Delay Research On Networked Control System Based On T-S Model

Wireless networked control system, Compared with traditional point-to-point mode, ismore advanced. However, the introduction of computer communication network has causedmany new inevitable problems, such as packet loss, transmission delay and so on. The datatransformed in wireless network must be quantized before transmission with constrained of thebandwidth, which will cause the quantization error. Therefore, there is a serials of new problemsas how to reduce the effect of quantization errors and delay on the system performance, andmany scholars have committed themselves to the study of the controlling problem. And theauthor of this thesis also conducts a series of studies on the issue.In this thesis, we put emphasis on the wireless networked control system. First, Due to thetime-delay of the system, the neutral system with mixed delays is chose for the study, which isa special case of networked control system. A novel delay-dependent robust stability criteria isproposed. Then, logarithmic quantizer is also chosen. Since it can cause quantization error, amethod has to be used to solve the problem; and this method can transform quantization error tothe uncertainty. Last, T-S fuzzy model is introduced into the networked control system. Someresults about T-S fuzzy model with a time-delay term are obtained to deal with the problem ofstability and stabilization problems.The content of this thesis is as follows:1.Some new delay-dependent robust stability criteria for neutral system with mixed delaysand time-varying structured uncertainties are proposed. The criteria have taken into accountthe lower bounds and upper bounds of mixed delays. The criteria proposed in this section areboth neutral-and discrete-delay-dependent. Some numerical examples have been provided toillustrate the improvement of the proposed results.2.A new delay-dependent stability criterion for the NQCSs has been proposed. Based onLyapunov-Krasovskii functional and LMI, the proposed criterion incorporates the asymptoti-cally stability, H∞performance analysis and H∞control analysis by introducing some slackmatrices and scalars. We have shown, by means of theoretical and numerical results, that isconstructive method to the system with time-delay often lends to less conservative results.3.A robust H∞control method has been proposed for a class of nonlinear NCSs, whichcan be represented by a T-S fuzzy model with uncertainties. The developed controller can beobtained by solving a set of LMIs based on a Lyapunov-Krasovskii functional method. In theproposed approach, the maximum allowance control interval δ is adopted to reflect both the network-induced delay and packet dropout problems. Moreover, the network-induced delay τkneeds not to be restricted within one sampling period.