复合多稳随机共振阵列微弱信号检测理论与应用研究

本项目基于随机共振原理，建立复合多稳随机共振模型，研究在多频微弱周期力和宽带随机力共同激励下多稳随机共振在不同稳态间的切换机制，探究在不同时间尺度的两种力作用下多稳系统的输出响应，解决具有极低信噪比的多频微弱信号检测难的问题；基于复合多稳模型构建随机共振阵列，研究α稳定分布噪声和多频微弱信号激励的随机共振阵列中噪声分布参数、多稳系统参数对系统共振输出响应的作用规律，解决淹没在具有尖峰脉冲特性噪声中的宽频带多频微弱信号检测难的问题；研究参数诱导多稳随机共振阵列中多个系统参数的自适应并行协同优化方法，建立多稳随机共振反演系统，实现微弱特征信号的快速增强检测与信号恢复；利用研究得到的非线性微弱信号增强检测理论和方法，以检测单晶炉中引起硅液面异常波动的振源信息为目标展开应用研究。

In this project, establishing the composite multi-stable stochastic resonance model based on the theory of stochastic resonance, the switching mechanism of multi-stable stochastic resonance excited by multi-frequency weak periodic force and broadband random force is explored, and the output response of multi-stable system under two forces at different time scales is studied, the problem of the detection of multi-frequency weak signal with very low SNR is solved. Besides, building the stochastic resonance array based on composite multi-stable stochastic resonance model, the evolution law of stochastic resonance output response with α stable noise distribution parameters and multi-stable system parameters is explored in stochastic resonance array excited by α stable noise and multi-frequency weak signal, the problem of the detection of multi-frequency weak signal submerged in the spike pulse noise is also solved. Then the adaptive parallel collaborative optimization method for multiple system parameters in parameter-induced stochastic resonance arrays is studied, and the multi-stable stochastic resonance inversion system is established, which realize the fast and enhanced detection of weak characteristic signal. The enhanced detection theory and method of non-linear weak signal is used to detect the information of vibration source which leads to an abnormal fluctuation of silicon liquid surface in single crystal furnace.