Application of the artifical neyral networks theory methods for generalization and representation charachteristics of aviation gas turbine engines axial compressors

The methods of generalization and representation characteristics of gas turbine axial compressor using the methods of neural network approximation.An algorithm for experimental performance generalization in mathematical packages on methods of artificial neural networks are submitted. We compare the results of approximation and interpolation characteristics by neural networks based on multi-layer perceptron and based on radial basis functions for different values of the design compressor pressure ratio. The expediency of applying the methods of generalization and representation compressor characteristics using a two-layer perceptron are shown.