Research On Particle Swarm Optimization Based Active Noise Control Method

As an urgently environmental problem, noise pollution seriously affects people’s life and production. It has become a prominent issue constraining on socioeconomic development. Comparing with the conventional noise control technology, active noise control technology has been widely used for its merits in low-frequency noise control. Secondary path estimation is required in the classic active noise control method, which not only increases complexity but also affects stability of the active noise control system. Besides, particle swarm optimization algorithm based active noise control method avoids the secondary path estimation, but its global convergence speed and optimization accuracy remains to be further improved.To improve the performance of active noise control system based on particle swarm optimization algorithm, an improved particle swarm optimization algorithm is proposed. What’s more, research on improved particle swarm optimization algorithm based active noise control method is mainly carried out in this paper. Firstly, theoretical basis of active noise control technology is introduced, including acoustic theory, adaptive filter theory and active noise control system theory. Combining with the actual situation, the design of active noise control system is provided.Secondly, the classic active noise control algorithm is discussed, and the reason of requiring secondary path estimation in advance is analyz ed. According to the basic principle and convergence condition of classic algorithm, several improved classic active noise control algorithms are given. The corresponding simulation results verify the theoretical analysis of classic active noise control algorithm and also show drawbacks in its strain capacity.Then, an improved particle swarm optimization algorithm is proposed according to the relative performance of the basic particle swarm optimization algorithm. Three improvements are made in this algorithm, which include dynamically changing the inertia weight, substituting the average of all individual optimal values for personal best value, and setting the trigger condition to deal with the mutations of acoustic path. Based on the improved particle swarm optimization algorithm, a new active noise control method is designed. Its main steps and differences from classic active noise control method are also analyzed. Comparing the simulation results, it is illustrated that this method has faster convergence speed and higher optimization accuracy than the existing particle swarm optimization algorithm based active noise control method, and it also has stronger strain capacity than the classic active noise control method and the existing particle swarm optimization algorithm based active noise control method.Finally, the experimental platform using audio interface and other related equipment is set up. The time-frequency domain analysis of the primary noise is completed and the corresponding experimental scheme is also designed. In the experiment of secondary path modeling, an accurate model is estimated by the least mean square algorithm. In the cancellation experiment, the result shows that the improved particle swarm optimization algorithm and classic active noise control algorithm can both perform well in noise reduction, and the former performs better. In the noise reduction experiments, two noise reduction systems are designed and developed by using the improved particle swarm optimization algorithm and the classic active noise control algorithm, the results show that both systems can significantly attenuate the primary noise and have a certain versatility.