Thermal systems include devices or processes where heat transfer plays a major role, such as heat exchangers, refrigeration units, combustion engines, solar collectors, cooling systems in electronics, etc. These systems often have multiple performance objectives like minimizing energy consumption, maximizing heat transfer, reducing costs, improving efficiency and reliability, etc. This paper presents a simple yet powerful optimization algorithm—Best-Current-Random (BCR)—developed to solve both unconstrained and constrained optimization problems in thermal systems involving single- and multiple objectives. Notably, this algorithm operates without the use of metaphor-based analogies and eliminates the need for algorithm-specific parameters, which are typically required in traditional metaheuristic methods and often introduce additional complexity. The BCR algorithm operates by leveraging the best, current, and randomly selected solutions from the current population. Comprehensive evaluations reveal that this novel and simple algorithm delivers competitive—and frequently superior—performance compared to conventional optimization techniques. The BCR algorithm is applied to the single-objective optimization of heat exchanger networks and an industrial refrigeration system. The multi-objective version of the BCR is named MO-BCR and is successfully applied to solve a 2-objectives problem and a 4-objectives problem of selected thermal systems. The proposed BHARAT (Best Holistic Adaptable Ranking of Attributes Technique) can be used to find out the best compromise nondominated solution from amongst a number of Pareto-optimal solutions. Researchers and practitioners across scientific and engineering domains may find these algorithms advantageous in solving real-world, constrained, and non-convex optimization problems.
Rao et al. (Thu,) studied this question.