基于MIMO模型的潜艇声学故障预报研究

Study on Prediction of Submarine Acoustic Fault Based on MIMO Model

利用多输入多输出(MIMO)量化分析模型对潜艇潜在故障源进行分析,并通过频段因果性检测解决M IM O模型各输入源的优先权问题,从而方便地得到了各源相对于叠加场的贡献百分比动态曲线,以用于潜艇声学故障的预报.给出了潜艇声学故障预报的一般流程,并阐述其应用步骤.用一个水池实验验证了文中所提出的方法的可行性.

The concept of submarine acoustic fault and its prediction is brought forward for the first time. Based on the causality checking of inputs, the multi-input/multi-output （MIMO） modeling is employed to analyze and obtain the contribution per cent curve of each noise source, and thus to predict acoustic fault. Then a general flow and process of acoustic fault prediction is also given. Finally, a model ship towing tank experiment validates the feasibility of the proposed method.