Dynamic obstacle avoidance for path planning and control on intelligent vehicle based on the risk of collision

摘要

To improve the autonomy of intelligent vehicle in complex or uncertain environment, a dynamic planning control method for obstacle avoidance has been studied. By leading the degree of risk collision into control system as the input and using an improved fuzzy control algorithm, the input and output of the fuzzy controller can be considered as extract amount. The concept of patterns and pattern matching was used, and according to the matching degree of each rule, the weighted average algorithm was applied to determine the output of control action, this method can avoid a dynamic obstacle in a timely manner and shorten the computing time. At last, by simulation, it was verified that the dynamic control method can make intelligent vehicle avoid the dynamic obstacles independently, and walking toward the target point exactly. And the results also provide a theoretical basis for the realization of the intelligent vehicle moving independently and safely in the complex and dynamic environment.