Updating final-state control methods taking input constraints at final time into account

抄録

In most motion control problems, precise settling in terminal time of control is a crucial specification. Totani and Nishimura, thus proposed the final-state control (FSC) technique which satisfies the required final-state conditions precisely. FSC has been applied to positioning control problems of hard disk drives and other servo systems. Moreover, the authors proposed a real-time updating version of FSC (updating final-state control: UFSC) by taking the application to time-varying final-state conditions into account. UFSC is also one of the promising solutions to many motion control problems. However, huge control inputs often occur at the final control time to suppress the error between the final-state conditions and real state variables, making it challenging to apply UFSC to practical problems. This paper discusses three improved control methods taking the input constraints at the final control time into account to improve feasibility. These are (i) frequency-shaping method (FSM), (ii) time-varying weighting method (TVWM), and (iii) input freezing method (IFM). The effectiveness of the methods has been discussed by comparing the numerical simulation results of a case study. The case study is a mid-air retrieval problem of low-speed-descent objects using fixed-wing unmanned aerial vehicles, which is also discussed in the authors’ prior study. As a result, this paper concludes the IFM based UFSC is the most realistic method for the control problem.