Adaptive control of vibration and acoustic waves in structural systems via singular value decomposition method

In engineering systems, it is possible for sound or vibration sources to generate elastic waves that can have disruptive or destructive effects. One of the important approaches utilized to prevent these situations is called active control. The concept of active control, in its most general definition, is the process of minimizing the amplitudes of elastic waves that occur in solids, liquids or gases due to sound or vibration, by applying other waves with altered amplitude, phase or frequency to the same environment. In order to carry out the aforementioned operations, the magnitudes of waves to be controlled must be measured within certain boundaries and artificial waves adjusted according to measured magnitudes must be reintroduced to the system as feedback or feedforward signals. In this thesis, a layout including vibration control sources and acoustic sensors is chosen and a workflow consisting of three stages has been planned for noise control. In the first part, the modal coupling governing equations between structural element and the acoustic field were obtained and the error criterion of the model was determined by applying the appropriate boundary conditions. In the second part, system parameters were optimized using single value decomposition for the specified error criterion, then verified with the finite element model results, and finally obtained data provided the configuration of the active control systems. In the third and last part experiments were carried out. The metal elements were used in the experiment were exposed to artificially created noise levels in the control environment. The sound levels occurring in the enclosed space due to the vibrating structure were measured with the calibration microphones, and the adapted waves were fed back to the system. As a result of detailed analysis and assessment of the findings, it was determined that the optimization of the control system parameters, the finite element model and the experimental data were compatible with each other, and it was observed that the sound pressure levels were reduced by 17 – 20 dB in the established active noise control system.