In this paper, characteristics in all working conditions of some selected turbine are acquired based on the information of several given operating points and an improved ISO temperature is proposed to calculate expander power of some turbine. An object-oriented approach is used to code the turbine calculation program. Firstly, characteristics in the same rotational speeds with given variable operating points are calculated by Stodola-Flügel equation and PSO based neural networks are used to predict more characteristics. Then, the turbine thermodynamic process is divided into blending of gas and cooling air and gas expansion. The blending process is calculated by improved ISO temperature method which adopts PSO method to calculate the optimal value of recovery coefficient of gas pressure and efficiency coefficient of the cooling air enthalpy. Finally, the above calculation procedure of the turbine is coded by an object-oriented approach and it is approved to be of high accuracy. Such a calculation process proposed in this article meets the requirement of turbine performance calculation with cooling air.

This content is only available via PDF.
You do not currently have access to this content.