex environment under the condition of limited communication. When the robot falls into deadlock in the process of covering the area, the optimal navigation escaping point of the robot is found by introducing a cellular automaton mechanism, and the robot escapes from the dead zone by improving a BRRT (bidirectional rapid spread star) navigation algorithm to plan the path, so that the high-efficiency covering of the swarm robot to the complex environment is realized.
Description
Technical Field
The invention belongs to the field of swarm robot control, and relates to a navigation escaping method under the coverage of a swarm robot complex environment area.
Background
The distributed swarm robot area coverage is to autonomously and efficiently explore an unknown area by using a large-scale robot so as to obtain map information which is as perfect as possible or realize a certain specific task. At present, heuristic algorithms are mostly adopted at home and abroad to solve the problem of robot swarm region coverage, compared with random coverage, the coverage efficiency of the methods is greatly improved, however, coverage to a simple environment can be mostly realized under the condition of limited communication and the like, and for the environment with complex obstacle distribution, a robot often gets into deadlock in the coverage process, namely, the robot gets into a certain narrow region and cannot escape to cause a large amount of repeated coverage, so that the robot cannot work normally, the coverage efficiency is low, and the corresponding navigation escaping technology is required to be adopted to enable the robot to escape from the deadlock.
The traditional navigation methods such as RRT and BRRT enable the robot to better escape from deadlock through path planning. However, most of the conventional navigation methods are only suitable for scenes with known coverage environments and target point positions, have the problems of instability and large calculation amount, and are difficult to be suitable for swarm robots with high real-time requirements and limited calculation capacity and complex and variable actual environments.
Disclosure of Invention
Technical problem to be solved
In order to avoid the defects of the prior art, the invention provides a navigation escaping method under the coverage of a swarm robot complex environment area, which solves the problem of low coverage efficiency caused by deadlock when the swarm robot under the limited communication executes the area coverage in an unknown complex environment.
Technical scheme
A navigation escaping method under the coverage of a swarm robot complex environment area is characterized by comprising the following steps:
step 1: the method comprises the steps of rasterizing a complex environment area into square discrete units with the same size according to an interval delta L to obtain a grid map, wherein the size of each grid is the coverage area of one robot, and endowing each robot with the grid map;
step 2: updating the neural network diagram according to the following formula of the bio-excitation neural network model
Wherein: h iscThe activity value of the neurons in the grid c is the stimulation amount; i iscIs an external input to grid c; l is the number of neurons in the neighborhood of grid c; w is ackTo connect the weight(ii) a A. B and D are positive constants;
by giving input Ic:
Obtaining the activity value of the neuron not covered with the grid to be close to 1, the activity value of the neuron covered with the grid to be close to 0, and the activity value of the neuron covered with the barrier grid to be close to-1;
and step 3: judging whether the current robot is in a deadlock state, namely judging whether the activity values of eight grid neurons around the robot are all less than or equal to 0, if the activity values of the eight grid neurons around the robot are all less than or equal to 0, indicating that neighbor grids around the robot are all detected or are obstacles, and executing the step 6 if the robot is in a deadlock state, otherwise executing the step 4;
and 4, step 4: determining the next decision of the robot according to the neuron activity value, namely selecting a grid with the neuron activity value close to 1 as a target point of the next step of the robot;
and 5: judging whether the robot meets the coverage requirement, if so, ending the coverage, otherwise, returning to the step 2;
step 6: the optimal navigation escaping point of the robot is obtained by adopting a cellular automata algorithm, and the following new rule of the cellular state value table is as follows:
1. initializing a cell state value table: the current grid cell state value of the robot is 1, the barrier grid cell state value is-1, and the other grid cell values are 0;
2. evolution rules of the cell state value table: when the state value of the central cellular is-1, the value is unchanged, otherwise, the state value of the central cellular is updated by adding 1 to the minimum value of the cellular values which are more than or equal to 1 in 8 neighbors of the central cellular;
updating a cellular state value table by using the rules, wherein each value in the table is the step number from the trapped point of the robot, and selecting the uncovered point with the least step number as the navigation trapped point;
and 7: executing the robot in the step 6, and performing path navigation by adopting a BRRT navigation algorithm based on the neuron network activity value to enable the robot to escape from a deadlock state; returning to step 5.
