Hybrid solar-gas-electric dryer optimization with genetic algorithms
To promote the hybrid solar dryers for use even under unfavorable weather and to overcome the intermittance state issue, the energy consumption should be optimized and the response time should be reduced. This work concerns a drying chamber connected to a solar absorber where the air can be heated also by combustion of gas and by electric resistance. To optimize the control parameters, an evolutionary optimization algorithm simulating natural selection was used. It was combined with a predictive model based on the artificial neural networks (ANN) technique and used as a fitness function for the genetic algorithm (GA). The ANN is a learning algorithm that needs training through a large dataset, which was collected using CFD simulation and experimental data. Then a GA was executed in order to optimize two objectives: The energy consumption and the t95% response time in which the drying chamber temperature reaches its set point (60°C). After optimization, a 30% decrease of the t95% response time, and 20% decrease of the energy consumption were obtained. Keywords: hybrid solar dryer; artificial neural network; temperature regulation; energy consumption; genetic algorithm.