论文部分内容阅读
提出一种基于时频展缩的随机共振算法,用于检测转子早期故障信号,该方法充分利用了随机共振检测弱信号的优势,通过引入时频展缩,消除了随机共振系统对待测信号频率的限制,实现了在绝热近似理论下,从强噪声中提取出转子微弱早期故障信号.理论分析和实测结果表明:对埋在强噪声中的未知频率的早期故障信号,采用连续的压缩算法,以获得一个适当的输入信号到随机共振体系,根据输出信号共振谐振峰的变化,经反变换运算可得待测微弱故障信号的未知频率.与传统方法相比,计算速度提高了4个数量级,能在极限信噪比(信噪比-50dB)下,提取出目标信号.
A stochastic resonance algorithm based on time-frequency shrinkage is proposed to detect early fault signals of rotors. This method makes full use of the advantages of stochastic resonance in detecting weak signals. By introducing time-frequency shrinkage, the stochastic resonance system can eliminate the influence of stochastic resonance on the signal frequency , The weak early fault signal of the rotor is extracted from the strong noise under the adiabatic approximation theory.The theoretical analysis and experimental results show that the continuous compression algorithm is applied to the early fault signal of unknown frequency buried in strong noise, In order to obtain an appropriate input signal to the stochastic resonance system, the unknown frequency of the weak signal to be measured can be obtained by inverse transform operation according to the resonant resonance peak of the output signal.Compared with the traditional method, the computational speed is improved by 4 orders of magnitude, In the limit signal to noise ratio (signal to noise ratio -50dB), extract the target signal.