Robust Control of Adaptive Model Predictive Control using Online Model Estimation

In recent years, designing an adaptive model predictive controller (AMPC) with varying nonlinear plant parameters in real-time has been a challenging problem. Estimating the model parameters under real-time variations require sufficiently excited signal. Hence, this paper proposes an online model estimation technique for adaptive model predictive control (AMPC) using Recursive Polynomial Model Estimator (RPME). Parameters of the system are continuously varied during real-time to validate the robustness of the controller. Linear plant model parameters are estimated online using RPME and fed to the adaptive model predictive controller to compute the control laws with reference to the step signal. Real-time simulation for nonlinear system response has been conducted using AMPC on Van Der Pol oscillator and Inverted Pendulum System.