卡尔曼滤波融合新算法及其船舶高精度液位测量的应用  

Kalman Filter Fusion Algorithm and Its Application in Highly Precise Measurement of Liquid Level in Ship Tank

作　　者：刘庆华[1,2] 仲海嘯 陶峰[1] 吴震宇[1] 

机构地区：[1]江苏科技大学计算机科学与工程学院,镇江212003 [2]同济大学交通运输工程学院,上海201804

出　　处：《中国造船》2016年第2期192-200,共9页Shipbuilding of China

基　　金：国家863高技术研究发展计划(2013AA12A206)

摘　　要：针对雷达液位计在船舶液位测量过程中由于液位波动、外界干扰或者其它突发状况导致的测量结果出现偏差、收敛速度慢以及不稳定等问题,提出了一种扩展卡尔曼滤波和加权算法相结合的融合算法。首先,采用动力学原理,对雷达液位计的测量过程进行建模,并对其测量结果进行扩展卡尔曼滤波;其次,引入多传感器融合思想,对滤波后雷达液位计的测量值与经过低通滤波后压力传感器的测量值进行加权融合。然后,采用实际船舶上的水舱进行了实验。实验结果表明:经过扩展卡尔曼滤波和多传感器融合后的液位值,能够具有更高的稳定性、可靠性和精度。In order to solve the problem that there exist deviation, slow convergence and instability for the measurement of fluctuated liquid level in a ship tank by radar level gauge, a new method that the extended Kalman filter algorithm is combined with weighted algorithm is proposed. A model for radar level gauge is built by using dynamic principle and then filtered results from extended Kalman filter can be obtained. Multi-sensor fusion theory is used to reconcile measurement results from radar level gauge and that from pressure sensors after low-pass filter. The methodology is validated with the test in a ship tank.

关 键 词：船舶 液位测量 扩展卡尔曼滤波 加权融合算法 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]