Research Of Path Planning For Small-Size Intelligent Soccer Robot In Complex And Dynamic Environment

As a symbolic and challenging subject in the field of artificial intelligence, the soccer robot game has received the world-wide recognition and interests. Besides, the development of the study in soccer robot is a strategic step from basic theories to practical application. What's more, path planning is mainly applied in the bottom strategy of robot, whose inferior and superior influence the real-time and the accuracy of the movement and whose performance influences directly the working efficiency of the robot. Till now, the algorithm of path planning in the static environment has received a mount of achievement, meanwhile, the path planning which satisfies the complicated and dynamic environment is still in the further stage of perfection. Therefore, the design of the path algorithm which can be in accordance with the complicated and dynamic environment has the theoretical significance as well as the application value.Path planning means finding an optimum way of keeping off all barriers in the working space from the initial point to the end point by one or some optimized principle (such as the minimum working price, vehicle path and walking time). The main concerned topics include that: a perfect mold for the working environment of the robot should be structured firstly and then a free-obstacle path which is optimized or shortest should be found by certain algorithm. Although the technology of the robot path planning has achieved a lot, no algorithm is perfect and everyone has its advantages and inescapable disadvantages, which needs continued growth. Therefore, this subject concerns about the path planning of the small-sized soccer robot based on the previous studies.The introduction part first talks about the background and the thematic significance of this subject, then, soccer robot and small-size soccer robot are briefly introduced, followed by a summary of the recent studies around the world, evaluation criteria and trends of path planning algorithms. The content and the structure of the thesis lie at the end of the first part.The second part will introduce the particle swarm algorithm used in the local path search. The basic theory of particle swarm algorithm is introduced at first, including the source of the algorithm, analysis of some parameters and the realized process of algorithm; then compare the PSO algorithm with genetic algorithm and ant colony algorithm in the advantages and disadvantages; after that the improvement direction for PSO algorithm are put forward by some scholars; Finally the application of Particle Swarm Optimization algorithm in various fields is presented.The third chapter is the key of whole thesis-the design of the algorithm to realize the path planning of the robot. The grid method in the environment modeling will be introduced first. The grid method is simple and can achieve the optimal solution if the algorithm designed appropriately. What's more, it can combine with several matured algorithms and techniques. Therefore, this subject will adopt the coordinate method of the grid method to construct the working space of the robot in order to plan the overall path. Then, the improvement of the initial parameters and the fitness function of the particle swarm optimization will be operated. Combined with the grid method, the optimal solution can plan the robot's walking path. This method improves the speed and the accuracy of the path planning, and solves the problem that path planning is easily trapped into local extremes and cannot keep up the dynamic obstructions, which will balance between the path length and the walking time to design a free-barrier and slide optimal path.The forth Chapter is a design part of simulation system. In order to verify the feasibility and effectiveness of this chapter, the simulation system is designed using MATLAB language. The interface and functionality of experimental system are described in detail.The fifth chapter shows the validity and practicality of the algorithm by the comparison of the standard PSO algorithm and this topic algorithm. First for proving standard PSO can solve various problems, test functions are applied to it. then the hypothesis of the standard PSO can applied to the path planning of small-size robot is testified by simulation experiment. Finally, standard PSO and improved algorithm are applied to the path planning separately, the simulation results show that improved algorithm is better than standard PSO and has strong reliability.The last chapter is the conclusion and the prospect of this subject. This thesis will conclude the research outcomes and the further improvement.