MEMS矢量水听器的温度特性及其补偿方法研究

Research on Temperature Properties and Compensation for MEMS Vector Hydrophone

文中以MEMS矢量水听器为研究对象,分析其温度特性,采用BP神经网络方法进行温度补偿。海水的温度范围为-2-30℃,在该范围内对水听器的温度特性进行测试。测试结果表明：随着温度的变化,MEMS矢量水听器产生温度漂移,使其灵敏度发生浮动,这样严重制约了水听器的测量精度和应用范围。采用BP神经网络算法对水听器进行温度补偿,将水听器测量电压值和实时温度进行数据融合,削弱了环境温度的影响。补偿后,水听器的温度漂移显著降低,不同工作温度下的灵敏度曲线高度重合,灵敏度浮动不超过2 d B。

The temperature properties of the MEMS vector hydrophone was analyzed in this paper and temperature compensation by using BP neural network method was realized.The sea water temperature is in the range of-2 - 30 ℃.Within the temperature scope,the temperature properties of hydrophone were tested.The result shows that the MEMS vector hydrophone occurs temperature drift and its sensitivity has obvious fluctuation with the change of temperature. This severely restricted the hydrophone measuring precision and its application range.BP neural network algorithm was used to realize the temperature compensation for hydrophone.This method fuses the voltage signals selected by the hydrophone with the real-time temperature,which weakens the interference of the temperature on the hydrophone.After compensation,the temperature drift of hydrophone greatly reduces. Compared with the sensitivity curve before compensation,the sensitivity after compensation highly coincidences and the floating of sensitivity is less than 2 d B under different temperatures.