Advantageous effects
The invention provides a navigation and escaping method under the coverage of a swarm robot complex environment area, aiming at solving the problem that the swarm robot is trapped in deadlock when the swarm robot executes the area coverage in an unknown complex environment under the condition of limited communication, so that the coverage efficiency is low. When the robot falls into deadlock in the process of covering the area, the optimal navigation escaping point of the robot is found by introducing a cellular automaton mechanism, and the robot escapes from the dead zone by improving a BRRT (bidirectional rapid spread star) navigation algorithm to plan the path, so that the high-efficiency covering of the swarm robot to the complex environment is realized.
On the basis of the traditional BRRT algorithm, the BRRT algorithm is improved by utilizing the characteristic that the grid neuron activity value closer to an undetected area is larger; in the BRRT navigation algorithm method based on the neuron network activity value, the magnitude between the neuron activity value of a new node and the neuron activity value of a father node of the new node is compared before collision detection is carried out, if the father node is a point on an initial point tree, the next collision detection is carried out when the neuron activity value of the new node is larger than the neuron activity value of the father node of the new node, and if not, the random point is generated again; if the father node is a point on the escaping point tree, performing next collision detection when the neuron activity value of the new node is smaller than the neuron activity value of the father node, and otherwise, returning to regenerate the random point; the method is used for guiding the generation of random points, so that blind sampling is reduced, and the path planning speed and the algorithm stability are improved.
Drawings
FIG. 1: BRRT navigation algorithm based on neuron network activity value
FIG. 2: flow chart of navigation escaping method under complex environment area coverage of swarm robots
FIG. 3: cell state value table following process
FIG. 4: BRRT navigation algorithm simulation contrast diagram before and after improvement
FIG. 5: grid map of complex environment
FIG. 6: robot stuck-in dead zone illustration
FIG. 7: initial position diagram of robot
FIG. 8: robot complete coverage map
FIG. 9: pre-and post-simulation contrast by adding navigation escaping mechanism
(a) Average visit number simulation comparison graph, (b) total moving step simulation comparison graph
Detailed Description
The invention will now be further described with reference to the following examples and drawings:
the invention provides a navigation trap-escaping mechanism which is applicable to unknown environments and has smaller calculation amount, so that a robot escapes from deadlock. When the robot falls into deadlock in the area covering process, the optimal navigation escaping point of the robot is found by introducing a cellular automaton mechanism, and the robot escapes from a dead zone by improving a BRRT (bidirectional rapid spread star) navigation algorithm to plan a path, so that the high-efficiency covering of the swarm robot on the complex environment is realized, and the swarm robot is more suitable for the actual environment.
The invention uses the size of the environmental stimulus to represent the coverage degree of the robot to the unknown environment in the process of area coverage, and uses the biological excitation neural network model to update the size of the stimulus, and the dynamic change rule is represented by the following formula:
wherein h iscThe activity value of the neurons in the grid c is the stimulation amount; i iscIs an external input to grid c; l is the number of neurons in the neighborhood of grid c; w is ackIs a connection weight; A. b and D are positive constants.
By giving input Ic:
It can be found that the uncovered grid stimulation amount is close to 1, the covered grid stimulation amount is close to 0, and the obstacle grid stimulation amount is close to-1. The robot is guided to move to a grid with the stimulation amount close to 1, so that the whole environment area is covered.
When the activity values of the grid where the robot is located and the 8 grids around the robot are less than or equal to 0, namely the neighbor grids around the robot are all detected or are obstacles, the robot is indicated to be involved in deadlock. The invention adopts a cellular automata algorithm to obtain the optimal navigation escaping point in real time. The following rule of the cell state value table is as follows:
1. initializing a cell state value table: the current cell state value of the robot is 1, the barrier cell state value is-1, and the remaining cell values are 0, as shown in fig. 3 (a).
2. Evolution rules of the cell state value table: when the central cell state value is-1, the value is not changed, otherwise, the central cell state value is updated by adding 1 to the minimum value of the cell values greater than or equal to 1 in the 8 neighbors of the central cell, as shown in fig. 3 (b).
And updating a cellular state value table by using the rule, wherein each value in the table is the step number from the trapped point of the robot, and the uncovered point with the least step number is selected as the navigation trapped point.
In order to reduce the calculation amount of the navigation planning process, the invention adopts a BRRT navigation algorithm based on the neuron network activity value to carry out path navigation, so that the robot reaches the escaping target point. The traditional BRRT navigation algorithm plans a path by randomly sampling in the target space, but blind sampling results in many invalid nodes. In order to reduce the calculation amount, the BRRT navigation algorithm is improved by utilizing the characteristic that the grid activity value is larger when the grid activity value is closer to an undetected area. The flow chart is shown in figure 1:
in the BRRT navigation algorithm method based on the neuron network activity value, the magnitude between the neuron activity value of a new node and the parent node activity value of the new node is compared before collision detection is carried out, so that blind sampling is reduced. The method greatly simplifies the calculated amount, improves the path planning speed and the stability of the algorithm, and a simulation comparison graph is shown in figure 4.
