一种自适应滤波与干扰观测器相结合的大型舰船状态估计算法

An Adaptive Filtering-disturbance Observer-based State Estimation Algorithm for Large Ships

为满足航母等大型舰船目标的状态估计要求,提出一种由非线性干扰观测器和强跟踪容积卡尔曼滤波算法融合形成的交互多模型强补偿容积卡尔曼滤波算法。引入非线性干扰观测器,完成由外界不确定因素引起的干扰总量的估计,并对观测器稳定性进行证明。使用估计的干扰值实时修正强跟踪容积卡尔曼滤波的过程参数,最终形成交互多模型强补偿容积卡尔曼滤波算法,完成对目标状态相对准确的估计。研究结果表明:新提出的滤波算法能够较为准确地完成对目标状态的估计,与变结构多模型粒子滤波算法、变结构多模型无迹卡尔曼滤波算法和交互多模型强跟踪容积卡尔曼滤波算法相比,在目标位置和速度估计上具有更高的估计精度。

In order to meet the state estimation requirements of large ship targets such as aircraft carriers,an interactive multi-model strong compensating cubature Kalman filtering algorithm is proposed,which is formed by the fusion of nonlinear disturbance observer and strong tracking cubature Kalman filtering algorithm.The nonlinear disturbance observer is introduced to estimate the total amount of disturbance caused by external uncertainties and prove the stability of the observer,and then the estimated disturbance value is used to modify the process parameters of the strong tracking cubature Kalman filter in real time,which finally forms the interactive multi-model strong compensating cubature Kalman filtering algorithm and completes the relatively accurate estimation of target state.The results show that the proposed filtering algorithm can complete the more accurate estimation of target state,and has higher estimation accuracy in the estimation of target position and velocity compared with the variable-structure multi-model particle filtering algorithm,the variable-structure multi-model unscented Kalman filtering algorithm,and the interactive multi-model strong tracking cubature Kalman filtering algorithm.