Model Optimization of a Shell and Tube Heat Exchanger by Using HGAPSO

Abstract

The demand to develop an efficient heat exchanger have been prompted in order to save the energy and materials as well as economic incentives within the industry. To establish a reliable heat exchanger that be able to operate under its normal operating conditions, a model should reflect the true behaviour of the process itself. Modelling of heat exchanger helps to understand process that occurs during the operation. Hence, the primary objective of the project is developing a structural model using an ARX (Auto-Regressive with eXogenous input) equation. The data from experiment was used to determine the parameter of ARX equation. Using HGAPSO (Hybrid Genetic Algorithm and Partical Swarm Optimization) algorithm, ARX parameters are optimized to obtain the transfer function that represents the plant for modelling. Validation test of autocorrelation and cross-correlation were used to validate between normalised data input and error. Based on the result on the six order model, parameter a0, a1, a2, a3, a4, and a5 values are -0.00083, -0.00202, -0.00253, -0.00095, -0.00213 and -0.00275 while parameter b0, b1, b2, b3, b4 and b5 values are -0.7642, -0.5759, 0.3118, 0.0715, 0.1140 and 0.0683 respectively. From validation test, all graphs are within the 95 percent of confident line.