| With the improvement of people's quality of life,the working scene of robots has become more diverse,which propose higher a level of requirements on the development of robot technology.As the basic function of mobile robots,robot path planning technology has always been a research hotspot in the field of robotics.With the continuous maturity of the robot path planning algorithm,people explore the expansion research field according to the application scenarios.Coverage Path Planning(CPP)is one of its expansion areas.Mission planning problems,endurance issues,etc.are hot issues that need to be addressed in this research area.The key reason for the difficulty of endurance is that the target area is too large,and the energy of the single robot is not enough to support the entire coverage task.Through multi-robot collaboration,the reliability and effectiveness of the task can be increased.The sensor data is more comprehensive,and the coverage time is shorter.It reduces the time that a single robot returns to charging and refueling due to insufficient power or fuel consumption,and improves the reliability and robustness of system decision-making and planning,thus overcoming the shortcomings of single robots.However,it also makes the task assignment and path planning problems of multi-robots become more complicated.This paper proposes a method combining the region decomposition technique based on DMCED method and the sub-region path planning method based on genetic algorithm to solve the two problems existing in the above multi-robot.The main work of this paper can be divided into several parts:Firstly,the original area coverage problem is analyzed and defined.The path of the single robot is regarded as the area to be covered by the robot to optimize the original area coverage problem.The regional coverage problem is decomposed into two parts: regional decomposition and sub-region coverage.Furthermore,solutions are proposed for the two parts of the regional coverage problem.Aiming at the problem of multi-robot area decomposition,we propose a method of region decomposition based on the partition matrix of Euclidean distance calculation(DMCED)to complete the task assignment problem of multi-robot.The algorithm calculates the distribution matrix according to the initial evaluation matrix,determines the optimal distribution matrix by adjusting the fair distribution parameters,and divides the target area by using the distribution matrix.Thereby the result of the region decomposition is obtained.For the problem of sub-region coverage,this paper uses genetic algorithm as the method of path planning in sub-region.The initial cell is used as the gene in the genetic algorithm,the initial mini-path is used as the chromosome,the total distance of the subregion path and the time required to cover the entire sub-region are used to calculate the population fitness,and each generation is evaluated according to the population fitness.The pros and cons of each generation of population are assessed based on population fitness.Until the stop condition is met,the last population is output,and the path of the sub-region is obtained.After the path planning is completed,the implementation of the prototype system of the area coverage algorithm is performed.The area coverage path planning tool is designed and implemented to make it easy for users to use the algorithm.The tool combines the DMCED region decomposition method with the genetic algorithm-based sub-region path planning method to path the multi-robot in the simulation environment.The tool not only completes the construction of the initial environment,the distribution and removal of robots and obstacles,presents the results of regional decomposition and the final robot planning path,but also the results of regional decomposition and statistical analysis of path planning.The topic further applies the regional coverage problem from theory to actual research,and has important theoretical and practical value for solving area coverage path planning problems. |
PDF Full Text Request