Several of the practical industrial control processes are multivariable processes. Due to the relation amid the variables (interaction), delay in the loops, it is very intricate to design a controller directly for these processes. So first, the interaction of the variables is analyzed using Relative Normalized Gain Array (RNGA), which considers the time constant, static gain and delay time of the processes. Based on the effect of RNGA, relative gain array (RGA) and NI, the pair (control configuration) of variables to be controlled by decentralized control is selected. The equivalent transfer function (ETF) of the process model is estimated as first order process with delay using the corresponding elements in the Relative gain array and Relative average residence time array (RARTA) of the processes. Secondly, a decentralized Proportional- Integral (PI) controller is designed for each ETF simply using frequency response specifications. Finally, the performance and robustness of the algorithm is comparing with existing related approaches to validate the effectiveness of the projected algorithm.
Control Configuration Selection and Controller Design for Multivariable Processes Using Normalized Gain
2015-01-19
oai:zenodo.org:1337867
International Journal of Electrical, Electronic and Communication Sciences 7.0(10)
Article (Journal)
Electronic Resource
English
DDC: | 629 |
Flight controller design using multivariable loop shaping
British Library Conference Proceedings | 1994
|An Algorithmic Approach for Multivariable Digital Controller Design
British Library Online Contents | 1996
|Distributed Fuzzy Control of Multivariable Processes
British Library Conference Proceedings | 1995
|Multivariable Aircraft Engine Controller Design Using an Optimal Loop Shaping Approach
SAE Technical Papers | 2011
|