An Expandable High-Precision Modeling Method for Gas Turbine

For the high precision requirement of the model for gas turbine,an expandable modeling method is presented by using the actual measurement operation data.The influence of the environmental factor and the measurement error is eliminated,and the normal operation data is obtained by using the BP neural network.Based on the low precision model of the design state,a modified factor vector is imported,the root mean squared error is target function,a updating algorithm is given,a new adaptive calculation model of the component performance corrector is constructed,and the thermodynamic model which reflects the current normal characteristics of the gas turbine is achieved.The research results shows that the method have high precision and good practicability.