Adaptive Kalman Filtering Based on Model Parameter Ratios

• 成果类型: Article
• 署名作者: Ge, Quanbo; Li, Yunyu; Wang, Yuanliang; Hu, Xiaoming; Li, Hong; Sun, Changyin
• 署名单位: Nanjing University of Information Science & Technology; Hangzhou Dianzi University; Shanghai Maritime University; Royal Institute of Technology; Southeast University - China
• 刊物名称: IEEE TRANSACTIONS ON AUTOMATIC CONTROL
• ISSN/ISSBN: 0018-9286
• DOI: 10.1109/TAC.2024.3376306
• 发表日期: 2024
• 页码: 6230-6237
• 关键词: Noise measurement; Kalman filters; Q measurement; estimation; Adaptation models; Covariance matrices; Time measurement; estimation error; inaccurate models; Kalman filter (KF); model parameter ratio (MPR); particle swarm optimization (PSO)

摘要

This article studies an adaptive Kalman filter method based on model parameter ratio. The model parameter ratio theory is proposed for the first time, and the adaptive estimation problem is transformed into a constrained optimization problem. Compared with the existing Sage-Husa adaptive filtering algorithm, it can be seen that the application of this theory can more accurately estimate the process noise covariance and measurement noise covariance matrix, so that the algorithm has better filtering accuracy and better state estimation performance, At the same time, it is also better in antidivergence and sensitivity to initial conditions.