Dynamic Obstacle Avoidance For Mobile Robot Research Based On Local Environment Information

Since the appearance of the mobile robot, the obstacle avoidance planning has becomeone of the burning issues of robotics research. Obstacle avoidance planning not only needs tomeet the demand to avoid obstacles, but also requires considering the condition factors such astime, distance, which should be optimized as possible. Based on mobile robot workingenvironment exists globally known static obstacles and unknown dynamic obstacles, thispaper studies a mobile robot path planning method based on immune network algorithm torealize the dynamic obstacle avoidance, which is combined with rolling planning while mobilerobot running.In order to fulfil the need for mobile robot obstacle avoidance planning in the complexworking environment, the method of rectangular box surrounding obstacles in mobile robotworking environment is adopted in this paper. Then, visibility graph builds environmentalmodel. To achieve the aim of path planning of mobile robot, this paper combines thecharacteristics of the work environment after treatment with immune network algorithm. Here,for the sake of avoiding algorithm converging to a local optimum and accelerate theconvergence speed of the algorithm, affinity calculating method of immune algorithm andvaccine extracting have been improved. The overall process of mobile robot path planning isperformed by four steps. First, based on known global static environmental information, therobot utilizes the improved immune network algorithm for global path planning; then, themobile robot detects surrounding environmental information while running; once a dynamicobstacle occurs into the sensing range, mobile robot must predict the trajectory of perceiveddynamic barriers and forecast the region where it may collide with mobile robot, this paperonly considers the case that dynamic obstacle moves uniformly in a straight line; moreover,the danger zones are dealt as static obstacles at a certain moment in the future; next, mobilerobot conducts local path planning while obstacle locate in a proper place which in the rangeof mobile robot’s perception; after arriving at local target, mobile robot will return to the pathof global planning; so it finally realizes the combination of local dynamic planning and globalstatic planning. This paper analyses and simulates the above algorithms in differentenvironmental case. The experimental results prove that the algorithm can meet the demand of avoidance planning in both static global environment and dynamic environment, and the pathplanned out is optimal.Based on the combination of visibility graph and immune network algorithm, the localobstacle avoidance planning method of a mobile robot can significantly improve the runningefficiency of the immune network algorithm and effectively overcome the slow convergencespeed of faults. Furthermore, the method has stable ability to work out a better path whether inthe static global environment or local dynamic environment which has unknown partialinformation.