Soft computing based optimization of a novel solar heliostat integrated energy system using artificial neural networks

Abstract

This study proposes and investigates a novel energy system based on biomass and solar energy. This plant is composed of a biomass unit, a solar unit, and a waste-heat recovery unit. This novel proposed integrated system can provide the needs such as electricity, hydrogen, freshwater, heating, and hot water production. For electricity generation, two gas turbines, one steam Rankine cycle, and one organic Rankine cycle are used. In contrast, for utilization of solar energy, a heliostat field, and for biomass conversion, a gasifier is used. In addition, the desalination unit and PEM electrolyzer are utilized to produce fresh water and hydrogen, respectively. Firstly, the present work aims to investigate the developed system from the exergoeconomic and environmental perspective. Multi-objective optimization is conducted to determine the maximum amount of exergetic efficiency and the minimum value of the cost rate. An artificial neural network (ANN) is employed as a mediator tool to accelerate the optimization process. The relation between objective functions and design parameters is studied utilizing ANN to obtain the plant optimal decision variables. Employing the Pareto Envelope-based selection algorithm II (PESA-II) method, the optimum amount for the total cost rate and exergy efficiency is found 224.1 $/h and 26.7%, respectively. In addition, three evolutionary-based optimization algorithms are applied to determine the optimum results of the suggested plant.