基于扩展卡尔曼滤波算法的船舶姿态监测预报系统设计（英文）

Monitoring and forecasting system for ship attitude motion based on extended Kalman filtering algorithm

为提高船舶姿态监测预报系统的测量精度,设计加速度传感器和陀螺仪捷联的惯性导航系统,利用互补处理和反馈调节对传感器输出参数进行校正。为得到精度较高的预估数据,通过广义最小二乘法进行系统辨识建立船舶状态模型和矫正的传感器测量模型,由扩展的卡尔曼滤波器得到最优估计状态。同时设计闭环系统,进一步提高系统的测量精度和预估精度。此外,针对当前监测预报系统普遍存在的软件界面差、交互困难等缺点,设计姿态监测预报友好型界面,可3D动画实时显示船舶当前的运动姿态,并具有数据存储、分析、操作处理等功能。最后,通过实验对不同倾角变化率下实际姿态、测量姿态和预估姿态进行对比,结果表明所设计的姿态监测预报系统具有较高的监测精度和预估精度。

To improve the measurement precision of ship attitude motion monitoring and forecasting system, a cascaded inertial navigation system with accelerator and gyroscope is developed, which utilizes complementary measurement and feedback adjustment to amend the parameters of sensor＇s output. Meanwhile, aiming at high-precision forecasting data, a generalized least squares method is utilized for modeling the ship status and the corrected sensor measurement, then an Extended Kalman Filtering（EKF） algorithm is utilized to optimize the estimate status. Moreover, the close loop control architecture is designed to further improve the precisions of measurement and prediction. Additionally, a friendly interactive interface is designed to facilitate the human interaction, which can illustrate the current attitude in 3 D with functions of store, analysis, and operation of real-time data. Comparisons among the actual data, the measured data and the predicted data from the experiment demonstrate the high precisions of supervision and prediction of the designed ship attitude motion monitoring and forecasting system.