Study On Optimal Path Planning For Landfill Leakage Detection

The output of domestic garbage is increasing day by day as urbanization continues to progress.Due to the relatively simple construction,low cost and large disposal capacity,the primary safe method of treating residential waste in China is a sanitary landfill.In the process of landfill construction,affected by a variety of factors,the anti-seepage layer will appear different degrees of damage,since the landfill leachate contains a large number of toxic pollutants,in order to prevent its leakage,it is of great significance to do a good bare membrane test in the early stage of landfill construction,find the damage and repair it in time.This thesis adopts infrared inspection robot instead of manual inspection based on infrared flaw detection theory,in order to improve the detection efficiency of the robot on the landfill,the inspection robot completes the optimal path planning for full coverage.Firstly,according to the environmental situation of the landfill site,the grid map method is used to complete the environmental modeling of the landfill site,and the grid size,state and the relationship between grids are clearly defined.Then based on the distribution of obstacles,the rectangular decomposition method is used to complete the regional decomposition of the landfill site,and finally several sub-regions are generated.Secondly,based on the generated sub-regions,the template model method and biological excitation neural network algorithm are respectively studied.A template model adapted to the environment was designed for the sub-area overview of the landfill;The bio-excitation neural network algorithm is also improved and optimized to solve the problem of path misjudgment in the boundary area and near the obstacle of the robot,so that the inspection robot can efficiently complete the internal detection of the sub-area.Thirdly,the process algorithm of the genetic algorithm is improved and optimized to complete the solution of the best traversal sequence of the sub-regions.In the selection,tournament method and random traversal selection method are combined,In the crossover and mutation,fuzzy optimization is introduced and combined with Sigmaid function to adaptively adjust the values of crossover probability and mutation probability,so as to enhance the convergence speed and global optimization ability of the algorithm.After that,the A* algorithm is used to complete the transfer path planning between nonadjacent subregions,and the direction angle is introduced in its evaluation function to improve the search speed of the algorithm.Finally,given the initial position of the robot,the full coverage inspection path of the sub-region is planned by the internal traversal method of the sub-region,so that the robot completes the inspection of each sub-region in turn according to the optimal traversal order of the sub-region solved by the genetic algorithm,and improves the A*algorithm to search for the optimal transfer path between the regions when the two subregions are not adjacent,and finally complete the optimal path planning for the full coverage of the landfill.The simulation results demonstrate the feasibility and effectiveness of the full coverage path planning algorithm developed in this thesis.In the two sub-regions internal traversal method,compared with the template model method,the full coverage path generated by the improved bioexcitation neural network algorithm was 1.874% lower in repeat coverage,the path length was 13.59 units less,and the quality of the path generated was higher.At the same time,the improved algorithm in this thesis is simulated and compared,and the efficiency of the improved algorithm in the thesis is verified to plan the full coverage path,and the generated path has a low repeat coverage rate and a shorter driving path.Finally,the test was carried out,and the robot could efficiently complete the full coverage of an area.