Hybrid PID-fuzzy control scheme for managing energy resources in buildings

• Authors: Benjamin Paris;Julien Eynard;Stéphane Grieu;Monique Polit
• Venue: Applied Soft Computing
• Year: 2011

Both indoor temperature regulation and energy resources management in buildings require the design and the implementation of efficient and readily adaptable control schemes. One can use standard schemes, such as ''on/off'' and PID, or ''advanced'' schemes, such as MPC (Model Predictive Control). Another approach would be considering artificial intelligence tools. In this sense, fuzzy logic allows controlling temperature and managing energy sources, taking advantage of the flexibility offered by linguistic reasoning. With this kind of approaches, both the specific use of a building and the specificities of a proposed energy management strategy can be easily taken into account when designing or adjusting the control scheme, without having to model the process to be controlled. PID controllers being commonly used in buildings engineering, the proposed control scheme is built on the basis of a PID controller. This allows implementing the scheme even if a control system based on such a controller is already in use. So, a hybrid PID-fuzzy scheme is proposed for managing energy resources in buildings, as the combination of two usual control structures based on PID and fuzzy controllers: the ''parallel'' structure (according to the current dynamical state of the considered process, either the PID or the fuzzy controller is selected) and the ''fuzzy supervision'' of a PID controller. To test the scheme in simulation, a building mock-up has been built, instrumented and modeled. Finally, criteria describing the way energy is used and controlled in real-time have been defined with the aim of evaluating both the proposed strategy and the control scheme performance.