Performance assessment and optimisation of a novel guideless irregular dew point cooler using artificial intelligence

Air Conditioners (ACs) are a vital need in modern buildings to provide comfortable indoor air for the occupants. Several alternatives for the traditional coolers are introduced to improve the cooling efficiency but among them, Evaporative Coolers (ECs) absorbed more attention owing to their intelligible structure and high efficiency. ECs are categorized into two types, i.e., Direct Evaporative Coolers (DECs) and Indirect Evaporative Coolers (IECs). Continuous endeavours in the improvement of the ECs resulted in development of Dew Point Coolers (DPCs) which enable the supply air to reach the dew point temperature. The main innovation of DPCs relies on invention of a M-cycle Heat and Mass Exchanger (HMX) which contributes towards improvement of the ECs’ efficiency by up to 30%. A state-of-the-art counter flow DPC in which the flat plates in traditional HMXs are replaced by the corrugated plates is called Guideless Irregular DPC (GIDPC). This technology has 30-60% more cooling efficiency compared to the flat plate HMX in traditional DPCs.Owing to the empirical success of the Artificial Intelligence (AI) in different fields and enhanced importance of Machine Learning (ML) models, this study pioneers in developing two ML models using Multiple Polynomial Regression (MPR), and Deep Neural Network (DNN) methods, and three Multi Objective Evolutionary Optimisation (MOEO) models using Genetic Algorithms (GA), Particle Swarm Optimisation (PSO), and a novel bio-inspired algorithm, i.e., Slime Mould Algorithm (SMA), for the performance prediction and optimisation of the GIDPC in all possible operating climates. Furthermore, this study pioneers in developing an explainable and interpretable DNN model for the GIDPC. To this end, a game theory-based SHapley Additive exPlanations (SHAP) method is used to interpret contribution of the operating conditions on performance parameters.The ML models, take the intake air characteristic as well as main operating and design parameters of the HMX as inputs of the ML models to predict the GIDPC’s performance parameters, e.g., cooling capacity, coefficient of performance (COP), thermal efficiencies. The results revealed that both models have high prediction accuracies where MPR performs with a maximum average error of 1.22%. In addition, the Mean Square Error (MSE) of the selected DNN model is only 0.04. The objectives of the MOEO models are to simultaneously maximise the cooling efficiency and minimise the construction cost of the GIDPC by determining the optimum values of the selected decision variables.The performance of the optimised GIDPCs is compared in a deterministic way in which the comparisons are carried out in diverse climates in 2020 and 2050 in which the hourly future weather data are projected using a high-emission scenario defined by Intergovernmental Panel for Climate Change (IPCC). The results revealed that the hourly COP of the optimised systems outperforms the base design. Moreover, although power consumption of all systems increases from 2020 to 2050, owing to more operating hours as a result of global warming, but power savings of up to 72%, 69.49%, 63.24%, and 69.21% in hot summer continental, arid, tropical rainforest and Mediterranean hot summer climates respectively, can be achieved compared to the base system when the systems run optimally