Gray Wolf Optimizer for the Cogeneration Power Plant Design Within a Framework of Commercial Software and Mathematical Platform

Abstract

In a realm where finite natural fossil fuel reservoirs coexist with escalating energy requisites and critical ecological contamination thresholds, matters pertaining to the configuration of thermal systems, encompassing energy efficacy, financial assessment, project intricacy, ecological consciousness, and fine-tuned optimization, have progressively piqued the scientific community's curiosity. Hence, thermoeconomic optimization emerges as a promising avenue for enhancing the efficiency of thermal system designs. Nevertheless, the intricacies of thermoeconomic optimization in thermal system design typically involve a multitude of components, interconnected processes, and flows, which collectively give rise to a complex system of nonlinear equations stemming from both thermodynamic and economic modeling. Moreover, the inherent objective functions in these optimization challenges are analytically daunting, characterized by traits like discontinuity, multimodality, and non-differentiability, further compounded by a multitude of decision variables. In this context, metaheuristic methods present themselves as promising and appealing tools for optimizing such intricate systems. In this study, we employ two metaheuristic methods, namely the Genetic Algorithm (GA) and the Gray Wolf Optimizer (GWO), to optimize the regenerative gas turbine cogeneration system, recognized in the literature as the CGAM problem. The thermoeconomic optimization challenge is tackled and resolved through the computational integration of a commercial software package (EES) and a mathematical platform (Matlab). Within this framework, the thermodynamic and economic modeling, as well as the thermoeconomic optimization components, are seamlessly integrated into the Engineering Equation Solver (EES). EES, in turn, calculates the thermodynamic properties for all streams within the cogeneration system while concurrently solving mass and energy balances as necessitated by the evaluation of the objective function. It is worth noting that the GA operates as an optimization tool within EES, whereas the GWO is implemented in Matlab and effectively integrated with EES. This study reveals that, despite GWO's relatively longer computational time attributable to the integration between Matlab and EES, it stands out as notably efficient in addressing the given problem, primarily owing to its reduced demand for objective function evaluations during the optimization process. Moreover, both the decision variables and the objective function tend to converge towards values closely aligned with those found in the reference literature.