Potential field aided navigation path planning algorithm

In potential field aided navigation, optimal trajectories is an important task. The decision system is a key component of optimal trajectories. Designing the way in which decisions are taken and the path length in decision making will influence the whole performance. When classical algorithm need complete and precise models of the working space, and in many real scenarios their application is not available. Thus, model free methods for path planning under uncertainty are favorable choice. This paper uses Dyna-H algorithm in potential field aided navigation. The Dyna-H algorithm chooses branches more likely to produce good outcomes than other branches. and it is also a model free online reinforcement learning algorithm. The result of simulation shows that the navigation error of the planned path is less than that of arbitrary path. Improved A* path plan algorithm could enhance the performance of gravity aided