Shield attitude prediction method based on self-encoding characteristics and deep learning regression model

Abstract

The invention discloses a shield attitude prediction method based on self-encoding characteristics and a deep learning regression model. The method mainly comprises the following steps: (1) collectingdata; (2) carrying out pretreatment; (3) generating a training sample and a test sample; (4) constructing an auto-encoder and constructing a deep learning regression model; (5) extracting data features; (6) training a deep learning regression model; and (7) predicting the shield attitude. According to the method, the equipment parameter information in the tunneling process of the shield tunnelingmachine is combined, the shield posture can be reflected through the parameter information, the tedious solving process is avoided, the response time to shield posture changes is shortened, the efficient data analysis capacity is achieved, reference and adjustment bases can be provided for construction operators, and the construction quality is guaranteed.

Description

technical field [0001] The invention belongs to the technical field of machinery, and further relates to a shield attitude prediction method based on self-encoded features and a deep learning regression model in the field of shield machine measurement technology. The invention can be used in the construction process of the shield tunneling machine to predict the attitude that the real traveling line of the shield tunneling axis line does not completely coincide with the design axis line. Background technique [0002] During the propulsion process of the shield machine, sometimes it will pass through complex geological layers, and the cutter head will wear out, which will cause uneven force, which will lead to changes in the posture of the shield machine, that is, the actual line of travel of the actual shield axis and the The design axes do not exactly coincide. At present, in the shield construction process, the automatic guidance method is mainly used to calculate the dev...

AI Technical Summary

Problems solved by technology

Although this method uses sensors to measure the included angle, even if the automatic guidance system of the shield machine fails, using this method to calculate the included angle can still ensure the accuracy of the included angle. However, this method still has the disadvantage that the included angle The calculation formula is derived on the basis of the ideal geometric relationship. The harsh construction environment of the shield has a great influence on the geometric relationship, and small position changes will cause calculation errors. And this method is to determine the shield tunnel through measurement and calculation after excavation. The specific position of the shield machine has a certain lag time for following the posture of the shield machine.

However, the shortcomings of this method are that the process of calculating the attitude angle is too cumbersome due to the complexity of the measurement system composed of the gyroscope and the inclinometer, and the static accuracy of the gyroscope's perception of the orientation is low, which will cause problems in dynamic environments. Larger drifts can only play a limited role, causing the system to fail to work for a long time, and in the face of complex construction environments, it is impossible to analyze the state trend of the shield

Images