Robot Path Planning Based On Genetic Algorithm

A robot is an agent that can sense space,dynamic decision-making,and behavior planning.It can be used as a repetitive and boring job instead of human beings or used to do many tasks that humans cannot perform normally.Robots work in a variety of different environments,avoiding obstacles and finding a viable path,which is a prerequisite for performing related tasks.Path planning is based on the known algorithm,the environment is known,and the environment is completely unknown.According to the corresponding algorithm,a feasible path is obtained from the starting point to avoid obstacles encountered and walk to the end point.Path planning can be static or dynamic.In a static environment,environmental information is known and the robot can plan the path ahead of time.However,in an environment where dynamic or partial information is known,the robot needs to obtain environmental information in time and correct the path in real time.Robots can generally obtain information about the surrounding environment through sensing devices such as cameras and sensors.The processing of environmental information is the basis of modeling.Select the grid method to process the environment information and build the space model of the robot.In order to complete the robot two-dimensional spatial path planning task,this thesis firstly uses the basic genetic algorithm to realize the path planning of the robot,and obtains a feasible path.However,the obtained path length is long and there are many turning points,which leads to low efficiency of the robot path planning.In response to this problem,improvements have been made on the basis of genetic algorithms.Improvements to the initial population,mutation operations,and penalty functions result in fewer robot path turning points,shorter paths,and faster convergence.In addition,the improved genetic algorithm can also make the robot jump out of the local optimal solution and get a feasible path.At present,the path planning of two-dimensional space robots cannot meet the needs of three-dimensional space occasions.This requires further extending the two-dimensional space to obtain a three-dimensional space.For the path planning of three-dimensional space robots,this thesis first constructs a three-dimensional space by grid method and establishes a three-dimensional map model.Then the genetic algorithm and ant colony algorithm are combined to apply to the path planning of three-dimensional space robot,and the feasible path of the robot in three-dimensional space environment is obtained.Finally,the MATLAB software is used to simulate the two-dimensional space and three-dimensional space robot path planning.The two-dimensional space robot path planning is simulated in static environment,dynamic environment and two-dimensional special environment map,and the optimal or sub-optimal path of the robot is obtained.Then,the improved genetic algorithm and the improved ant colony algorithm are simulated respectively,and the simulation results are compared.Simulation results show that the improved genetic algorithm has a shorter path and fewer iterations.In the three-dimensional space environment,the robot path planning in two different environments was simulated.The simulation results show that the robot can find a feasible path in the three-dimensional space environment.