This study proposes an effective thermal control method for thermally comfortable and energy-efficient environments in buildings with double skin envelopes. Four rule-based control logics and an artificial neural network (ANN)-based control logic were developed for the integrated control of openings and cooling systems in summer. Using numerical computer simulations, the performance of the proposed control logics was comparatively tested in terms of thermal performance and energy efficiency.
Analysis results imply that the more detailed rules of thermal control logic were effective to maintain the indoor temperature conditions within comfortable ranges. The ANN-based predictive and adaptive control logic presented its potential as an advanced temperature control method with an increased temperature comfort period, decreased standard deviation of temperature from the center of the comfortable range, and decreased number and ratio of overshoots and undershoots out of the comfort range. The additional rules embedded for control logic or ANN applications yielded a more comfortable temperature environment in an integrated manner according to the properly designed operations of envelope openings and the cooling system. However, logics with additional rules and ANN models consumed more energy for space cooling. Therefore, the rule-based controls with advanced logics or an ANN model are required in case occupant comfort is a primary factor to be satisfied. In other cases, the simple rule-based logic is effectively applied. (C) 2014 Elsevier B.V. All rights reserved.