A Trained Neural Network for the Prediction of Optimum Operation Mode of a Stirling Engine

In this paper, a prototype Stirling engine is fabricated and tested. In order to find the optimum operation mode of the engine, a neural network is trained and developed using the experimental data, which are obtained from the test of the engine. The trained neural network can predict the thermal efficiency of the Stirling engine for input parameters such as angular speed and gas chamber temperature. The neural network results are validated by the experimental data. Finally, the optimization of the thermal efficiency of the Stirling engine is carried out using the genetic algorithm. The results show that the neural network is a good tool for the prediction of the thermal efficiency of the Stirling engine. The optimum value of angular speed and gas chamber temperature obtained is 240.5 rpm and 489.5 K, respectively. Also, the maximum thermal efficiency of the Stirling engine is 31.05%.