A Mechatronic System Design Case Study: Adaptive Modeling and Control of an Inverted Pendulum

In this paper, we address a class of underactuated nonlinear systems having fewer measurable state variables than the number of actuators. In some practical systems, the problem of variation of frequency of sampling of measurable state variables finally leads to failure in the process of control. We present a hybrid method of control for a typical class of nonlinear systems that successfully handles this problem. The practical aspects of the designed experiment, both in hardware and software, are addressed. The proposed algorithm is implemented on a single rode inverted pendulum (IP) as a typical example of the mentioned class of nonlinear systems. A software process of modeling, adaptive modeling (AM), is explained to reduce the density of noise in hardware process of sampling.