A network application for modeling a centrifugal compressor performance map

10th International Conference on Compressors and their Systems

Nikiforov, A.^1 ; Popova, D.^1 ; Soldatova, K.^2

Department of Mechanization, Smolensk State Agricultural Academy, Bolshaja Sovetskaja, 10/2, Smolensk

214000, Russia^1

RandD Laboratory Gas Dynamics of Turbo Machines, Peter the Great St.Petersburg Polytechnic University, Polytechnical st. 29, St.Petersburg, Russia^2

关键词: Aero-dynamic performance;    Centrifugal compressor stage;    Compressor performance;    Flow coefficients;    Friction coefficients;    Mean squared error;    Network applications;    Vaneless diffuser;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/232/1/012046/pdf
DOI  :  10.1088/1757-899X/232/1/012046

来源: IOP

【 摘 要 】

The approximation of aerodynamic performance of a centrifugal compressor stage and vaneless diffuser by neural networks is presented. Advantages, difficulties and specific features of the method are described. An example of a neural network and its structure is shown. The performances in terms of efficiency, pressure ratio and work coefficient of 39 model stages within the range of flow coefficient from 0.01 to 0.08 were modeled with mean squared error 1.5 %. In addition, the loss and friction coefficients of vaneless diffusers of relative widths 0.014-0.10 are modeled with mean squared error 2.45 %.