不完全测量下的复杂网络状态估计

发布时间：2018-07-26 14:38

【摘要】：网络监控和故障诊断往往需要实时掌握网络中的所有节点状态信息。而复杂网络规模大,节点多且复杂,因此直接测量网络中各个节点状态信息是不现实。为了获得网络中每个节点的状态信息,继而深入地了解网络的行为特征,就需要我们通过已测节点的输出信息对整个网络状态信息进行估计。此外,现实网络会由于种种因素导致信息在传输过程中不可避免地存在时延和丢包,所以研究不完全测量下的复杂网络状态估计显得更具有现实指导意义。本文的主要研究工作及取得的成果如下:(1)现实网络在传输过程中会受到丢包、时延等不确定因素的干扰,这些干扰的出现势必会对复杂网络的状态估计产生影响。本文在之前已有的理论基础上,设计了一种鲁棒保性能状态估计器,旨在减少数据丢包和诱导时延对状态估计的影响。基于Lyapunov稳定性理论和随机分析方法,以线性矩阵不等式的形式给出了鲁棒保性能状态估计器的设计准则。最终,通过对不同规模的复杂网络进行数值仿真,验证了设计方案的有效性。(2)现实网络中可能存在部分节点输出不可测的现象,同时网络时延在k时刻的输出更依赖于(k-1)时刻的输出情况。本文将牵制控制的方法运用到具有一步诱导时延的离散复杂网络的状态估计中。通过对部分节点施加反馈控制,以达到对整个复杂网络进行状态估计的目的,从而大大降低了状态估计的成本,更加符合实际的工程应用。利用Lyapunov稳定性理论,推出牵制控制状态估计器的设计准则。最后,通过数值仿真以验证结果的可行性。(3)根据结构可控性理论选取驱动节点进行牵制控制。利用Lyapunov稳定性理论,得出增益反馈控制器设计准则。最后,通过实例仿真,验证所设计反馈控制器的有效性和可行性。
[Abstract]:Network monitoring and fault diagnosis often need to grasp the status of all nodes in the network in real time. The complex network is large in scale and complex in nodes, so it is not realistic to directly measure the state information of each node in the network. In order to obtain the state information of each node in the network and further understand the behavior characteristics of the network, we need to estimate the state information of the whole network through the output information of the measured nodes. In addition, due to various factors, delay and packet loss will inevitably exist in the transmission of information in real networks, so it is more important to study the state estimation of complex networks under incomplete measurement. The main research work and results obtained in this paper are as follows: (1) in the process of transmission, real networks will be disturbed by uncertain factors such as packet loss, delay and so on, which will inevitably affect the state estimation of complex networks. Based on the previous theories, a robust guaranteed cost performance state estimator is designed to reduce the influence of data packet loss and induced delay on state estimation. Based on Lyapunov stability theory and stochastic analysis method, the design criteria of robust guaranteed cost state estimators are given in the form of linear matrix inequalities (LMIs). Finally, the effectiveness of the design scheme is verified by numerical simulation of complex networks of different sizes. (2) the phenomenon of unmeasurable output of some nodes may exist in the real network. At the same time, the output of network delay at k time is more dependent on the output of (k-1) time. In this paper, the containment control method is applied to the state estimation of discrete complex networks with one step induced delay. By applying feedback control to some nodes, the state estimation of the whole complex network is achieved, and the cost of state estimation is greatly reduced, which is more in line with the practical engineering application. Based on the Lyapunov stability theory, the design criteria of the control state estimator are derived. Finally, the feasibility of the results is verified by numerical simulation. (3) according to the theory of structural controllability, the driving nodes are selected to control. Using Lyapunov stability theory, the design criteria of gain feedback controller are obtained. Finally, the effectiveness and feasibility of the designed feedback controller are verified by simulation.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O157.5

【相似文献】