Research On Path Planning Of Agricultural Robot Based On Artificial Fish Swarm And Dynamic Window Approach

The development of mobile robots and the urgent needs of agricultural development have made it a trend as improving the agricultural efficiency with mobile robots,and the issue of path design becomes an important task in the domain of mobile robot.Path design refers to the process that the robot figures out an obstacle-free path based on certain performance indicators(working cost,path length and walking time)in the environment with obstacles.Based on the environment information-obtaining degree,path designing can be categorized as full-scale path planning and partial path planning.There are varied path-planning algorithms at present,and artificial fish swarm algorithm turns out to be a new group improved smart algorithm.Based on the specific environment of the orchard,this thesis made improvement based on the artificial fish swarm algorithm,which would obtain a good effect in accomplishing the path-planning research of robots for agricultural purpose.This thesis expands research as follows:(1)The production of vector electronic map of the orchard environment based on ARCGIS.It,based on the specific experimental environment of the orchard and starting from the three aspects as the data collection,map symbolization and the final setting and output,analyzed the production method of the vector electronic map on ARCGIS and accomplished the final sketch.(2)The full-scale path planning based on artificial fish swarm algorithm.Targeted on the issues as the easy-to-mature for the artificial fish swarm algorithm in the full-scale path planning and the issue that it could easily get in parts,this thesis introduced the immune mediation mechanism and immune memory characteristics into the artificial fish swarm algorithm,produced a full-scale path planning method named immune-artificial fish swarm algorithm,and conducted analysis on the improved artificial fish swarm algorithm with the three testing functions as Rosenbrock?Griewank?Ackley,proving that the improved method could obtain a best result in preventing it from falling into the parts.At last,a gird environment model was proposed to make comparisons between its path-planning effects and the artificial fish swarm algorithm.The simulation effects prove that the improved immune-artificial fish swarm algorithm could perform better in the speed and performance.(3)The partial path planning based on the improved dynamic window approach.The traditional algorithm considered obstacles as material points while neglecting the factors as the dimensions of the obstacles.This thesis proposed an obstacle-expansion model,which would greatly improve the safety of the robot in operation.(4)The path planning of agricultural robots based on mixed path planning algorithm.This method figured out a full-scale path with the improved artificial fish swarm full-scale planning algorithm under the static environment in the first place.Then it conducted the timely obstacle-escape with the improved dynamic window approach,established an agricultural robot experiment platform,verifying efficiency of this algorithm.