Neural-Network-Based Distributed State Estimation Under Encoding-Decoding Schemes: Probabilistic-Constrained Cases
针对传感器网络中的非线性系统，提出一种基于神经网络的分布式状态估计方法，在编码解码通信下以指定概率将估计误差限制在给定区域内，并保证有限时间指数有界性能。
In this article, a neural-network (NN)-based approach of distributed state estimation with probabilistic constraints is proposed for a class of nonlinear systems over sensor networks. For the discussed plant, the unknown nonlinear dynamics are approximated by resorting to NNs and the communication among estimators and sensors is scheduled by encoding–decoding schemes. The goal of the addressed problem is to design a distributed estimator such that, in the presence of the bounded noises, all possible errors are confined to some certain region in a predetermined probability while achieving the exponentially bounded performance in a finite time domain. In light of the matrix operation, some sufficient conditions are obtained to ensure the existence of the desired gains of estimators, which are computed by dealing with the corresponding matrix inequalities in an iterative way. The effectiveness of the proposed distributed state estimation method is verified by presenting an example of a one-track model.