Developing a machine learning model for fast economic optimization of solar power plants using the hybrid method of firefly and genetic algorithms, case study: optimizing solar thermal collector in Calgary, Alberta

Abstract

Due to the depletion of fossil fuels and environmental concerns, renewable energy has become increasingly popular. Even so, the economic competitiveness and cost of energy in renewable systems remain a challenge. Optimization of renewable energy systems from an economic standpoint is important not only from the point of view of researchers but also industry owners, stakeholders, and governments. Solar collectors are one of the most optimized and developed renewable energy systems. However, due to the high degree of nonlinearity and many unknowns associated with these systems, optimizing them is an extremely time-consuming and expensive process. This study presents an economically optimal design platform for solar power plants with a fast response time using machine learning techniques. Compared with traditional mathematical optimization, the speed of economic optimization with the help of the machine learning method increased by up to 1100 times. A total of seven continuous variables and three discrete variables were selected for optimization of the parabolic trough solar collector. The objective functions were to optimize the exergy efficiency and the heat cost. As part of the environmental assessment, the cost of carbon dioxide emission was calculated based on the system's exergy and energy efficiencies. According to the sensitivity analysis, the mass flow of working fluid and the initial temperature of the fluid play the most significant roles. A simulated solar collector in Calgary was optimized in order to evaluate the applicability of the proposed platform.

Keywords

• machine learning model
• firefly optimization algorithm
• solar collector