Active intelligent control of vibration of flexible plate structures

The development of intelligent control approaches for vibration reduction of flexible plate structures are investigated and reported in this thesis. In this work active intelligent control comprises a set of control techniques based on particle swarm optimisation (PSO), real coded genetic algorithm (RCGA) and artificial immune system (AIS). A traditional method of recursive least squares ( U S ) is investigated as a comparison to the proposed intelligent techniques. The aim of this work is to assess the potential applicability of the intelligent techniques in the active vibration control (AVC) of flexible structures. A simulation environment characterising the dynamic behaviour of a flexible plate structure is developed using finite difference methods and state-space formulation. This is realised within MatlabISIMULINK as a testbed for verification of control designs. The plate is subjected to different disturbance signals and linear parametric models characterising the inputloutput dynamic behaviour of the plate, between two measurement points, is developed using RLS, RCGA, PSO and AIS algorithms. It is demonstrated through time-domain and frequency-domain analysis and tests that the RCGA, PSO and AIS approaches perform very well in modelling of the flexible plate. The modelling approach is extended to a model-based AVC strategy using the principle of wave interference. The approach is first realised within a single-input single-output (SISO) control configuration with RLS, RCGA, PSO and AIS algorithms. Tests with various disturbance signals show that good vibration reduction is achieved with the developed model-based SISO-AVC algorithms. The approach is then realised within a singleinput multi-output (SIMO) control configuration, and exemplified in vibration control tests with two control sources. It is shown that higher levels of vibration can be achieved with model-based SIMO-AVC algorithms as compared to those modelbased SISO-AVC algorithms. A further strategy based on non-model based control is developed for vibration reduction in flexible structures. The approach is realised within SISO and SIMO AVC structures using RCGA, PSO and AIS algorithms. These are implemented on the flexible plate structure with various disturbance signals. It is demonstrated that the non-model based AVC algorithms perform comparatively similar to their model-based AVC counterparts in terms of amount of vibration reduction. However, the non-model based AVC algorithms are faster than their model-based AVC counterparts.