针对不可靠网络传输下的离散奇异马尔科夫跳变系统,提出了H_∞鲁棒滤波器设计,基于网络在模态跳变以概率方式接近观测下,建立了一个全新的具有部分模态依赖的网络化滤波器存在的充分条件,其滤波器参数可以通过求解线性矩阵不等式而易于得到,且不需要对原系统矩阵分解。在系统模态跳变并以概率方式接近观测下,滤波误差系统是正则、因果和随机稳定且满足H_∞性能指标,并给出数例证明了所提出的网络化鲁棒H_∞滤波器设计是有效的。 Aiming at the discrete singular Markov transition system under unreliable network transmission, a H_∞ robust filter design is proposed. Based on the network’s mode hopping approach and probabilistic approach, a new model with partially modal dependence The sufficient condition for existence of the networked filter is that the filter parameters can be easily obtained by solving the linear matrix inequality without decomposing the original system matrix. Under the system modal jump and the probabilistic approach, the error filtering system is regular, causal and stochastic stable and satisfies the H_∞ performance index, and gives several examples to prove the proposed network robust H_∞filter design It is effective.