Font Size: a A A

Research On Path Planning Method Of Mobile Robot Based On Neural Network

Posted on:2018-07-16Degree:MasterType:Thesis
Country:ChinaCandidate:M M GongFull Text:PDF
GTID:2348330536981920Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the progress of science and technology,robotics has been considerable development,the robot can free humans from repetitive work in heavy,from industry to public life,robots play an increasingly important role.Path planning of mobile robot is one of the core content of robotics,many scholars have conducted in-depth research,has very important significance.his paper mainly studies a kind of mobile robot path planning method based on the combination of neural network and hybrid particle swarm optimization algorithm.In the environment of information representation,this paper study the multilayer feedforward network,Hopfield neural network and ART neural network in robot path planning,the multilayer feedforward network in describe obstacle information has the advantages of simple calculation,easy parallel,and without the weight training.consider the characteristics of path planning in unknown environment,finally determined to use the multilayer feedforward network.This paper uses a hybrid particle swarm optimization(DHPSO)algorithm to perform the path planning.The standard particle swarm algorithm(SPSO),whose inertia weight decreasing with the iteration number,has a strong global convergence but slow convergence rate in certain iterations.The particle swarm algorithm with a compression factor(PSOCF)has a weak global convergence but fast convergence rate,proposes a hybrid 2-population cross particle swarm algorithm DHPSO,Population one and population two are iterated by SPSO and PSOCF respectively,the superior particles of population one is exchanged to population two in certain number of iterations.DHPSO combines the advantages of SPSO and PSOCF,DHPSO not only has the fastest convergence rate,there is also a strong global convergence.Finally,the simulation experiments of particle swarm optimization and path planning are carried out on the MATLABR2016 a experimental platform.The simulation experiments under the single mode function and multi-modal function prove the superiority of the DHPSO algorithm.The simulation experiment of path planning shows that the algorithm can find a good path in the simple and complex environment,and it shows the effectiveness of the path planning algorithm this paper proposed,and has some practical value.
Keywords/Search Tags:robot, path planning, neural network, particle swarm optimization(PSO)algorithm
PDF Full Text Request
Related items