Research On Active Noise Control Based On Iterative Learning Control

Noise pollution of modern industrial activities is becoming a serious environmental pollution. Traditional passive noise cancellation technology can’t suppress low-frequency noise effectively. And it’s costly. Active noise control(ANC) technology is an effective noise suppression technology for low-frequency noise. It has been widely used in industrial production, daily life, national defense and military fields. ANC system cancels primary noise by generating an anti-noise of equal amplitude and opposite phase. The cancellation performance is influenced by noise tracking precision. A new type of tracking technology called iterative learning control(ILC)is to get national attention in recent years. It can improve the performance of ANC system on noise tracking. ANC system based on ILC has high research value.This thesis establishes a generalized iterative learning control scheme(GILCS) and proposes two design strategies for learning filter of GILCS. In the designs, we use an IIR filter to realize learning filter, then use Particle swarm optimization(PSO) and Quantum-behaved particle swarm optimization(QPSO) to optimize filter parameters. The research focuses on three questions: how to combine the two techniques; how to establish and demonstrate the system model; how to optimize the system parameters to enhance performance. In this paper, research content and results are as follows:1. Establish an active noise control system based on iterative learning control, derive convergence condition, convergence error and convergence rate formula of the system.2. Propose a design strategy for learning filter of GILCS based on PSO. The learning filter is designed to ensure that the system can converge to a tiny noise and the convergence rate reaches a maximum value at the same time.3. Enhance the performance of design strategy based on QPSO. The enhanced strategy can search the system parameters more effectively.