| As an important research area of the artificial intelligence, this half century has witnessed the rapid development in robotics. The trajectory planning of the autonomous robot and so on, now the advancing front topic of the artificial intelligence domain, have been attached much importance by scholars of various countries. In the research concerning autonomous mobile robot, navigation technology is the core of its research. The path planning in uncertain and dynamic environment is a key link to the navigation of the autonomous mobile robot, which is an important area and a hotspot of research field. The path planning in environment with numerous obstacles, especially in dynamic environment, is a very complicated issue. A number of scholars are committed to research in this area, but so far the issue has remained improperly settled. In this dissertation an intensive study has been made regarding this issue and the research is as follows:Firstly, after conducting an in-depth research into robust chaos optimization and Alopex optimization algorithm, a hybrid algorithm of chaos and Alopex is proposed, which combines both algorithms' advantages and avoids their weaknesses in view of the Logistic mapping probability distribution non-uniformity and the optimal solution of sketchy searching may deviate from the global optimal solution so as to affect the algorithm search speed. The hybrid algorithm fully elaborates the fast searching power of the improved Alopex and the excellent searching quality of robust chaos optimization so as to improve convergence rate and avoid getting optimal solution only in local interval by using general optimization algorithms. The convergence of both robust chaos optimization and hybrid algorithm are also proofed. With the aid of the proposed algorithm, a path is planned in terms of actual task for mobile robot in typical industry environment with static barriers and mobile obstacle in which the mobile robot keeps away from the stationary barriers and mobile obstacles without collisions, and meets time requirements, dynamics constraints, and kinematical constraints on the optimal planning path from the initial point to the target point, which illustrates that the algorithm has the capability of solving actual problems on a large scale. Secondly, chaos artificial potential field is presented by combining robust chaos optimization algorithm with artificial potential field which is very practical and effective in robot local path planning. A nearly optimal path beyond local interval can be developed effectively in dynamic environment by using this method, which avoids traditional artificial potential field's limitations that it is easy to get optimal solution in local interval and that it is unable to find a path between two nearer obstacles. The chaos artificial potential field has achieved satisfactory results in New NEU robot football team systems of 5 to 5 simulation platform in Northeast University, which has established solid foundation for application of chaos artificial potential field in soccer robot and other robots' path planning.Thirdly, the complexity is probed about the autonomous mobile robot navigation by use of sensor information in dynamic environment. By using the two methods of computing the maximum Lyapunov exponent and drawing power spectrum, it is proofed that Chaotic phenomenon exists in time sequence of distance information that autonomous mobile robot has got from sensors between robots and obstacles. It reveals the navigation of autonomous mobile robots in dynamic environments is the cause of a complex problem.Finally, on the premise of the mobile robot' priority of evading the obstacles in the dynamic environment, the path planning strategy is discussed. The strategy makes the mobile robot evade the obstacles correctly in more and uncertain moving obstacles circumstance. At the same time the obstacles position expression of forecasting k sample is presented when the obstacles movement is uncertain.
|
PDF Full Text Request