Regression-Assisted Ant Lion Optimisation of a Low-Grade-Heat Adsorption Chiller: A Decision-Support Technology for Sustainable Cooling

Growing cooling demand and environmental concerns motivate research into alternative technologies capable of converting low-grade heat into useful cooling. This study proposes a regression-assisted multi-objective optimisation framework using the Ant Lion Optimiser and its multi-objective variant to jointly maximise the coefficient of performance (COP), cooling capacity (Qcc) and waste-heat recovery efficiency (ηe). Pareto-optimal solutions exhibit a one-dimensional ridge in which ηe declines, and COP and Qcc increase simultaneously. Within the explored bounds, non-dominated ranges span COP = 0.674–0.716, Qcc= 18.3–27.5 kW and ηe= 0.118–0.127, with a practical compromise near COP ≈ 0.695, Qcc ≈ 24 kW and ηe ≈ 0.122–0.123. Compared to the typical reported COP band for single-stage silica-gel/water ADCs, the practical compromise solution (COP ≈ 0.695) offers a conservative COP improvement of approximately 16% when benchmarked against COP = 0.6, while the compromise Qcc (Qcc ≈ 24 kW) represents a conservative increase of approximately 20% relative to the upper product-class reference (20 kW). A one-at-a-time sensitivity analysis with re-optimisation identifies the hot- and chilled-water inlet temperatures and exchanger conductance as the dominant decision variables and maps diminishing-return regions. This framework can effectively utilise low-grade heat in future low-carbon buildings and processes, supporting the configuration of ADC systems.