Robot Path Planning Based On Improved Ant Colony Algorithm

Path planning is a important part in the field of robot navigation technology research, isindispensable. The so-called robot path planning is to point to in unknown environments, to findout a safe path from start to finish, In recent years, research experts and scholars from all over theworld dedicated to the study of robot path planning, and achieved good results, many research outof the path planning method, mainly including artificial potential field method, neural network andgenetic algorithm, the optimization algorithm in path planning problem or other applications, has agood results.Italian scholar m. own et al., first proposed the ant colony algorithm, this algorithm source ofthought is was inspired by the catalytic behavior of ants. Ant colony algorithm has been applied inmany research fields, it not only has the characteristics of positive feedback, but also has theparallelism and the advantages of strong robustness, in solving the problem of robot path planninghas successful application. However, according to the algorithm in robot path planning problem,ant colony algorithm there are still shortcomings, mainly in the process of dealing with problems,easy to fall into local optimal solution and the convergence speed is slow.In this article, aiming at the problem of robot path planning, through the application of antcolony algorithm in robot path planning and analysis, put forward the applicability andeffectiveness of the improved ant colony algorithm. First of robot work space environmentmodeling, the space of the3d model into two plane, using the method of grid of two-dimensionalplane grid division, as environment model of robot path planning. Respectively analyzed thetraditional ant colony algorithm and adaptive ant colony algorithm in the application of robot pathplanning problem, and according to the basic ant colony algorithm, the advantages anddisadvantages of relative to the ant colony algorithm was improved. Secondly, in view of the antcolony algorithm in robot path planning in the slow convergence speed, easy to fall into localoptimal solution of the problem, in order to better optimization results, therefore, in eachoptimization iteration of the ant colony algorithm in the process of introducing the immuneoperator, namely ant colony search process, according to the node selection probability to choosethe next node, until all the ants to the finish, then all constitute the path called the initialpopulation, and then selected, the excellent individuals chosen, after crossover and mutation, tochange some of the genes on the individual, to form a new individual, in order to improve thequality of the solution. Every after ten times of circulation, through nonlinear optimization method,choose the optimal individual. As a result, the improved ant colony algorithm is improved inglobal search space ergodicity and convergence rate, and avoid falling into local optimal solution,it is a kind of convergence and optimization ability are better optimization method, can get theoptimal path of the robot.Respectively in this article, in the environment of the robot’s static and dynamic analysis ofimproved ant colony algorithm for validation, first of all, in static path planning, different sizes ofenvironmental information collected, through the basic ant colony algorithm, an adaptive antcolony algorithm and the improved ant colony algorithm analysis and comparison, the improvedant colony algorithm in convergence speed and the shortest path have achieved good results, thus the effectiveness of the improved ant colony algorithm have been verified and adaptability.Secondly, studied the improved ant colony algorithm in the application of dynamic programming,first by Dijkstra algorithm in the robot working space search out a global shortest path, in theprocess of walking robot, once detected a mobile obstacles ahead, the use of the improved antcolony algorithm in the mobile obstacles around to find out the local target, achieve the result ofdynamic obstacle avoidance.