Study On Feedback Design For Active Control Of Transformer Noise SPECIALIZATION:Signal And Information Processing

Based on the feedback behavior, a feedback design method is proposed and is applied to active control for transformer noise. The performance of decentralized multi-channel feedback analog control systems is analyzed. From the application point, a method of using an internally synthesized signal of preset frequencies as the reference signal for transformer noise control is proposed. The behavior of the hybrid active noise control system is explored and the performances of various control methods are compared.Firstly, a simple and effective approach for designing feedback controllers is proposed based on analyzing the waterbed effect of the feedback active noise control system. By flattening the noise amplification due to the waterbed effect in the whole disturbance amplification frequency band intuitively, the designed feedback controller can suppress the disturbance amplification to almost the minimum value for the given disturbance attenuation and ensure good stability of the system. Being compared with the existing approaches, the main advantage of the proposed approach is its simplicity for implementation, which avoids complicated weight function selections and the numerical optimizations of the existing approaches. Detailed design procedures and implementation of the digital controller are illustrated, simulation and experiment results are provided to show the effectiveness of the proposed approach.Secondly, the performance of decentralized multi-channel feedback analog control systems is analyzed with the performance and the geometrical configuration of the independent controllers. The generalized Nyquist theorem and Gerschgorin circle theorem are used to derive a sufficient stability condition in terms of the predefined maximum noise amplification and the geometrical configuration of the independent controllers, and the noise reduction performance of the multi-channel system is predicted with the design and geometrical configuration of the independent controllers. Simulation and experimental results are presented to illustrate the effectiveness of the proposed analysis.A method of using an internally synthesized signal of preset frequencies as the reference signal for transformer noise control is proposed. The feasibility of applying the method in a specific transformer algorithm is investigated, where the effects of frequency mismatch and secondary path delay on the steady state behavior of the transformer algorithm are analyzed and compared with the filtered-x least mean square algorithm. It is found that the noise reduction performance degrades with the increase of secondary path delay and frequency mismatch, but can be improved by using a larger convergence step size and a faster update rate. The feasibility of applying the transformer algorithm with an internally synthesized reference signal to practical transformer noise control is verified by both numerical simulations and experiments.Finally, the behavior of the hybrid system using a broadband FIR feedback controller and an adaptive feedforward controller with the FXLMS algorithm is studied. It is found the plant together with the feedback controller can be seen as the equivalent secondary path controlled by the feedforward controller and the performance of the hybrid system is possible to be greater than that of the feedback controller and the feedforward controller. Through the simulation and experiment of active headset for broadband white noise, the performances of the hybrid system with different feedback control design and different primary source position are explored. The performances of various methods are compared with active control for the transformer noise.