Li Xiao-Meng, Zhang Bin, Li Panshuo, Zhou Qi, Lu Renquan
IEEE Trans Neural Netw Learn Syst. 2020 May;31(5):1450-1460. doi: 10.1109/TNNLS.2019.2920368. Epub 2019 Jul 1.
The problem of finite-horizon H state estimator design for periodic neural networks over multiple fading channels is studied in this paper. To characterize the measurement signals transmitted through different channels experiencing channel fading, a multiple fading channels model is considered. For investigating the situation of correlated fading channels, a set of correlated random variables is introduced. Specifically, the channel coefficients are described by white noise processes and are assumed to be correlated. Two sufficient criteria are provided, by utilizing a stochastic analysis approach, to guarantee that the estimation error system is stochastically stable and achieves the prescribed H performance. Then, the parameters of the estimator are derived by solving recursive linear matrix inequalities. Finally, some simulation results are shown to illustrate the effectiveness of the proposed method.
本文研究了多衰落信道上周期神经网络的有限时域H状态估计器设计问题。为了刻画通过经历信道衰落的不同信道传输的测量信号，考虑了一种多衰落信道模型。为了研究相关衰落信道的情况，引入了一组相关随机变量。具体而言，信道系数由白噪声过程描述，并假设它们是相关的。通过利用随机分析方法，提供了两个充分准则，以保证估计误差系统是随机稳定的，并实现规定的H性能。然后，通过求解递归线性矩阵不等式来推导估计器的参数。最后，给出了一些仿真结果以说明所提方法的有效性。
