Networked Control Of A Three-tank System

With the rapid development of the computer-controlled technology andcomputer networks technology, it has come true that having one or more controlloops closed via a serial communication channel. Feedback control systemswherein the control loops are closed through a real-time network are callednetworked control systems(NCSs). Integrating computer networks into controlsystems to replace the traditional point-to-point wiring has enormous advantages,including lower cost, reduced weigh and power, simple installation andmaintenance, and higher reliability.The insertion of the communication network in the feedback control loopmakes the analysis and design of a NCS complex. The first issue is thenetwork-induced delay that occurs while exchanging data among nodes ofnetworks connect to the shared medium. The transmission delays are often randomand mostly longer than one sampling period. Secondly, Due to the bandwidthconstraint of the network, network scheduling which appears when reasonablyallocating the shared network resources. At the same time, data packet dropout mayoccur during data transmission. Because of the bandwidth constraint and capacitylimit of data packet, one-packet transmissions and multiple-packet ones areinevitable. These problems not only degrade the performance of the controlsystems, even make the systems unstable. So the stability analysis and design of thenetworked control systems have gained more and more attention.The multi-vessels flow system is a typical object in the process control. It hasextremely widespread application background in the production. The three-tanksystem is a nonlinear control system with double input and double output, whichsimulate the multi-vessels flow system. In the article, the three-tank system is usedto study the control algorithm. Firstly, we have modeled and linearized the system.Then we get the discrete model of three-tank system. At the same time, weintroduce the hardware and the setup of Three-Tank system based networkedcontrol system.The network-induced delay occurs while exchanging data among sensor,controller and actuator through networks. The network-induced delays areinevitable, which are the major cause of the deterioration of system dynamicperformance and potential system instability. Therefore, it is imperative to analyzethe stability of NCS with network-induced delay. In Chapter 3, we have modeledcontrol systems with network-induced delay. During the course, the delays aredivided into two kinds: delays less than one sampling period and longerdelays(longer than one sampling period). For delay less than one sampling period,one can always receive a control sampling in a single sampling period. However,this may not be true for longer delays. For longer delays, one may receive zero, one,or more than one control samples in a single sampling period. Then, we analyzetheir stability using stability region and hybrid system stability analysis techniqueand give the stability region of Three-Tank system. Estimation schemes forcompensating network-induced delay have also been presented: full-state feedbackand output feedback. Simulation results prove the validity of the design method.When NCS transmissions are non-ideal, not only network-induced delay, butalso data packet dropout must be considered. Networks can be viewed as unreliabledata transmission paths, where congestion and node failure can occasionally occur.In Chapter 4, we have modeled control systems with data packet dropout as aswitch closing at a certain rate and give the stable determination method. Then weapply the conclusion to Three-Tank system and show the validity of it. We alsodiscuss the stability of NCSs with multi-packet transmissions.We begin to investigate the control scheme based on the networked three-tanksystem advanced in chapter 2. When we discuss the delay compensation,time-varying delays are regarded as the sum of the mean delays and uncertaindelays. Accordingly, networked control system with delays is modeled as adiscrete-time model with structural uncertainty for its time-varying delays. Basedon the model, a new control law via an iterative linear matrix inequality approach ispresented which can stabilize the closed-loop system. When we discuss data packetdropout compensation scheme, NCS with arbitrary but finite data packet dropout ismodeled as switched system. The corresponding state feedback controller can alsobe constructed by using the feasible solutions of a group of linear matrixinequalities. Based on the data packet-dropout compensator, we also consider thenetwork-induced delay to arrive the delay-packet dropout compensation algorithm.Here, we take the constant delay into account, which often appears in thescheduling networks. The controller can be presented in the similar way as the datapacket-dropout. For every kind of control scheme, we apply it to the three-tankcontrol system and the simulation results show that the control algorithms areeffective.