Inverse dynamics compensation via ‘simulation of feedback control systems'

Abstract

In many applications within the fields of science, engineering, mathematics, and socio-economics, the ‘inverse problem' (i.e. the problem of determining a system's input from its output) is commonplace. The current paper presents a new concept for solving the inverse problem, which the authors call inverse dynamics compensation via ‘simulation of feedback control systems' (IDCS). IDCS is a numerical method that obtains an approximate inverse dynamic solution via a feedback control simulation, using only the forward dynamic model of the linear or non-linear system. Simulations are ideal environments for feedback control, in the sense that they can be completely noise-free and precisely known, so that the IDCS method can be tuned for the best possible performance. It is seen that IDCS can use the advantages of such ideal control simulation performance in the real world of physical implementations.This paper first introduces the basic idea of IDCS. Then, the effectiveness of the proposed method is demonstrated via three illustrative examples, in which the systems to be controlled have continuous and discontinuous non-linearities. IDCS performance is also investigated via physical experimentation on a non-linear servohydraulic test rig. Finally, extensions of the basic IDCS formulation are introduced which are potentially useful for the solution of additional practical problems.