Research On Path Planning Of Robots In Three Dimensional Space Based On An Improved Ant Colony Algorithm

Mobile robot path planning technology is one of core-technology in robot research domain.3-D path planning problem is planning an optimal path from the starting position to the finishing position according to a certain evaluation criteria (such as the shortest path length) in the restricted environment (such as only have fixed obstacles). Recently, There is a growing concern about the study of mobile robots in three dimensional space(3-D) environment, like the MAV(micro air vehicle),the UAV(unmanned aerial vehicle),the UUV(unmanned under water vehicle),the AUV(autonomous underwater vehicle),the wall climbing robot and so on. The problem of the3-D path planning is a basic one, and also is a mark of the intelligence. People have made a lot of study about the3-D path planning problem and got many achievements, but there is also some problem deserved to be studied deeply.Ant colony optimization(ACO) is a metaheuristic approach for solving combinatorial optimization problems. Because of the great performance of ACO, researchers show great interest in the research and applications on ACO nowadays.According to the present condition of the studying and the intellective development, this paper presents a new path planning approach based on the improved ACO. The following is the main works:Firstly, this paper introduces the common used planning algorithms for intelligent mobile robot in the country and abroad, and the researching situation is elaborated and summarized. It also analyzes the advantages and disadvantages of each algorithm which build an important basis for the research on mobile robot path planning in the thesis.Secondly, it introduces the ideas of basic ACO and the mathematical model, and describes the influence on the performance of the basic ACO which comes from parameters. It also introduces the performance indexes of the algorithm, and summarizes the advantages and disadvantages of the basic ACO.Thirdly, it helps set up a simple3-D workspace, in which grid is used to model the robot global path planning workspace. We improved the basic ACO,and the improved ACO is used for the path planning of mobile robot in3-D;Matlab software is used to make simulation experiment,the simulation results show that the improved ACO is better than the basic ACO in the path of the optimal value and planning time.Finally, by setting up a3-D terrain environment, grid is also used to model the robot global path planning workspace; the improved ACO is used for the path planning of mobile robot in3-D terrain environment, Matlab software is used to make simulation experiment, and the simulation results show that the improved ACO is better than the basic ACO in the path of the optimal value and planning time; and also optimum performance, time performance and robust performance of improved ACO are also compared with those of basic ACO, the results showed that the improved ACO is better than the basic ACO; then by analyzing the convergence of the improved ACO, the conclusion showed that the improved ACO is feasible.In this thesis, ACO is improved as follows:(1) In the process of applying ACO in3-D path planning, ants releasing pheromones on nodes rather than in the edge between nodes to reduce the huge distance matrix, saving storage space and reducing computational complexity.(2) The design of heuristic function contains the shortest path and the roadside constraints, and improves the validity of the algorithm (3) The bringing in of global pheromone and part pheromone update mechanism can overcome the shortcoming of the basic ACO such as taking a long time, emerging the stagnation easily etc.