Networked Control Systems With Delay-dependent Stability

In the modeling of network control system and the analysis of the model stability, the common method is to identify model changes. But in this method, relationship of the objects in the Newton-Leibniz formula is fixed, therefore, the derived results lack universality and the maximum allowable delay bound is relatively small. For the limitations of the identifying model changes, this thesis has researched on the stability and the stabilization of the network control system by the free-weight matrix method and the improved free-weighting matrix method.This thesis has analyzed the effect caused by the time-delay on the performance of the network control system, has extracted several important time-delays of the network control system, and has given the network control system model and analyzed its stability based on the time delay performance and existing models.For the continuous network control system which does not consider the data packet loss in the transmission process, this thesis has established the network control system with single input and output and the one with multi-input and output and their corresponding uncertain system model, then with the integral inequality method and the free-weight matrix method, this thesis has worked out the stability criterion of the network control system with single input and output and its corresponding maximum upper limit of the time-delay, and has extended to the system with multi-input and output. And the data example shows that the obtained results are more conservative than the existing ones.For the network control system with single-packet transmission, this thesis has considered the network induced time-delay the data-packet loss in the transmission courses, and has gained the stability criterion of the system by the method of the improved free-weight matrix, at the same time, under the situation of given controller and based on the linear matrix inequalities, this thesis has also obtained the calculation method of the maximum allowable network induced time-delay.