An improved fuzzy logic controller using genetic algorithmic techniques for building management systems

摘要

This paper presents the design and control performance of a fuzzy logic controller (FLC) based on the system's output errors for a building managmeent system (BMS). The system comprises a heating, ventilating and air conditioning (HVAC) plant which regualtes the air temperature and relative humidity in a full-scale office room. The plant has 3 control inputs, namely the heater, cooler and humidifier. It has been shwon that good set point regulation can be achieved by employing 5 fuzzy input set membership functions for each output error and 3 fuzzy output set membership functions for each incremental contorl input as well as having good relational rules to determine the control strategy. It is also shown that the performance can be further improved by employing genetic algorithms to search for the 9 optimal fuzzy set output membership functions while maintaining the 10 fuzzy input sets constant. Copyright direct C 1998 IFAC