Experimental comparison of integral approximations for state predictive control : - Stabilization control of the inverted pendulum with flexible arm -

This paper presents the experimental results of an inverted pendulum with a flexible arm and artificial time delay. State predictive control requires integral approximation. Depending on the approximation method, the approximation error can cause an unstable response. This problem has been solved in two ways: the low-pass approximations reported by Mirkin and the convolution approximation reported by Zhong. Experimentally, we show that the inverted pendulum with a time delay is stabilized by approximating the distributed-delay control laws. Although both approximations stabilize the system, a combination method that merges the low-pass and convolution approximations is more effective than either method alone.