Font Size: a A A

Study On State Prediction Of Networked Control System With Delay

Posted on:2013-03-18Degree:MasterType:Thesis
Country:ChinaCandidate:Y N JiaoFull Text:PDF
GTID:2248330371487136Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Networked Control System (NCS) is a complex system, it is a crossing research filed of Control Theory, Network Communications, and Computer Technique and so on, its design and implementation is restricted by many factors, such as the needs of system, network environment, economic benefits, etc. Therefore, the research of NCS is multi-layered and diversified. In NCS, because of the irregular variation in load, the mode of communication and the sharing of the bandwidth, when the controller, sensor and actuator nodes change data via the network, there will be several negative effects, such as network congestion, multi-path communication of packets, disordering of packets time, packet dropout, retransmission, collision and network interruption and so on. So, when the NCS nodes communicate with each other, the network induced delay is inevitable.In NCS, one of index to measure the QoS is delay, which indicates the time that exchange one bit data between two nodes, the delay will be different when the number and the position of nodes are different.At the first, we review the existing problems and the lately research status of NCS. Then analyze the existing reason and characteristics of time-delay. And its effect on performance of system, the relationship between sampling period and controller parameter is analyzed by Matlab simulation. Aimed at the problem of delay, we summarize the lately research result and the major research methods on NCS.In this paper, we study the modeling of NCS with network delay, and analyze the system performance. We bring up a method of modeling based on the piecewise continuous control signal, which establish a function relationship between discrete state equation and delay interval to provide a new way to the delay research. Based on that model, we design the state observer for prediction referring to Kalman filter. In order to demonstrate its validity, we analyze the system stability by Lyapunov method, and simulate the tracking error between output and reference input by supposing the controlled plant is DC motor, after that the applicable range of our method is given.Finally, we primitively introduce the Truetime toolbox, which is a real-time simulator. Simulating the system performance by aiming at an output feedback NCS, which with PID controller when controlled plant is DC motor, while sensor node is clock-driven, and the controller node and actuator node are event-driven. In this paper, we compare the output of system with reference input under CSMA/CD and Profibus. At last, we replace the code of controller with our model and state observer. By simulation, we find that our method can reduce the impact of delay on NCS(CSMA/CD) effectively.
Keywords/Search Tags:networked control system, system modeling, status observer, prediction, Truetime
PDF Full Text Request
Related items