Using GMDH Neural Networks to Model the Power and Torque of Stirling Engine

The Stirling engine is a simple type of external-combustion engine that uses a compressible fluid as a working fluid. The Stirling engine can theoretically be very efficient to convert heat into mechanical work at Carnot efficiency. It is an environmental friendly heat engine which could reduce CO2 emission through combustion process. Various parameters could affect the performance of the addressed Stirling engine which is considered in its optimization for designing purpose. Through addressed factors, torque and power have the highest effect on the robustness of the Stirling engines. Due to this fact, determination of the two referred parameters with low uncertainty and high precision are needed. In this communication, the distribution of torque and power are represented based on experimental evidence. A new polynomial model is suggested to calculate torque and power, based on experimental data. This study addresses the question of whether GMDH-type neural networks could be used to estimate the torque and power based on specified variables.