Mobile Robot Vision Tracking And Path Planning Based On Rough Set

The mobile robotics represents the development frontier of high-tech which is a hot research direction at present, while the path planning, as a crucial link of the autonomous mobile robotic research, is the key technology for intelligent robots, so it also becomes a hotspot.Based on the analysis of the advantages and disadvantages of the various path planning methods, the genetic algorithm is adopted as the base of path planning research. However, the planning speed of existing genetic algorithm for path planning is slow, and in complex environment especially a dynamic environment it is difficult to meet the real-time requirement of mobile robot. Although there are many improved genetic algorithm, but the effect is also not satisfactory. To solve this problem, rough set is introduced into the genetic algorithm in this paper. It focused on the applications of rough set genetic algorithm in a static global path planning of mobile robot, and proposed rough set genetic algorithm based path planning of mobile robot in a dynamic environment.In the static global path planning, at first, under the grid model and by the feasibility of the grids, the initial decision-making table of the robot is obtained, which can be simplified by the rough set theory to extract the minimal decision-making rules. We use these rules to train the initial population of the genetic algorithm (GA), and then solve the best path using GA. Compared with the initial population of GA generated randomly, the initial population of GA simplified by rough sets suggest that the effect of this suggested RSGA is significant at optimizing the robot path planning speed, especially in the complicated environments.In the simulation results and analysis, the experiments are carried out in different environments based on VC++ simulation system. By comparing the experimental data, the influence of population size, crossover probability and mutation probability with different to genetic algorithms is argued which provides a basis for following experiments. Based on the above discussion, fix up the values of the parameters, and choose C++ as the simulation platform, according to the ideas of static global path planning and local path planning, the static and dynamic simulation are carried out respectively, and path planning based on rough set genetic algorithm and simple genetic algorithm are compared. It is obvious that the proposed approach can find the optimal path with good success rates. It is observed that the performance of the proposed algorithm surpasses those of simple genetic algorithm based approaches for this problem.