抄録
The feasibility in optimal control applications is increasing in recent years with rapid improvement of computational capability. User-friendly practical tools are highly required to meet complex design requirements in the field of aerospace engineering and dynamic programming is regarded as a powerful tool that satisfies these necessities. On the contrary, dynamic programming possesses a drawback so called ‘menace of the expanding grid’ or dimensional difference problem in applications where the number of control variables is fewer than that of state variables. This paper proposes a unique method to approach this issue and reviews its validity and feasibility by discussing a classical problem of Linear Quadratic Regulator (LQR) and an optimal gliding flight of an airplane that the analytically given exact solutions are checked for clarity.
