A New Approach of Signal Extension and its Application in Internet-based Remote Diagnosis Systems

摘要

In Internet-based Remote Diagnosis System, signal analysis is one of the important approaches to extract the system knowledge and fault information. Wavelet analysis and Empirical Mode Decomposition can be used to process the nonstationary and short time signals, but both of them have serious boundary effect. If the problem is not solved, the transform results will be meaningless. Conventional signal extension approaches, such as even-extension and odd-extension, are unsatisfactory in practices. Thus a new approach of signal extension is presented in the paper. It proves to be more effective to eliminate the boundary effect through the experiments. The implement principle and procedure about the new approach are also given in this paper.