Development and Analysis of a Predictive Control Algorithm for Embedding in a Microprocessor Controller

Abstract

The aim of this paper is to develop a predictive control algorithm and compare it with thermostat based control. The approach is based on modelling thermostat based control in EnergyPlus, and creating a predictive model using Fast Fourier Transform (FFT). Thermostat based control was modelled in EnergyPlus using a simple single zone building with 42% of south glazing. Predictive model was developed as an FFT response function to heat input, thus obtaining an inverse model and applying it several time steps ahead. The heating load and predictive percentage of dissatisfied of thermostat based control and of the predictive control cases were compared. The results show that in this particular study the predictive control approach leads to just over 20% reduction of energy consumption whilst keeping similar level of thermal comfort. In addition to investigating the modelling of predictive control using Fourier series, the aim of the research is to assess the feasibility of embedding this type of control algorithm in a microprocessor controller.