函数模型误差弹性自适应滤波修正

Resilient Adaptive Filtering Based on Model Modification and Underwater Application

多源导航信息融合过程中,观测模型和动力学模型随时间和空间变化复杂,高精度的动态载体导航与定位需要观测模型和动力学模型具有准实时或实时修正的能力。针对包含观测模型误差以及动力学模型误差的滤波系统,提出了一种基于信息滤波的弹性自适应滤波算法。所提算法以不含模型误差的标准信息滤波器为主滤波器,分别构造了观测函数模型及动力学函数模型误差补偿滤波器,对两类模型误差进行补偿。所提方法强调模型补偿项的弹性自适应估计和状态参数的弹性组合,提高了时变模型误差估计的稳定性。半物理仿真实验结果表明,基于函数模型补偿的弹性自适应滤波算法可以有效地估计观测模型和载体动力学模型误差项,水下拖体的三维位置偏差在0.2 m以内,两类模型误差的影响基本消除,明显提高了载体动态参数的估计精度。

Objectives:Multi-sensor information integration provides important support for positioning,navigation and timing services in extreme environments.High-precision dynamic vehicle navigation and localization requires observation models and dynamic models with the capability of quasi-real-time or realtime correction.In the process of multi-source navigation information fusion,the observed model and noise of navigation sensors vary in time and space,so it is of great theoretical importance and practical value to construct accurate functional and stochastic models.Methods:A resilient adaptive filtering algorithm based on information filtering is proposed for filtering systems containing both observational and dynamic model errors.In the proposed algorithm,the standard information filter without model errors is used as the primary filter,and error compensation filters for the observed and dynamic function models are constructed to compensate the errors of the two types of models,respectively.Results:The proposed approach emphasizes the elastic adaptive estimation of the model compensation term and the elastic combination of the state parameters,which improves the stability of the error estimates for time-varying models.Semi physical simulation experimental results show that the elasticity of the adaptive filter algorithm based on function model compensation can effectively estimate the observation model error term and vehicle dynamics model,with the error of three-dimensional position of the underwater towed body deviation less than 0.2 m.Conclusions:It can fundamentally eliminate the influence of two kinds of model errors and clearly improve the estimation accuracy of vehicle dynamic parameters.