Predictive Control of Power Converters and Electrical Drives

12.1 Introduction

When control schemes based on model predictive control (MPC) are implemented experimentally, a large number of calculations are required, introducing a considerable time delay in the actuation. This delay can deteriorate the performance of the system if not considered in the design of the controller.

Compensation of the calculation delay has been considered in several works published to date [1–5]. Similar compensation methods have also been proposed for other predictive control schemes such as deadbeat control [6].

Another source of delay in these types of control schemes appears due to the need for future values of the reference variables in the cost function. Usually, the future reference is considered to be the same as the actual reference, which is a good assumption when the reference is a constant value or the sampling frequency is much higher than the frequency of the reference variable. However, during transients and with sinusoidal references, a delay between the controlled and reference variables appears. In order to eliminate this delay, the future reference variables need to be calculated. Some simple extrapolation methods for calculating the future reference variables are presented in this chapter.