Parameter Identification of Horizontal Flexible Plate System Using Cuckoo Search Algorithm

Abstract

AbstractThe flexible structure offers various advantages such as lightweight, efficient, fast system response and low energy consumption. However, this structure produces too much vibration which lead to system failure. This thesis presents a modelling of the horizontal flexible plate system based on cuckoo search algorithm. The aim of this research is to eliminate vibrations produced by the flexible plate structure in order to preserve the performance and lengthen the life of the system. Initially, the system models are developed based on the vibration data collected experimentally. The dynamic models were developed based on auto-regressive with exogenous (ARX) using system identification technique. The system identification technique applied in this study to develop the mathematical models of the system are the conventional technique known as recursive least squares (RLS) and swarm intelligence algorithm known as cuckoo search algorithm (CS). Then, the developed system models were evaluated and validated by mean squared error (MSE), correlation test and also pole-zero diagram. The performance of developed model was compared between the conventional and swarm intelligence algorithms. The best system model was verified based on lowest MSE, within 95% confidence level of the correlation test and high stability of the pole-zero diagram.KeywordsFlexible plateSystem identificationRecursive least algorithmCuckoo search algorithmSwarm intelligence algorithm