摘要
水中兵器爆炸试验测量是获得兵器爆炸威力的重要手段,而数据的预处理是水中兵器毁伤评估的重要环节和基础.本文针对自由场压力数据中常见的时域曲线趋势项、工频干扰、通道串扰、上升沿误差等干扰,研究了相关的预处理方法,并以自由场压力传感器实测数据实例,分别以串联共轭梯度去除法用于去除工频干扰,将多项式拟合用于去除低频趋势项,将动窗稳健中值滤波法用于去除脉冲干扰,将指数函数拟合用于冲击波脉冲上升沿修正.设计了对自由场压力测量数据的预处理流程,并以时频域特性、统计特性等评估了预处理算法的合理性.其中,串联共轭梯度法对于固定频率工频干扰去除效率高,而对其他频率成分影响很小.对于冲击波上升沿的指数拟合修正,通过使用牛顿截断法对指数拟合中需要用到的参数进行优化,并以拟合前后的能量误差和峰值误差评价修正效果.
Experimental measurements in the explosive testing of underwater weapons are important tactics to obtain the explosive power of a weapon. Preprocessing measurement data in the free-field pressure test is an important basis for approaching and extracting the true values of the measurement signals. By analyzing and calculating characters from the processed data, the explosive power value of weapons can be reasonably assessed. In this paper, we investigated preprocessing methods to eliminate common interference noise in free-field pressure tests, such as time-domain curve tendency, power frequency interference, interference between test channels, and rising edge error of shock wave impulses. The underwater free-field pressure data tested by the pressure sensor obtained in the explosion test was used to study the preprocessing flow design. It was processed by polynomial fitting to remove tendency item, serial conjugate gradient method to remove power frequency interference, moving window medial filtering to reduce pulse interference, and exponential function fitting to correct the rising edge of shock wave impulses. A preprocessing flow was designed for underwater free-field pressure measurement data. Its validity was evaluated by the comparative analysis of time domain and frequency domain characteristics and statistical characteristics of the data before and after preprocessing. The serial conjugate gradient method significantly reduces power frequency interference at certain frequencies and affects signals with other frequencies scarecely. In the exponential fitting of shock wave impulses, the relevant parameters of the fitting curve were optimized by the truncated Newton method. The effect of the exponential fitting of the peak value correction was evaluated by power error and peak value error.
作者
权琳
陈梦英
陈琦
郭颖
程素秋
QUAN Lin;CHEN MengYing;CHEN Qi;GUO Ying;CHENG SuQiu(Unit 91439 of China PLA,Dalian 116041,China;Shanghai Acoustics Laboratory,Chinese Academy of Sciences,Shanghai 201800,China)
出处
《中国科学：物理学、力学、天文学》
CSCD
北大核心
2021年第12期97-104,共8页
Scientia Sinica Physica,Mechanica & Astronomica
关键词
水下爆炸
压力
预处理
underwater explode
pressure
pre-process
