Estimation of Unknown Inlet Temperature Profile Using an Improved Gbest-PSO

In this study, an improved gbest-PSO is proposed to overcome the shortcoming of earlier convergence of classical gbest-PSO. Then the improved gbest-PSO is used to identify the unknown inlet temperature profile in a plate channel flow. The effects of measurement position and measurement error on the accuracy of prediction are studied thoroughly. Analysis of computational results of two test problems shows that the improved gbest-PSO proposed in this paper has an excellent smooth convergence characteristic. The local refine mechanism introduced in the improved gbest-PSO increases the opportunity of finding the global optimum greatly especially for high dimensional multimodal optimization problems. Accurate results are obtained even when the measurements contain a 10% noise. Consequently, the inverse convection heat transfer problem is successfully solved by the improved gbest-PSO.