基于卡尔曼滤波的水下航行器水动力系数辨识方法

Hydrodynamic Coefficients Identification Method for Underwater Vehicles Based on Kalman Filter

文中建立了水下航行器垂直面内机动水动力系数辨识模型,分别采用扩展卡尔曼滤波(EKF)和平方根无迹卡尔曼滤波(SRUKF)算法,对模型中突出影响操纵性能的水动力导数进行了辨识.仿真结果表明:当首尾升降舵都采用幅值为10°正弦规律操舵,将水下航行器的状态参数作为测量输入,两种算法都有很高的辨识精度,与水动力导数试验测定值进行比较,SRUKF和EKF的最大辨识误差仅2.98%和0.91%.此外,由于水下航行器垂直面运动模型线性度较高,无论是辨识精度还是收敛速度,EKF算法的性能都优于SRUKF,更适用于线性水动力系数的在线识别.

In this paper,the hydrodynamic coefficient identification model of underwater vehicle maneuvering in vertical plane was established.The extended Kalman filter(EKF)and square root unscented Kalman filter(SRUKF)algorithms were used to identify the hydrodynamic derivatives that significantly affect the handling performance in the model.The simulation results show that the two algorithms have high identification accuracy when the head and tail elevators are steered with sine law with amplitude of 10 and the state parameters of underwater vehicles are taken as measurement inputs.Compared with the hydrodynamic derivative test,the maximum identification errors of SRUKF and EKF are only 2.98%and 0.91%.In addition,because of the high linearity of the vertical motion model of underwater vehicle,the performance of EKF algorithm is better than SRUKF in both identification accuracy and convergence speed,and it is more suitable for online identification of linear hydrodynamic coefficients.