本文選題：Volterra濾波器 + 非線性主動噪聲控制��； 參考：《西南交通大學》2015年碩士論文

【摘要】：在非線性信號處理領域中一個重要的研究對象是非線性主動噪聲控制,由Volterra濾波器本身包含的非線性特性,適合創(chuàng)建非線性系統(tǒng)模型,能夠有效的實現(xiàn)非線性主動噪聲控制。因此,基于Volterra濾波器的非線性主動噪聲控制就成為自適應信號處理領域中備受關注的熱點。論文在選題的背景和意義及基于Volterra濾波器的非線性主動噪聲控制國內外研究現(xiàn)狀的基礎上,對濾波器結構進行深入研究,來實現(xiàn)低計算復雜度的非線性主動噪聲控制效果。利用Volterra濾波器和DCT技術兩個方面,探究一些新型結構的非線性主動噪聲控制器。主要是在標準的二階Volterra濾波器和三階Volterra濾波器結構、擴展流水線Volterra濾波器結構基礎之上,適當?shù)膽肈CT技術來改進繼而提出相應的新型結構的非線性主動噪聲控制器。本文又對所提出的DCT域二階Volterra濾波器結構的主動噪聲控制中應用的FXLMS算法進行改變,采用NLMS算法和變步長算法,目的是為進一步改善系統(tǒng)的性能,計算機仿真證明了其可行性、有效性及優(yōu)越性。論文的主要研究成果包括：1研究了兩種DCT域Volterra濾波器的非線性主動噪聲控制結構,即基于DCT域二階和三階Volterra濾波器的非線性主動噪聲控制結構,此主動噪聲控制器最大特點是結構簡單,實現(xiàn)容易,能夠減少嚴重的計算量負擔和提高濾波器算法的收斂速度。2在提出的基于DCT域二階Volterra濾波器的非線性主動噪聲控制結構基礎之上,對其應用的FXLMS算法和算法中固定步長的不足進行改進,采用NLMS算法和變步長算法來改變系統(tǒng)權系數(shù)的更新方式。計算機仿真結果表明,NLMS算法和變步長算法不僅在新的結構中有效可行,而且保證系統(tǒng)低計算復雜度的情況下,提高了收斂速度和保證了較低的穩(wěn)態(tài)誤差。3基于擴展流水線Volterra濾波器結構的基礎之上應用DCT技術,提出基于DCT域擴展流水線Volterra濾波器非線性主動噪聲控制結構及改進結構。實現(xiàn)了系統(tǒng)性能的改善,更重要的是進一步達到了非線性主動噪聲控制的低計算復雜度需求。
[Abstract]:Nonlinear active noise control is an important research object in the field of nonlinear signal processing. The nonlinear characteristics contained in Volterra filter are suitable for the establishment of nonlinear system model and can effectively realize nonlinear active noise control. Therefore, nonlinear active noise control based on Volterra filter has become a hot topic in the field of adaptive signal processing. Based on the background and significance of the topic and the research status of nonlinear active noise control based on Volterra filter at home and abroad, the structure of the filter is deeply studied to achieve the low computational complexity of nonlinear active noise control effect. Using Volterra filter and DCT technology, some new nonlinear active noise controllers are studied. Based on the standard second-order Volterra filter and third-order Volterra filter structure and extended pipelined Volterra filter structure, a new nonlinear active noise controller is proposed by using DCT technology. In this paper, the FXLMS algorithm used in the active noise control of the second-order Volterra filter structure in DCT domain is changed, and the NLMS algorithm and the variable step size algorithm are adopted in order to further improve the performance of the system. The computer simulation proves its feasibility. Effectiveness and superiority. The main research results of this paper include: 1, the nonlinear active noise control structure of two kinds of Volterra filters in DCT domain is studied, that is, the nonlinear active noise control structure based on DCT second-order and third-order Volterra filters. This active noise controller is characterized by its simple structure and easy implementation. It can reduce the heavy computational burden and improve the convergence speed of the filter algorithm .2 based on the proposed nonlinear active noise control structure based on the second-order Volterra filter in DCT domain. In order to improve the FXLMS algorithm and the deficiency of the fixed step size, the NLMS algorithm and the variable step size algorithm are used to change the updating method of the system weight coefficient. The computer simulation results show that the NLMS algorithm and the variable step size algorithm are not only effective and feasible in the new structure, but also ensure the low computational complexity of the system. The convergence rate is improved and the steady-state error is guaranteed. 3 based on the extended pipelined Volterra filter structure, the nonlinear active noise control structure and the improved structure of the extended pipeline Volterra filter based on DCT domain are proposed by using DCT technology. The performance of the system is improved and, more importantly, the low computational complexity of nonlinear active noise control is achieved.
【學位授予單位】：西南交通大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TB535;TN713

【參考文獻】

相關期刊論文 前2條

1 虞曉,胡光銳,吳小滔;自適應噪聲對消中的 ELMS 算法及其變步長算法[J];上海交通大學學報;1998年04期

2 張家樹,李恒超,肖先賜;連續(xù)混沌信號的離散余弦變換域二次實時濾波預測[J];物理學報;2004年03期

，

本文編號：1923559