| Summary: | Being reliant on Air Conditioning (AC) throughout the majority of the year, desert countries with extremely hot weather conditions such as Qatar are facing challenges in lowering weariness cost due to AC On-Off switching while maintaining an adequate level of comfort under a wide-range of ambient temperature variations. To address these challenges, this paper investigates an optimal On-Off control strategy to improve the AC utilization process. To overcome complexities of online optimization, a Elman Neural Networks (NN)-based estimator is proposed to estimate real values of the outdoor temperature, and make off-line optimization available. By looking up the optimum values solved from an off-line optimization scheme, the proposed control solutions can adaptively regulate the indoor temperature regardless of outdoor temperature variations. In addition, a cost function of multiple objectives, which consider both Coefficient of Performance (COP), and AC compressor weariness due to On-Off switching, is designed for the optimization target of minimum cost. Unlike conventional On-Off control methodologies, the proposed On-Off control technique can respond adaptively to match large-range (up to 20 <inline-formula> <math display="inline"> <semantics> <msup> <mrow></mrow> <mo>∘</mo> </msup> </semantics> </math> </inline-formula>C) ambient temperature variations while overcoming the drawbacks of long-time online optimization due to heavy computational load. Finally, the Elman NN based outdoor temperature estimator is validated with an acceptable accuracy and various validations for AC control optimization under Qatar’s real outdoor temperature conditions, which include three hot seasons, are conducted and analyzed. The results demonstrate the effectiveness and robustness of the proposed optimal On-Off control solution.
|
|---|
