摘要
针对现有随机共振以信噪比为测度，难以进行量化和在实际工程中应用，提出一种符号序列改进型香农熵结合随机共振的微弱信号检测方法。介绍了随机共振和符号序列化的基本原理，以输出信号的符号序列熵作为判断是否达到最佳随机共振的测度，自适应的调节系统参数a和b，使系统达到最佳共振并进行频谱分析。结果表明输出信号的符号序列熵可有效反映共振状况，能调节系统达到最佳信噪比输出，而且易于工程实现，证实了该方法的有效性，进而为衡量随机共振提供一种新方法。
A method of weak features detection for mechanical fault data is proposed, which combined the stochastic resonance and the improved Shannon entropy of symbolic series. Firstly, the basic principles of SR and symbolic sequence analysis （SSA） is briefly introduced. Secondly, symbolic sequence Shannon entropy of output signal is the measure whether the state of this system of SR is the best one. Finally, according to adjust system parameters a and b adaptively, the system of SR achieves the best state and analyzes signal to noise ratio （SNR） and spectrum of output signal. Information of output signal show that Shannon entropy of symbolic series can reflect the state of SR , the best SNR of output signal can be achieved and it can be easily used in practical projects, which confirms the new method proposed is correct and useful.
出处
《计量学报》
CSCD
北大核心
2015年第5期496-500,共5页
Acta Metrologica Sinica
基金
甘肃省省自然科学研究基金(1308RJZA273)
