Minimizing Averaged Error For Dynamic Matrix Control

摘要

This paper presents a new and robust measuring mechanism to the least-squares approach used in Dynamic Matrix Control by minimizing distinctively upcoming errors. This approach entails individual recommendation in which a subsequent move is an average of all individual recommendations. This leads to control the set-point and zone where the differential equation yields an overdamped solution. This solution is then used to approximate the time constants and the analytical solution. Time constants are ultimately derived from the quadratic characteristic equation.