Cross-Gramian based lower-order controller design

The controller designed for practical systems require less computational complexity and saving the hardware resources, for implementation. Model order reduction techniques play a vital role in achieving these requirements by deriving a reduced order model that accurately approximates the original system. Since the operating frequency range of the plant or controller are critical, therefore, the response of the reduced model should be similar to the original system in the emphasized frequency region. In this paper, we propose a cross-gramian based frequency-weighted model order reduction technique for plant-compensator reduction. The Proposed algorithm does not impose the minimality condition on the original system and produces better results with reduced computational effort as compared to the existing algorithm. The significance of the Proposed technique is shown by numerical examples.