Active vibration control of flexible beam system based on cuckoo search algorithm

Abstract

<p>A flexible beam is recognized as a lightweight structure that is prone to excessive vibration, resulting in poor performance. Thus, controlling unwanted vibration is necessary to maintain the system’s performance. Therefore, this study presents a technique to suppress undesired vibration in a flexible beam structure by introducing active vibration control (AVC). However, to develop an effective controller, an appropriate flexible beam model must first be obtained. In recent times, one of the best methods employed to model a flexible beam structure is system identification via a swarm intelligence algorithm. In this study, an intelligent algorithm acknowledged as cuckoo search (CS) was acquainted. The capability of the proposed algorithm was verified using three robustness techniques which were correlation test, pole-zero diagrams and mean square error (MSE). The simulation result showed that the CS algorithm achieved superior performance by achieving the lowest MSE of 6.1547x10<sup>-9</sup>, a correlation test between a 95% confidence level and high stability. Next, a proportional-integral-derivative (PID) controller tuned by the Ziegler-Nichols method was developed using the transfer function accomplished from the CS model. Two types of interference, namely single and multiple sine waves were introduced to validate the effectiveness of the controller. The controller successfully achieved a 30.2 dB of attenuation level for both disturbances.</p>

Mayfly Algorithm for Modelling a Horizontal Flexible Plate Structure

“…Finally, the discrete transfer function resulting from the optimal RLS model is defined in Eq. (11).…”

“…The parameter estimation to construct mathematical model can be acquired from traditional and optimization approaches. Nonetheless, in situations where obtaining a model structure using traditional techniques proves challenging, intelligent methods become a desirable option [11][12][13]. Various forms of artificial intelligence, including glowworm swarm optimization (GSO), neural network systems, and mayfly algorithm (MA), have proven effective in parameter identification [8].…”