The general flow chart of the navigation escaping method under the complex environment area coverage of the swarm robot is shown in the following figure 2.
The navigation escaping method under the coverage of the swarm robot complex environment area comprises the following specific steps:
the complex environment area is rasterized into square discrete units with the same size according to an interval delta L to obtain a grid map, wherein each grid is the coverage area of one robot, and the grid map is given to each robot, as shown in FIG. 5.
Updating the neural network diagram according to the formulas (5) to (6):
and thirdly, judging whether the current position is in a deadlock state, namely whether the activity values of eight grid neurons around the robot are all less than or equal to 0, as shown in the attached figure 6. If so, executing step (c), otherwise, executing step (d).
Fourthly, determining the next decision of the robot according to the neuron activity value, namely selecting the grid with the neuron activity value close to 1 as the target point of the next step by the robot
Judging whether the robot reaches the covering requirement, if so, ending the covering, otherwise, returning to the step II.
And sixthly, obtaining the optimal navigation escaping point of the robot through a cellular automaton algorithm.
And seventhly, executing the robot after the step sixthly, and performing path navigation by adopting a BRRT navigation algorithm based on the neuron network activity value to enable the robot to escape from the trapped point. And returning to the step (v).
Simulation analysis is carried out on the swarm robots according to the steps, and simulation results are shown in the attached figures 7, 8 and 9.
In the embodiment, a grid map with a coverage area of 100m × 100m and a discrete coverage area of 100 × 100 is selected, the robot system is composed of 10 to 50 isomorphic robots, and the initial positions of the robots are randomly generated in the coverage area, as shown in the green grid in fig. 7. The external input E is 100, the decay rate A is 5, the upper limit of the activity value B is 1, and the lower limit of the activity value-D is-1.
The independent repeated experiments are carried out for 50 times under the same condition, and the experimental result shown in the attached figure 9 in the simulation effect diagram shows that the provided navigation escaping mechanism solves the problem that the swarm robots with limited communication are trapped in deadlock and cannot escape when the swarm robots carry out regional coverage in an unknown complex environment, so that the coverage efficiency is greatly improved, and the high efficiency of the invention is verified by comparing with the coverage simulation result under the basic biostimulation neural network model.
Variations, modifications, substitutions and alterations of the above-described embodiments may be made by those skilled in the art without departing from the principles and spirit of the invention.
Claims (1)
1. A navigation escaping method under the coverage of a swarm robot complex environment area is characterized by comprising the following steps:
step 1: the method comprises the steps of rasterizing a complex environment area into square discrete units with the same size according to an interval delta L to obtain a grid map, wherein the size of each grid is the coverage area of one robot, and endowing each robot with the grid map;
step 2: updating the neural network diagram according to the following formula of the bio-excitation neural network model
Wherein: h iscThe activity value of the neurons in the grid c is the stimulation amount; i iscIs an external input to grid c; l is the number of neurons in the neighborhood of grid c; w is ackIs a connection weight; A. b and D are positive constants;
by giving input Ic:
Obtaining the activity value of the neuron not covered with the grid to be close to 1, the activity value of the neuron covered with the grid to be close to 0, and the activity value of the neuron covered with the barrier grid to be close to-1;
and step 3: judging whether the current robot is in a deadlock state, namely judging whether the activity values of eight grid neurons around the robot are all less than or equal to 0, if the activity values of the eight grid neurons around the robot are all less than or equal to 0, indicating that neighbor grids around the robot are all detected or are obstacles, and executing the step 6 if the robot is in a deadlock state, otherwise executing the step 4;
and 4, step 4: determining the next decision of the robot according to the neuron activity value, namely selecting a grid with the neuron activity value close to 1 as a target point of the next step of the robot;
and 5: judging whether the robot meets the coverage requirement, if so, ending the coverage, otherwise, returning to the step 2;
step 6: the optimal navigation escaping point of the robot is obtained by adopting a cellular automata algorithm, and the following new rule of the cellular state value table is as follows:
1. initializing a cell state value table: the current grid cell state value of the robot is 1, the barrier grid cell state value is-1, and the other grid cell values are 0;
