應用於汽車車艙之多通道主動式噪音控制系統

Abstract

本研究探討了應用於汽車車艙之多通道主動式噪音控制系統。在第一部份，我們將三種控制架構以及兩種控制器演算法提出來作比較。我們使用合成參考輸入之最小均方法 (least-mean-square, LMS) 以及線性二次高斯法（linear quadratic gaussian, LQG）來合成控制器，同時比較回饋式（feedback），前饋式（feedforward）以及回饋前饋混合式（hybrid）控制架構對噪音控制效果之影響。然而在眾多主動控制問題中，倘若效能感測器與量測感測器置於同一位置或是控制喇叭與噪音源位置相同，則由於波德積分的限制（Bode’s integral constraint）會有溢出（spillover）的情況產生。因此在第二部分，我們利用空間回饋（spatially feedforward）架構和線性二次高斯法，設計一多通道之零溢出控制器。在第一和第二部分中皆使用頻率域法建立需要之數學模型，同時本研究所提出的控制器均以一數位訊號處理器（DSP）來加以實現。在第一部份，我們以一汽車內部空間作為實驗聲場，並僅使用汽車噪音。實驗結果顯示回饋與前饋混合型架構對於消音有最好的效果與收斂速度。在第二部分，我們使用與第一部份相同之聲場，但使用寬頻隨機噪音以及引擎噪音當作主要噪音源。實驗結果指出所設計出來之控制器對於寬頻隨機噪音有抑制的效能；而當我們以假想的噪音模型設計控制器時，所設計的控制器亦能有效地抑制汽車噪音。

This paper proposes a multiple channel active noise cancellation system for car cabins. In part I, various active techniques are compared in terms of three control structures and two control algorithms. The least-mean-square (LMS) algorithm with synthetic reference and the linear quadratic gaussian (LQG) control are employed for controller synthesis. Feedback, feedforward and hybrid configurations are used as the major control structures. In many active noise control problems, however, spillover will occur because of the Bode integral constraint if the performance sensors and measurement sensors are collocated, or if the control speakers and the noise sources are collocated. In part II, therefore, we will focus on the implementation of a multiple-channel zero-spillover ANC controller using a spatially feedforward structure for a car cabin. The linear quadratic gaussian algorithm is used in controller design. Both the plant models in part I and II are realized by using a frequency-domain procedure. The controllers are implemented by using a digital signal processor (DSP). In part I, experiments are carried out to evaluate the proposed approaches for attenuation of the internal field in a car cabin, where only the engine noise is used. The experimental results indicate that the hybrid structure yields the best performance in terms of attenuation and convergence speed. In part II, experiments are carried out to verify the proposed technique, where broadband random noise and engine noise are chosen as the primary noise. The results indicate that the proposed controller yields a broadband attenuation for the random noise. The proposed method is also effective in attenuating the engine noise if an assumed model of the noise is incorporated into controller design.