TED-AJ03-258 USING FUZZY LOGIC IN HEATING CONTROL SYSTEMS :

This paper is focused on the temperature control of heating systems. As a dynamic system, heating equipments and systems usually presents high order dynamics with slow time constants, well-damped behavior, non-linear and/or time-varying characteristics. Moreover, heating systems usually includes amplitude and rate constraints on the control signal, that is, there are limits on the energy available to heat up the environment. All these mentioned characteristics and also the need of energy demand reduction while maintaining a certain level of thermal comfort justifies the use of advanced control strategies instead of the simple and reliable on-off controller. The main objective of this work is to analyze the use of fuzzy logic based controllers in this kind of systems. A methodology for the controller membership function adjustment is proposed, keeping in mind thermal applications. For instance, in the thermal comfort field, there is no need to keep the indoor temperature in a rigid fixed value, a range of temperatures is sufficient to create a comfortable situation. From an economic point of view, there is a need to search for a good trade off between thermal comfort and energy costs. We are acquainted that thermal comfort depends on many variables such as air temperature, radiant temperature, relative humidity, air velocity, metabolism rate and clothing thermal resistance. It is well know that thermal comfort is a wider concept but in the present paper it is represented only by temperature control. Future works will also include humidity control together with temperature control. Thus, the aim is to reduce energy demand and energy costs by keeping the temperature behavior within admissible band. The paper analysis is performed on a lumped approach based non-linear model for room air temperature and humidity. The heat transfer calculation is made considering multi-layer envelope for discretization of energy equation (Fourier's law ), internal radiation effects and air infiltration. The results are compared with the performance of classic strategies and simulations results show the effectiveness of proposed strategy.