【作者】 刘月欣；

【导师】 徐新盛；

【作者基本信息】 哈尔滨工程大学 ， 水声工程， 2008， 硕士

【摘要】 随着现代工业的发展,噪声控制问题日益引起了人们的关注。基于自适应控制策略的有源噪声抵消技术已经成为有源噪声控制领域的重要研究内容之一。在自适应有源噪声控制中,控制算法是直接影响自适应控制品质的重要因素。目前,国内外在控制算法的研究方面已经取得了丰硕的成果,但其中一些算法存在着明显的不足和特定性,因此运用新的信息分析手段对控制算法进行研究是一个非常有潜力而且具有重要意义的工作。针对这一问题,论文选择丁以控制算法为切入点,主要研究了两类自适应有源噪声抵消方法:基于神经网络的自适应有源噪声抵消方法和基于同步方程方法的自适应有源噪声抵消方法。为了有效的控制非线性噪声,论文深入研究了FXBPNN算法,该算法采用BP(back-propagation)神经网络作为自适应有源噪声抵消系统的控制器,它能够抵消非线性噪声,但运算量大。因此在这种算法的基础上,论文提出了一种运算量小的FEBPNN算法。并对FEBPNN算法、FXBPNN算法和FXLMS算法进行了计算机仿真比较分析。另外,论文深入研究了基于同步方程方法的自适应有源噪声抵消算法,该算法通过估计附加滤波器避免了次级通道建模,但运算量也很大。因此在这种算法的基础上,论文提出了一种改进的基于同步方程方法的自适应有源噪声抵消算法,通过存储少量的主输入噪声和误差信号,避免了估计附加滤波器,因而有效地降低了运算量。通过计算机仿真,将两种同步方程方法与FXLMS算法进行比较分析。最后,利用文中研究的几种自适应有源噪声抵消算法对真实的海试数据进行处理分析。更多还原

【Abstract】 With modern industrial development, the subject of noise control has attracted people’s growing attention and concern. Adaptive active noise cancellation based on the adaptive control strategy has become one of important researches in the field of active noise control. Control algorithm is an important factor of directly affecting the adaptive control quality. So far, a lot of great achievements have been reached in the study of control algorithm all over the world. However, there are obvious deficiency and specificity in many of these algorithms. Therefore, it is a very potential and meaningful work to improve the performance of control algorithm by means of new information analytical tools.Aimed at this point, the control algorithm is chosen to be studied in this paper. Two kinds of adapative active noise cancellation algorithms which are based on neural network and simultaneous equations are discussed in this paper.To control nonlinear noise effectively, FXBPNN algorithm is studied. This algorithm using BP (back propagation) neural network as a controller of an adapative active noise cancellation system can cancel nonlinear noise. But it is has large amount of computation. Therefore, on the basis of this algorithm, FEBPNN algorithm with lower computational load is proposed in this paper.Then, Computer simulations are carry out to compare the FEBPNN algorithm with FXBPNN algorithm and FXLMS algorithm.In addition, an adaptive active noise cancellation algorithm based on the simultaneous equations method is introduced in this paper. This algorithm can avoid secondary path modeling by system identification of an auxiliary filter. As a result, the amount of computation is increased. In this paper, an improved adaptive active noise cancellation algorithm based on the simultaneous equations method is proposed. By storing a small number of an input signal and an error signal, it can avoid this identification. Therefore, the amount of computation can be reduced greatly. Computer simulations are carry out to compare the two algorithms with FXLMS algorithm.At last, the performance of several adaptive active noise cancellation algorithm studied in the paper is proved through analysis and trealing with the sea trial data.更多还原