Switching-Time Computation for Time-optimal Trajectory Planning of Wheeled Mobile Robots

This paper is to present an approach to solve time-optimal collision-free trajectory planning problems for wheeled mobile robots based on switching time computation (STC) method, with smoothed penalty function method for state inequality constraints. By applying the smoothed penalty function method, the state inequality constraints are aggregated and posed as a final time constraint. Then gradients of the final state with respect to each arc time are deduced, which are used in BFGS to solve the time optimal problem. Based on trajectory planning especially, an effective approach to design the two inputs is also proposed. Numerical results are presented for illustration.