【作者】 杨鹏；

【导师】 应黎明；

【作者基本信息】 武汉大学 ， 电力系统及其自动化， 2017， 硕士

【摘要】 随着中国经济飞速发展,人们越来越关注生活的品质。降低变电站噪声,可以减轻噪声对周围群众生活的干扰,大幅提升其生活质量。有源噪声控制系统能有针对性的降低变电站低频噪声,因此,简单、经济、有效的有源降噪系统在电力变压器噪声控制领域有着广阔应用前景。目前,大部分研究以线性系统为基础,该系统收敛速度与稳态误差之间存在不可调和的矛盾,而且在实际运行中,控制系统中不可避免的存在非线性因素,使这类线性系统的控制性能大打折扣。如果能解决系统中的非线性问题,有源降噪系统就能够选用含有非线性失真的低价电声器件,不仅能提高降噪性能,对降低系统的成本同样具有特殊意义。因此,论文以国家自然科学基金项目《基于内模控制的阵列式电力变压器有源消声技术研究》为依托,对电力变压器非线性有源噪声控制技术及系统收敛速度与稳态误差的协调方法进行了研究。论文针对系统收敛速度与稳态误差之间的矛盾,将凸组合滤波器引入有源噪声控制系统,对该结构所采用的组合算法及系统稳定性、收敛速度、稳态误差进行了详细的理论推导,理论证明组合算法能够兼顾稳态误差与收敛速度取得良好的综合性能。分别对目前非线性噪声控制系统中采用的两种主流方法进行了研究,一种是基于函数连接型神经网络(FLANN)的方法,另一种是基于Volterra滤波器的方法。分别阐述了它们的控制系统基本结构及经典算法,并做了可行性分析。分别将这两种滤波器的组合结构引入非线性有源噪声控制系统,并对它们采用的组合算法进行了理论推导,针对传统组合算法在收敛过程中出现的停滞现象,采用附加瞬时转移结构对算法进行优化。研究发现FLANN结构具有结构简单、计算量较小等优点,主要应用于非线性程度较弱的情况。而Volterra滤波器具有较强的非线性处理能力,但核函数的数目和计算量都随输入信号长度呈指数式增加。因此,将这两种滤波器通过凸组合后引入有源噪声控制系统,以结合它们在非线性处理能力,计算复杂度方面的优势,对该组合结构及算法进行了推导。为降低算法计算复杂度,采用修正箕舌线函数对联合系数进行选取,采用sign函数对混合系数进行更新,降低计算量的同时,能够避免联合参数趋向0或1时,混合系数更新缓慢甚至停滞的现象。最后通过大量仿真实验对文中研究内容进行了验证。更多还原

【Abstract】 With the rapid development of China’s economy,people pay more and more attention to the quality of life.Reducing the noise of substation can reduce the impact of noise for the life of people around,which will greatly improve their quality of life.The active noise control system can reduce the low frequency noise of transformer substation.Therefore,the simple,economical and effective active noise reduction system has broad application prospects in the field of power transformer noise control.At present,most research of the active noise control system is based on linear system,of which the irreconcilable conflicts between the convergence speed and steady-state error exist.In actual operation,there are non-linear factors inevitably existing in the control system,which will bring about the great reduction of control performance for those linear systems.If he nonlinear problem of the active noise control system can be solved,the system can use the cheap electro acoustic device containing nonlinear distortion features,which not only can improve the noise reduction performance,but also is the special significance to reduce the cost of the system.Therefore,in this paper,the nonlinear active noise control technology of power transformer and the coordination method of convergence speed and steady-state error are studied,which is supported by the National Natural Science Foundation of China "Research on active noise control technology of array type power transformer based on internal model control"In order to solve the contradiction between convergence speed and steady-state error,the convex combination filter is introduced to the active noise control system.Detailed theoretical derivation of the stability,convergence speed and steady-state error of the control system,and the combined algorithm used in the system are carried out,which proved that the combinatorial algorithm can achieve the comprehensive performance of steady-state error and convergence speed perfectly and acquire good overall performance.In this paper,two main methods used in the current nonlinear noise control system are studied,one is based on the function link artificial neural network(FLANN)and the other is based on the Volterra filter.The basic structure and the classical algorithm of the control system are described,and the feasibility of two methods are analyzed.The convex combination structures of the two main methods are introduced into the nonlinear active noise control system respectively,and the combinatorial algorithms adopted are deduced theoretically.In order to solve the stagnation phenomenon appeared in the process of convergence according to traditional combination algorithm,the additional instantaneous transfer structure algorithm is used to optimize the algorithm.It is found that the FLANN has the advantages of simple structure and less computation,and is mainly used in the case of the weak nonlinearity.However,the Volterra filter has strong nonlinear processing ability,but the number of the kernel function and the amount of computation increases exponentially with the length of the input signal increasing exponentially.Therefore,the two kinds of filters are introduced into the active noise control system by convex combination to combine their advantages in nonlinear processing ability and computational complexity,of which the combination structure and algorithm are deduced.In order to reduce the computational complexity,the modified versiera function is used to select the association coefficient function and the sign function is used to update the mixing coefficient,which is possible to avoid the slow or even stagnation of the mixed coefficient and prevent the association coefficient from tending to be 0 or 1 when the amount of calculation is reduced at the same time.Finally,a large number of simulation experiments are carried out to verify the research contents.更多还原