Five-hole pneumatic probe calibration method based on two-stage artificial neural network

Abstract

The invention discloses a five-hole pneumatic probe calibration technical method based on a two-stage artificial neural network. The calibration parameters of a five-hole pneumatic porous probe are predicated through employing the artificial neural network, a first reference artificial neural network ANN1 and a second reference artificial neural network ANN2 are constructed, the input feature parameters are the feature parameters kMa, kalpha and kgamma obtained through the measurement of a pneumatic probe, calibration parameters output by the ANN1 are error-free predicted values of Mact, alpha ct, gamma ct, kpt and kps, calibration parameters output by the ANN2 are error predicted values of Mact, alpha ct, gamma ct, kpt and kps, calibration data measured by the five-hole pneumatic probe are used for training and testing the ANN1 and the ANN2 respectively, and system analysis is carried out through different combinations of the number of network hidden layers, a training algorithm, an activation function and the number of neurons, to find an optimal network parameter solution compared with the reference network; and the optimal network structure is used to realize two-stage calibration prediction, and the prediction sum of the two neural networks is a prediction calibration value. The method has the advantages of being easy to implement, good in generalization, high in testing precision and the like.

Description

Technical field [0001] The present invention relates to the field of gas turbine engine impeller mechanical refined test technology, and it is related to a five-hole pneumatic probe calibration method, and specifically, a five-hole pneumatic probe standard method based on artificial double-stage neural network, for impeller mechanical measurement pneumatic The high-precision wide range of probes is calibrated to provide advanced testing for impeller mechanical refinement design. Background technique [0002] Modern gas turbine engine must optimize the impeller mechanical blade performance under the design phase in order to obtain as high efficiency as possible. In order to design the optimal blade gas layout, the blade is required to measure the flow of the flow path in detail to understand the boundary layer, secondary flow development, and the measurement accuracy of the porous air probe is critical, which requires high-precision calibration of the probe. . [0003] The probe i...

AI Technical Summary

Problems solved by technology

The accuracy of this method is adequate for the central sector where the flow direction is approximately aligned with the probe axis, but is poor for other locations outside the sector, making the angle error outside the sector significantly increased and not satisfying the turbomachinery interior. Measurement of the position of the equal major turning angle of the secondary flow

Images