船体梁爆炸损伤识别方法研究

Research on identification method of explosion damage of ship hull girder

针对目前在船舶爆炸损伤试验中不能有效获得船舶损伤数据,将爆炸过程中船舶船体梁总振动固有频率与爆炸损伤前完好船体总振动固有频率进行对比,建立完好船体和损伤船体多种假想损伤的固有频率数据库,采用神经网络智能方法得到频率改变特征参数与损伤参数的映射关系,对船舶结构损伤状态进行判断。以一艘1500吨级的船舶为研究对象,通过仿真计算船舶结构损伤前后各阶固有频率,基于频率改变特征参数与损伤参数的神经网络智能方法识别船舶结构损伤位置和损伤程度,损伤位置全部定位正确,损伤程度平均精度达97.91%。该方法适用于爆炸损伤条件下的船舶结构损伤识别,新颖有效。

Aiming at the problem that the ship damage data can not be obtained effectively in the ship explosion damage test at present,based on the ship hull girder in the process of explosion vibration natural frequency and compared with the natural frequencies of the intact hull before explosion damage,the natural frequency databases of various hypothetical damage of intact and damaged hulls were established.The mapping relationship between frequency change characteristic parameters and damage parameters was obtained by using neural network intelligent method,and then the identification method of ship structural damage state was obtained.Taking a 1500-ton ship as the research object,through the simulation calculation of ship structural damage before and after each order natural frequency,the damage parameters of the neural network intelligent method can be a very good identification of ship structure damage location and damage degree based on the frequency change characteristic parameters.The damage location are all correct orientation and the damage degree average accuracy is 97.91%.This method is suitable for damage identification of ship structure under explosion damage condition and provides a novel and effective method for damage identification of ship structure.