Optimal Model Order Reduction of Heavy-Duty Gas Turbine Power Plants

Abstract

The real time analysis of gas turbine power plants with higher order system would be dreary and costly. To tackle this difficulty, reduced-order 18.2MW rated heavy-duty gas turbine (5001M) has been attained by using different model reduction methods. Routh Approximation method, Routh stability criteria and Padé Approximation mix method, Clustering technique and Padé Approximation mix method are used to reduce the order of the heavy-duty gas turbine power plant system. The results obtained from the methods are compared and analyzed on MATLAB. It is found that Routh stability criteria and Padé Approximation mix method based reduce order model retains the characteristics of 5001M gas turbine. Afterwards the result is improved for this model of reduced order by optimizing their coefficient using Genetic Algorithm. The results indicate that the Genetic Algorithm based reduced order model obtained has response more closer to that of the heavy-duty gas turbine power plant than any other methods used in this paper.