Volume 11,Issue 01

Optimization of Operating Conditions of Beta-Type Stirling Engine with Regenerator Using Artificial Neural Network and Response Surface Method

Authors

Ece Ozgoren Unlu , Berna Yazici, Yasar Onder Ozgoren

Abstract
The aim of this study is to determine the optimal operating conditions of a beta-type Stirling engine and the main factors affecting these conditions. The main factors affecting the engine power are engine speed (750-900 rpm), pressure (6-8 bar) and temperature (600-700 °C). Experimental design was conducted applying Box-Behnken Design, to obtain engine power values via selected factors. In addition, these experiments were carried out on a beta-type Stirling engine with a regenerator, which has not been previously studied in the literature. The engine power was analyzed using the Response Surface Method and Artificial Neural Network, and the most appropriate model was achieved. The desirability function approach was used to determine the optimal engine operating conditions and the optimum values of the main factors of pressure, engine speed and temperature were determined to be 8 bar, 900 rpm and 650 °C, respectively. Under optimal engine operating conditions, the engine power was determined to be 56.736 W. Besides, the Determination Coefficient (R2) and Mean Square Error values for the Response Surface Method were 0.898 and 6.47, respectively, while for the Artificial Neural Networks method, they were 0.975 and 2.11, respectively. The results obtained indicate that the developed Artificial Neural Network model is an acceptable and more powerful modeling technique than the Response Surface Method for predicting power values of the beta-type Stirling engine.

Keyword: Optimization, Artificial Neural Network, Response Surface Method, Box-Behnken Method, Stirling Engine.

PDF [ 769.5 Kb ] | Endnote File | XML