Advanced controllers for building energy management systems. Advanced controllers based on traditional mathematical methods (MIMO P+I, state-space, adaptive solutions with constraints) and intelligent solutions (fuzzy logic and genetic algorithms) are investigated for humidifying, ventilating and air-conditioning applications.

摘要

This thesis presents the design and implementation of control strategies for buildingudenergy management systems (BEMS). The controllers considered include the multi PI-loop controllers, state-space designs, constrained input and output MIMO adaptiveudcontrollers, fuzzy logic solutions and genetic algorithm techniques. The controludperformances of the designs developed using the various methods based on aspects suchudas regulation errors squared, energy consumptions and the settling periods areudinvestigated for different designs. The aim of the control strategy is to regulate the roomudtemperature and the humidity to required comfort levels.udIn this study the building system under study is a 3 input/ 2 output system subject to external disturbances/effects. The three inputs are heating, cooling and humidification,udand the 2 outputs are room air temperature and relative humidity. The externaluddisturbances consist of climatic effects and other stochastic influences. The study isudcarried out within a simulation environment using the mathematical model of the testudroom at Loughborough University and the designed control solutions are verifiedudthrough experimental trials using the full-scale BMS facility at the University ofudBradford.