Distributed Receding Horizon Estimation for Time Invariant Discrete Time Linear Systems Based on Substate Decomposition

作者：Zenghong Huang, Zijie Chen, Yong Xu, Chang Liu, Peng Shi · 发表于：IEEE Transactions on Network Science and Engineering · 年份：2025 · DOI：10.1109/tnse.2025.3590754 · 被引用次数：3 · 研究领域：Fault Detection and Control Systems、Target Tracking and Data Fusion in Sensor Networks、Advanced Control Systems Optimization

This paper investigates distributed receding horizon estimation (DRHE) for time-invariant discrete-time linear systems over a sensor network. The original system is decomposed into several low-dimensional subsystems, where each sensor is capable of observing only one specific subsystem. The observable substate for each node is estimated by minimizing a local cost associated with receding horizon estimation (RHE), while the prediction of unobservable substates is updated through one-step weighted fusion. A maximal directed acyclic graph (MDAG) is introduced to facilitate the construction of the weight values for fusing these predictions, which is a more general method compared to directed spanning trees. Additionally, we propose a novel algorithm for identifying an MDAG across the network. We establish sufficient stability conditions for the proposed estimator under the assumption of collective observability. Finally, a numerical example of temperature monitoring is presented to demonstrate the effectiveness of the developed method.