Balancing indoor thermal comfort and energy consumption of ACMV systems via sparse swarm algorithms in optimizations

This paper proposes a systematic modelling and optimizing of energy consumption and indoor thermal comfort for air-conditioning and mechanical ventilation (ACMV) systems. The models of extreme learning machines (ELM) and neural networks (NN) are established and evaluated. These well-trained models are then integrated with the computational intelligence techniques of sparse firefly algorithm (sFA) and sparse augmented firefly algorithm (sAFA). The sFA and sAFA aim to locate the global optimal operating points of the ACMV systems in real-time and predict energy saving rate (ESR) with a third order polynomial regression based on minimizing the mean squared errors (MSE) of the cost functions. This study also covers different indoor scenarios, such as general offices, lecture theatres and conference rooms. Given the well trained models, the maximum prediction of potential ESR can be −30% via the sparse AFA optimizations while maintaining indoor thermal comfort in the pre-defined comfort zone.