非线性时间序列的动力学混沌特征自动提取技术

Dynamic chaos features auto-extracting technique of non-linear time series

College of Civil Aviation, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China

摘要: 提出了一种非线性时间序列混沌特征的自动提取方法.该方法直接根据非线性时间序列, 依次计算出延迟时间、嵌入维数、相关维数、最大李雅普诺夫指数、相轨迹特征以及Poincare特征等混沌特征, 整个特征的提取过程自动完成, 毋须人工干预.最后用算例表明了该方法的有效性和正确性.该方法对于利用非线性混沌特征进行非线性系统故障诊断和趋势预测具有重要意义.
- 航空、航天推进系统 /
- 非线性时间序列分析 /
- 混沌 /
- 特征提取 /
- 故障诊断
Abstract: A new method is put forward,which can implement auto-extracting of chaos features of nonlinear time series.It can directly from non-linear time series compute the chaos features such as delay time,embedded dimensions,correlation dimension,maximum Lyapunov exponents,phase track features and Poincare map features,and the whole features extracting process is carried out automatically without manpower intervention.Finally,examples show the validity and correctness of the new method.The new chaos features extracting method has important meaning to fault diagnosis of nonlinear system by nonlinear chaos feature.
- aero space propulsion system /
- nonlinear time series analy sis /
- chaos /
- feature extracting /
- fault diagnosis

[1]	徐章遂,房立清,王希武,等.故障信息诊断原理及应用[M].北京:国防工业出版社,2000:185-224.
[2]	任辉.转子碰摩故障诊断的信号分析方法研究[D].西北工业大学,2001:1-82.
[3]	Grassberger P,Procaaaia I.Measuring the strangeness of strange attractors[J].Physica,1982,D9:189-208.
[4]	Abraham N B,Albano A M.Calculating the dimension of attactors fromsmall data sets[J].Physics Letter,1986,114A(5):217-221.
[5]	Rosensrein M T,Collins J J and Luca C J D.Reconstruction expansion as a geometry-based framework for choosing proper delay times[J].Physia,1994,D73:82-98.
[6]	张雨.设备状态信息的现代辨识[M].长沙:国防科技大学出版社,2002:1-52.
[7]	关惠玲,韩捷.设备故障诊断专家系统原理及实践[M].北京:机械工业出版社,2000:56-72.
[8]	顾玉娣,过天晴,杨纪龙.概率论与数理统计[M].北京:航空工业出版社,2002:174-179.