PCA-SVM-based shield tunneling machine fault data cleaning method and device

Abstract

The invention discloses a PCA-SVM (Principal Component Analysis-Support Vector Machine)-based shield tunneling machine fault data cleaning method and a PCA-SVM-based shield tunneling machine fault data cleaning device. The PCA-SVM-based shield tunneling machine fault data cleaning method comprises the following steps: acquiring each state parameter data set in an operation process of a to-be-diagnosed shield tunneling machine; performing dimension reduction processing on each state parameter data set by utilizing a PCA algorithm And carrying out secondary cleaning on fault data in the state parameters after dimensionality reduction by the PCA algorithm by using an SVM algorithm. According to the method, firstly, a principal component analysis (PCA) algorithm is utilized, the dimensionality of a state parameter data set in the operation process of the shield tunneling machine is reduced, and the complexity of the data set is reduced; and then fault data of different fault types at the fault type classification boundary is found out by using the SVM algorithm principle and the hyperplane equation obtained by classification, and the fault data is eliminated. Through the two times of data cleaning, state parameters capable of correctly reflecting faults in the operation process of the shield tunneling machine are found out, data at fault type classification boundaries are removed, the over-fitting phenomenon is eliminated, and the subsequent fault diagnosis precision of the shield tunneling machine is improved.

Description

technical field [0001] The invention relates to the field of data processing of shield machines, in particular to a method and device for cleaning fault data of shield machines based on PCA-SVM. Background technique [0002] At present, with the continuous development of society, people's demand for complex projects such as railways and water conservancy projects is constantly increasing, and the safety requirements for projects are also constantly improving. Shield machine is a large-scale high-tech comprehensive construction equipment, which is specially used to excavate tunnels and carry out underground construction. It has the advantages of fast excavation, high quality, safety and economy. Therefore, the shield machine has great application potential in tunnels, water conservancy projects and other fields. Doing a good job in the fault diagnosis of the shield machine is of great significance to the project safety, progress and cost. [0003] Generally speaking, in the...

AI Technical Summary

Problems solved by technology

[0004] In view of the above problems, the present invention provides a shield machine fault data cleaning method and device based on PCA-SVM, which can solve the problems of low generalization ability and low diagnostic accuracy of the current shield machine fault diagnosis model

Images