Sampled-data H2 and H∞ control under non-uniform sampling and variable communication delay – Part2

Conventional sampled-data control theory assumes a periodic operation of sampling with disregard of variable communication delay. However, there are many emerging engineering applications where it is not possible to periodically sample the measurements and update control input under the influence of random delay as mostly occur in Networked Control Systems (NCS). NCS being convenient in terms of reliability, scalability and usage of less wiring over traditional control is preferred. However, some issues like limited bandwidth, communication delays and packet dropout might occur. In this talk we will focus on designing the controllers that will compensate the issues, for this strictly causal sampled data controller is designed by help of lifting technique which translates the LTV system to LTI sampled data system having controller based on sensor and actuator side part. Further, it is parametrized with help of Youla parameterization to obtain the multiple controllers. Once the SDDTS controllers are obtained, discrete-time predictor is designed based on the given random delay measurement bounds, its stability range is obtained by LMI stability test. After having SDDTS controllers for variable delays in feedback, Network friendly implemeSamntation is proposed and implementable. Finally, sampled-data H2 and H∞ control optimization problems are addressed for derived family of controllers.

Co-sponsored by: Cyber-Physical Systems Laboratory

Speaker(s): Mr. Salman Ahmad,

Cyber-Physical Systems Laboratory, Islamabad, Islamabad Capital Territory, Pakistan