System Identification of Rotary Double Inverted Pendulum using Artificial Neural Networks

System Identification has been widely used in obtaining the mathematical model of nonlinear systems. Nonlinear system identification is challenging because of its complexity and unpredictability. The nonlinear system considered in this paper is Rotary Double Inverted Pendulum which is unstable and non-minimum phase system. Inverted pendulum is a well-known benchmark system in control system laboratories which is inherently unstable. In this work full dynamics of the system is derived using classical mechanics and Lagrangian formulation. Artificial neural network is used to identify the model.