2. evolution rules of the cell state value table: when the state value of the central cellular is-1, the value is unchanged, otherwise, the state value of the central cellular is updated by adding 1 to the minimum value of the cellular values which are more than or equal to 1 in 8 neighbors of the central cellular;
updating a cellular state value table by using the rules, wherein each value in the table is the step number from the trapped point of the robot, and selecting the uncovered point with the least step number as the navigation trapped point;
and 7: executing the robot in the step 6, and performing path navigation by adopting a BRRT navigation algorithm based on the neuron network activity value to enable the robot to escape from a deadlock state; returning to step 5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111384047.9A CN114237226B (en) | 2021-11-10 | 2021-11-10 | Navigation escaping method under complex environment area coverage of group robot |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111384047.9A CN114237226B (en) | 2021-11-10 | 2021-11-10 | Navigation escaping method under complex environment area coverage of group robot |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114237226A true CN114237226A (en) | 2022-03-25 |
CN114237226B CN114237226B (en) | 2023-06-30 |
Family
ID=80750303
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111384047.9A Active CN114237226B (en) | 2021-11-10 | 2021-11-10 | Navigation escaping method under complex environment area coverage of group robot |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114237226B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016045615A1 (en) * | 2014-09-25 | 2016-03-31 | 科沃斯机器人有限公司 | Robot static path planning method |
CN106843216A (en) * | 2017-02-15 | 2017-06-13 | 北京大学深圳研究生院 | A kind of complete traverse path planing method of biological excitation robot based on backtracking search |
-
2021
- 2021-11-10 CN CN202111384047.9A patent/CN114237226B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016045615A1 (en) * | 2014-09-25 | 2016-03-31 | 科沃斯机器人有限公司 | Robot static path planning method |
CN106843216A (en) * | 2017-02-15 | 2017-06-13 | 北京大学深圳研究生院 | A kind of complete traverse path planing method of biological excitation robot based on backtracking search |
Non-Patent Citations (1)
Title |
---|
陆亮;许双伟;: "基于Hopfield神经网络的路径规划研究", 电脑编程技巧与维护, no. 12 * |
Also Published As
Publication number | Publication date |
---|---|
CN114237226B (en) | 2023-06-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112461247B (en) | Robot path planning method based on self-adaptive sparrow search algorithm | |
CN109164810B (en) | Robot self-adaptive dynamic path planning method based on ant colony-clustering algorithm | |
CN109116841B (en) | Path planning smooth optimization method based on ant colony algorithm | |
CN106695802A (en) | Improved RRT<*> obstacle avoidance motion planning method based on multi-degree-of-freedom mechanical arm | |
CN107992040B (en) | Robot path planning method based on combination of map grid and QPSO algorithm | |
CN105527964A (en) | Robot path planning method | |
CN113917925B (en) | Mobile robot path planning method based on improved genetic algorithm | |
Fetanat et al. | Optimization of dynamic mobile robot path planning based on evolutionary methods | |
CN115297484B (en) | Sensor network coverage rate optimization method based on novel compact particle swarm algorithm | |
Li et al. | Path planning of mobile robot based on genetic algorithm and gene rearrangement | |
Yu et al. | A novel parallel ant colony optimization algorithm for warehouse path planning | |
CN111474925B (en) | Path planning method for irregular-shape mobile robot | |
CN116772880A (en) | Unmanned aerial vehicle path planning method based on unmanned aerial vehicle vision | |
CN113467481B (en) | Path planning method based on improved Sarsa algorithm | |
Zong et al. | Whale optimization algorithm based on Levy flight and memory for static smooth path planning | |
Shibata et al. | Coordination in evolutionary multi-agent-robotic system using fuzzy and genetic algorithm | |
CN114237226A (en) | Navigation escaping method under complex environment area coverage of swarm robots | |
Zhan | Research on path planning method of humanoid robot based on improved genetic algorithm | |
CN112148030B (en) | Underwater glider path planning method based on heuristic algorithm | |
Chen et al. | Path Planning of Mobile Robot Based on Improved Differential Evolution Algorithm | |
CN111796597A (en) | Robot multi-target path optimization method based on novel firefly group optimization | |
Shibata et al. | Coordinative behavior in evolutionary multi-agent system by genetic algorithm | |
Figurowski et al. | Hybrid path planning for mobile robot using known environment model with semantic layer | |
Shiltagh et al. | A comparative study: Modified particle swarm optimization and modified genetic algorithm for global mobile robot navigation | |
Li et al. | Path planning algorithm for mobile robot based on improved ant colony algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |