Large high-speed rotary equipment depolarization and declination method based on deep belief neural network

Abstract

The invention discloses a deep belief neural network-based large-scale high-speed rotary equipment depolarization and inclination elimination method. The method comprises the steps that 1, influence factors of all machining errors are analyzed; 2, evaluating the real machining error of the large high-speed rotary equipment, and obtaining the measurement data of the influence factors correspondingto the machining error when the center and the inclination are not adjusted; 3, taking the normalized data of the influence factors of the processing errors as input of the networks, taking the processing errors as output of the networks, and dividing the data into a training set and a test set according to a certain proportion; and step 4, respectively constructing a deep confidence prediction neural network of each processing error, realizing automatic depolarization and deinclination, and performing verification through a test set. According to a traditional machining error measurement method for single-stage large high-speed rotary equipment, a complex aligning and inclination adjusting process needs to be carried out, and the invention provides a large high-speed rotary equipment depolarization and inclination eliminating method based on a deep belief neural network by utilizing a powerful data internal characteristic mechanism of deep learning.

Description

technical field [0001] The invention belongs to the technical field of processing errors of large-scale high-speed rotary equipment, and specifically relates to a method for eliminating deflection and tilting of large-scale high-speed rotary equipment based on a deep confidence neural network. Background technique [0002] Multi-stage equipment is assembled by stacking single-stage large-scale high-speed rotary equipment. In order to improve the assembly quality and first-time assembly pass rate of multi-stage equipment, it is necessary to obtain the machining error of each single-stage large-scale high-speed rotary equipment. At present, the machining errors of single-stage large-scale high-speed rotary equipment are passed such as figure 1 The two-column aeroengine ultra-precision assembly measuring instrument shown is used for measurement. However, due to the limitations of the instrument error evaluation method, it is impossible to accurately compensate the eccentricity...

AI Technical Summary

Problems solved by technology

Therefore, the existing methods for evaluating machining errors of large-scale high-speed rotary equipment require a cumbersome and time-consuming process of centering and tilting, which greatly reduces the assembly efficiency of multi-level equipment

Images