Dynamic path planning of mobile robots using adaptive dynamic programming

Dynamic path planning has gained increasing popularity in mobile robot navigation. Some of the current path planning methods require a priori information about the motion space and are easily affected by the distribution of obstacles. To address the above limitation, this paper presents a novel dynamic method that transforms path planning into an optimal control problem and solves it dynamically through adaptive dynamic programming and artificial potential field. The proposed method can obtain optimal paths for a differentially-driven mobile robot model in an unknown environment with many irregular obstacles. First, by combining path optimization and kinematical constraints of the mobile robot, the original problem is transformed into a new problem. Second, the total distance traveled, the effect of heading angle, the distance from the target to the robot, and the resultant force of the artificial potential field are included in the new performance index function. Third, the method based on adaptive dynamic programming is developed to avoid obstacles and guarantee the safety of autonomous navigation. The convergence analysis provides theoretical guarantees for our method, and the iterative control sequence will converge to the optimal control. Furthermore, simulation results and analyses under different complexity levels demonstrate that our method has promising performance in exploring and exploiting dynamic path planning problems.