基于小波包和混沌理论的状态预测方法研究

Research on Prediction Method based on Wavelet Packet Transformation and Chaos Theory

通过对某系统的压力时间序列进行多层次的小波包分解,对分解得到的各层次的时间序列进行分析,建立相应的混沌预测模型并进行预测,最后应用小波包重构算法,将各层次的预测结果进行重构,从而获得时间序列的预测结果。试验表明,该方法具有较高的预测精度,可有效地应用于系统的状态监测和故障预测。

This paper presented a kind of prediction method which combines wavelet packet transformation with chaos theory. Firstly, it used wavelet packet to decompose the time series of pressure for a certain system and analyzed every level time series. Secondly, it built the corresponding chaotic prediction model and did prediction. Finally, it applied the recon- struction algorithm of wavelet packet to reconstruct all levels of prediction results and obtained the prediction results of the time series. Experimental test showed that the method has highly precise and can be effectively applied in the system of con- dition monitoring and prediction.