Development of embedded model predictive controller

Abstract

The rapid improvement of computational power and performances of embedded systems, combined with development of new optimization algorithms with reduced complexity extends the application area of model predictive control. In this paper we describe a method of developing a model predictive controller which can be deployed on general purpose microcontroller without a floating-point unit and used to control systems with small response time. The presented method exploits the benefits of discrete-time model predictive control using Laguerre functions and fixed-point format for representing real numbers with the aim of reducing the computational load. The proposed method is verified by implementing embedded model predictive controller for DC motor angular velocity control.