Study On Real-time Path Planning Algorithm Of Mobile Robot In Unknown Environment

With the fast development of new technologies in control theory, sensor, computer science and artificial intelligence, many researchers all over the world are paying attentions to the mobile robot technologies. Mobile robot can be applied widely in many fields, such as scientific research, military support, industrial production and personal services. Real-time path planning is one of the most basic and essential technologies of the mobile robot, and is to find the obstacle-free path from the starting position to the destination based on the environment information which detected by the real-time sensor. Based on the platform of Pioneer3-AT mobile robot, the research on the mobile robot real-time path planning in unknown environment is carried out in this paper. A set of software of real-time path planning is designed and realized, and experiments about real-time path planning are implemented. The main work and achievements of this paper are shown as follows:1. Due to the mobile robot and unknown environment are difficult to accurately model, the real-time path planning method based on fuzzy control is studied. Using the hierarchical fuzzy control structure, the "rule explosion" problem is avoided and the autonomously obstacle-free running of the mobile robot is realized.2. In order to deal with the lacking of "predictability" in real-time path planning method, a new real-time path planning algorithm based on the tabu search and fuzzy control is presented. This algorithm first identifies the visible sub-goal by analyzing the data collected from laser radar and gets the optimum sub-goal through the tabu search which based on the value of heuristic function. Then the sub-goal tracing using the fuzzy control is realized. At the same time, the obstacle-free and optimization path of the mobile robot is implemented by gradual way. The algorithm overcomes the shortcoming of repetition visit, symmetric indecision, local minimum and oscillation. Simulation results verify the correctness and effectiveness of the proposed algorithm.3. The real-time path planning software based on the Client/Server architecture is developed. By using the simulation and real-world experiments, the correctness and effectiveness of the proposed algorithm are verified.In addition, the problems of the mobile robot in practical experiments which discuss the direction of future research are established.