The Research Of Robot Path Planning Of Multi-step And Colony Algorithm

Robot reflects the development leval and ability of science and technology and innovation, The Path planning of the mobile robot is a import field of robot.There are so many obstacles in a enviroment, according to some requirements, robot can find a path form start point to the end point.Ant colony algorithm(ACA) is a kind of parallel and robust evolutionary algorithm, and has been widely used in path planing. but has some disadvantages such as long-time searching and local best solution and so on. Some scholars put forward some improvement methods to solve these problems. for example the elitist strategy was introduced into ACA, but easy to premature convergence. in this paper we put forward an adapt algorithm, the main research as follow:Modeling the environment with grid method, Robot’s environment is divided into shape and size are the same and the ability to fully accommodate robots the size, Robot use path recognition technology to identify path, the Gauss filter process was used to smooth images. convert the RGB color space to the HLS color space, computes and contrast histograms of H and S separately. to find free grid;We put forward a new algorithm which have choose step length, robot not only can choose neighbour grid, but also can choose other grid which is far. which can increase flexibility. Robot can move to the goal grid more quickly;We proposed a new search strategy, according to the principle of shortest straight line between two points, connect the current grid and the goal grid, to find the next grid;Improve heuristic information, the distance between the candidate grid and the target grid should be introduced into heuristic information. which make robot willing to choose the grid which is close to the target and far form the current grid;Improving the way of updating pheromone, when the length of robot’s step longer than 1, the kind of grid which robot don’t stop on it but go through it, adopt the improved update way update pheromone;Using matlab software, carry out a large number of simulation experiments, the improved ant colony algorithm can find a shorter and less moves path, and the algorithm is more intelligent, efficient. which improves the performance of path planing.