中文摘要：

设备状态信号的处理是状态监测及故障诊断的基础。在实际运行环境中，通过传感器检测的信号有可能是振源信号的非线性混合、畸变信号。传统的振动信号处理方法对此种非线性混合信号处理效果并不理想。非线性盲源分离技术由于自身独特的盲处理优势，可以有效的去除外来干扰并分离出源信号，有助于提高诊断的准确性。针对直升机齿轮箱振动信号的非线性混叠进行盲源分离，分离出轴承故障振动信号，表明盲源分离技术是机械故障诊断领域的一个有效的信号处理方法。

英文摘要：

Equipment condition signal processing is the foundation of condition monitoring and fault diagnosis. In the actual environment, the mechanical vibration signals collected by the sensors are probably the nonlinear mixture of the source vibration signals or distorted signals. The traditional vibration sig- nal processing methods are not good enough for processing the nonlinearly mixed signals. However, nonlinear Blind Sources Separation （BSS）, due to its unique superiority in blind processing, can remove noises and external perturbations in the collected signals effectively and extract the source signals. As a result, it raises the accuracy of the diagnostic performance. In this paper, applying the nonlinear BSS technique, the vibration signals of bearing faults in a helicopter gearbox were extracted. The results show that the method of nonlinear BSS is an effective method for signal processing in the field of mechanical fault diagnosis.