Statistical model identification for cogeneration system using statistical analysis

Cogeneration system is nonlinear, wide range of operating regime, and multivariate. The generating efficiency and thermoelectric conversion coefficient of turbine generator are affected by many factors such as load variation, extracted steam change, and unstable supplied power etc. In this paper, a statistical model identification algorithm has been proposed to build the model of generating efficiency. Firstly, key variables were selected by using stepwise regression method. Then, statistical models were identified in different operating regimes. The experimental results demonstrated the performance of the presented algorithm. The field engineer can take the developed model as a reference and determine the operating regime of cogeneration system. As a result, the power management is improved and the operating cost is reduced.