论文部分内容阅读
发动机表面振动是一种多振源、激励复杂的振动形式,在发动机测试时目标振动信号受到很强的噪声干扰,必须通过有效方法进行信噪分离。介绍了一种利用人工神经网络进行自适应滤波的信噪分离方法,根据自适应噪声抵消原理建立了ADALINE自适应神经滤波器模型,并使用该模型滤除坦克发动机汽缸盖振动信号中的机体振动噪声,提高了信噪比,为汽缸盖振动信号的进一步分析处理奠定了基础。
Engine surface vibration is a multi-vibration source that excites complex vibrational forms. When the engine test is performed, the target vibration signal is strongly disturbed by noise and the signal-to-noise separation must be effectively performed. A method of signal-noise separation based on artificial neural network for adaptive filtering is introduced. According to the theory of adaptive noise cancellation, an ADALINE adaptive neural filter model is set up. The model is used to filter the vibration of tank engine cylinder head vibration Noise, improve the signal to noise ratio, the vibration signal for the cylinder head further analysis and processing laid the foundation.