Effectiveness evaluation of updating final-state control for automated guided vehicles motion control with collision avoidance problems

Abstract

This paper discusses the effectiveness of using Updating Final-State Control (UFSC) for Automated Guided Vehicle (AGV) collision-avoidance problems, as an example of rigid body motion control problems that have time-varying states. This paper broadly comprises two parts: 1) clarification of the characteristics of the UFSC through numerical simulations in an ideal system and 2) experimental demonstrations. Regarding the former, this paper performs several numerical simulations and demonstrates the effectiveness of the UFSC through comparison with other methods: PID (Proportional-Integral-Derivative) control with a safety length constraint and MPC (Model Predictive Control). The results show the acceptable performance of the UFSC. In addition, the UFSC has some advantages with respect to the computation time, the effort required for control parameter tuning, and the performance retention of prior obstacle's information and control parameters. For the experimental demonstrations, the authors add the realistic frictions, corresponding to the controlled object, to the one-dimensional model. The experiments are performed under four different conditions, and their results show the validity of the numerical simulations. The discussion in this paper supports the acceptable performance and effectiveness of the UFSC for these types of problems.