说明：  采用基于奇异值分解和人工神经网络的多传感器数据融合方法对喷水推进泵的空化状态进行了分类识别研究。首先利用基于奇异值分解的权值估计算法分别对水声信号和振动信号在时间上进行数据级融合，提取出各自的特征，然后将所有特征组合起来作为神经网络的输入，利用ＢＰ网络和ＲＢＦ网络进行特征级融合和分类识别。
(The use of water jet propulsion pump cavitation state multi-sensor data fusion method based on singular value decomposition and artificial neural network classification and recognition. First, based on the singular value decomposition weights estimation algorithm level data fusion underwater acoustic signals and vibration signals in time, extract individual characteristics, then combined all features as the input of the neural network, using BP and RBF network feature fusion and classification.)