Inverse Heat Condution Analysis of Composite Layers with Genetic Algorithm

• Main Author: 盧嘉聰
• Other Authors: 洪振益
• Format: Others
• Language: zh-TW
• Published: 1997
• Online Access: http://ndltd.ncl.edu.tw/handle/02282871386922422405

碩士 === 國立成功大學 === 機械工程學系 === 85 === In this present dissertation, a methodology is used to solve one-dimensional transient inverse heat conduction problem (IHCP) with composite layers on the estimation of the unknown boundary conditions or the contact heat conductance. This methodology is the combination of Genetic Algorithms (GAs) with the finite difference method and the relatively least square scheme. The advantage of GAs is that the mathematical model is simple and the solution is global optimum and robust. This thesis presents three examples of transient one-dimensional IHCP for composite layers. Example one is the estimation of the "prefect contact" plate surface temperature with known function form, example two is the estimation of the "prefect contact" cylinder surface temperature with known function form, and example three is the estimation of the contact heat conductance with known function form of the "imprefect contact" cylinder. We will discuss the influences of the measurement error, and the numbers of measurement data on the predicted values. The results reveal that the better solution is obtained for the maximum measurement error is under 5% and the numbers of measurement data are about 30 to 40.