Design, development and testing of a fuzzy logic and rough set controller for HVAC systems

An HVAC (Heating, Ventilating and Air Conditioning) system is essential to a building or a house in order to keep the occupants comfortable. However, many HVAC systems do not maintain a uniform temperature throughout the structure, and have cold or drafty sections, because they employ either an on-off control method based on thermostats or a proportional control method. In addition, energy is wasted because their inability to maintaining a steady temperature. On-off control or proportional control systems are used because of their simplicity and low cost. To maintain a steady temperature, a more sophisticated control system must be used.; This thesis presents a new controller design for an HVAC system which employs one of two types of inference algorithms. One is based on fuzzy approximate reasoning, and the second is an application of rough set theory. The controller uses a microprocessor and deduces the appropriate control gains from sensors. The system is capable of controlling temperature and humidity. To keep the temperature constant, the controller adjusts the flow of hot water for heating in winter, and the flow of chilled air for cooling in summer. Humidity is controlled by turning a humidifier on and off.; The system performance for a proportional control method, fuzzy approximate reasoning and rough set method are examined and compared. The results show that fuzzy approximate reasoning and rough set method perform much better than the proportional control method. The control system using fuzzy approximate reasoning shows slightly better performance than the system using rough set method.