In the project, we focused on both the design and the theoretical analysis of the Networked control systems in various scenarios. The main contributions of the project are as follows.
1. The effect of delay on the stability of a NCS was implemented on Matlab. Controllers to stabilize the system in the case of integral delay, random delay with & without limits, multi sampling, noise & constant delay were designed & implemented on Matlab
2. The effect of delay was studied based on a new model with a clock-driven controller. It was shown that system can be modeled as a switched system and the effect of dwell time on the stability was studied. The system was found to be stable for a dwell time 1/5th that of the plant sampling period. It was also seen that the system could tolerate a constant delay of up to 20 sampling periods, random delay of up to 6 sampling periods.
3. In the case of random delay with limits, GPC & Neural network functionality was compared. It was shown that Neural Network performs better. This analysis & implementation on MATLAB included parameter uncertainty also.
4. For non linear systems, modelling based on TS-Fuzzy method was implemented and the effect of delay was studied. A controller was designed to minimise the effect of delay on stability.
5. The effect of packet drop on the stability of NCS was studied & implemented on Matlab. An LMI & lyapunav stability approach was used to design a controller to stabilize such a system. The controller designed could stabilize a system with upto 75% packet drop. However the system could tolerate only a 1% variation in parameter. Thus proving the algorithm not to be robust.
6. The effect of random packet drop on the stability of NCS was studied & a Robust LMI based controller capable of establishing stability under random packet dropout, parameter uncertainty & disturbance was designed & implemented on Matlab.
7. The effect of alternating packet drop and delay on the stability of a system was implemented. A controller was designed that stabilised the system even in such a scenario using LMI.
8. The effect of both time delay & packet drop on NCS was studied & an LMI based PID controller was designed & implemented to provide exponential stability to the system under 50% packet dropout & time delay less than 1 sampling period. However the system could tolerate up to 5% parameter uncertainty.
9. In the case of delay & packet drop the H∞ output tracking performance was implemented on Matlab, the design of a controller in such a scenario was implemented & tracking was verified. Further the system was found withstand 20% parameter uncertainty
10. A Real Time NCS was simulated using TrueTime toolbox, to establish results under the following constraints- Data Rate, Frame size, Scheduling Policy, Type of Network, Data Packet loss, Distance between Controller & Plant
References of our project are the following papers
1. Qixin Zhu, Hongli Liu, Juan Jiang, Yanping Wang, Stabilization of networked control systems with long varying time delay based on two step transformation, IEEE, 2008
2. Jin Wu, Fei-Qi Deng, Jing-Guang Gao, Modeling and Stability of Long Random delay Networked Control Systems, Proc of fourth international conference on Machine Learning & Cybernetics, 18-21, Aug 2005
3. Tian-Kun Wang, Li-Hui Zhou, Pu Han, Qian Zhang, Complete compensation for the time delay in networked control system based on GPC and BP neural network, Proc of sixth international conference on Machine Learning & Cybernetics, 19-22, Aug 2007
Zuo You, Yin, Huaguang Zhang, Robust tolerant control for a class of nonlinear systems with uncertainities and input time delays based on T-S Fuzzy Model,
4. Mei Yu Long, Wang Tianguang Chu, Guangming Xie, Stabilization of Networked Control Systems with Data Packet Dropout and Network Delays via Switching System Approach, 43rd IEEE Conference, Dec 2004
5. Zidong Wang, Fuwen Yang, Robust H∞ control for NCS with Random Packet Losses, IEEE Trans, Vol 37, No. 4, Aug 2007
6. Xiandong Dang, Quingling Zhang, Stability & PID control for Networked Control Systems, Proc of 8th world congress on intelligent control & automation.
7. Huijun Gao and Tongwen Chen, Network based H∞ output tracking control, IEEE Trans, Vol 53, No.3, Apr 2008
8. Truetime toolbox developed by Lund University
