通信工程毕业设计（论文）-振动信号AR模型谱估计算法研究【全套设计】.doc

本科毕业设计(论文)
题目:振动信号AR模型谱估计算法研究
系别: 电子信息系
专业: 通信工程
班级:
学生:
学号:
指导教师:

2013年06月

摘要
振动信号的功率谱估计技术在电力、化工、冶金、矿山、船运等行业应用广泛,是相关领域中故障诊断及在线监测系统实施的基础。针对经典功率谱估计算法应用于振动信号谱估计存在方差与分辨力性能矛盾而导致谱估计质量不高的问题,论文研究了如何将现代谱估计算法中的AR模型谱估计算法应用于振动信号谱估计中。
AR模型谱估计算法属于参数模型谱估计范畴,论文首先对AR模型功率谱估计算法进行了深入研究与分析,包括算法模型、模型参数计算的Levision-Durbin递推算法和Burg算法,以及最优模型阶次确定的AIC(Akalke Information Criterion)准则。在理论分析的基础上,基于MATLAB进行了AR模型谱估计算法的仿真程序的编写。最后,基于MATLAB的GUI编程设计了图形用户界面,形成了一套完整的振动信号AR模型谱估计算法程序。
论文研究所设计的算法程序基于实测汽轮机振动信号进行了测试及性能分析。理论分析及仿真结果表明,AR模型谱估计算法在阶次合适的情况下,能够取得较好的方差及分辨力性能,适合于振动信号的功率谱估计,且AIC准则为确定合理的模型阶次提供了有效依据。
关键词:振动信号;AR模型谱估计;Levision-Durbin递推算法;Burg算法;MATLAB;信息论准则
全套设计,加153893706
Research of Vibration Signal’s AR Model Spectral Estimation Algorithm
Abstract
The power spectrum estimation techniques of the vibration signal are widely used in electric power, chemical industry, metallurgy, mining, shipping and other industries, is the basis for the implementation of fault diagnosis and online monitoring system in the related fields. Classic power spectrum estimation algorithm is applied to the vibration signal contradictions which led to the problem of spectral estimation is not high quality, the paper studies how modern spectral estimation algorithm AR model spectrum estimation algorithm applied to the vibration signal spectrum estimation.
AR model spectrum estimation algorithm is parameter model spectrum estimation category Firstly, the AR model power spectrum estimation algorithm is researched and analyzed in depth, including the algorithm model, the model parameters Levision-Durbin recursive algorithm and Burg algorithm, and the AIC(Akaike Information Criterion) criterion of determine the optimal model order .The preparation of the AR model spectrum estimation algorithm simulation program , on the basis of theoretical analysis, based on MATLAB, MATLAB-based GUI programming designed graphical user interface, plete set of vibration signal AR model spectrum estimation algorithm.
The al