Method for Determining Bridge Health Level Based on Semi-supervised Error Correction Learning

Abstract

The invention discloses a bridge health level determination method based on semi-supervised error correction learning, which separates the original label and the feature of the original bridge label data; utilizes a non-supervised learning method to perform clustering operations on data features; and performs clustering operations on The resulting clustered data with category information is labeled with the original bridge label data; according to the label unification results, the data with the same cluster label and the original label in the training set is used as a high-confidence bridge data set, and the data with different labels As a low-confidence data set; according to the preset iteration parameter k, through the supervised learning method, use the high-confidence bridge data set to train the classifier, and use the trained classifier to iterate the low-confidence bridge data set to update the high-confidence degree bridge data set; use the updated high-confidence bridge data set to train again to obtain the final classifier; use the final classifier to input the bridge data of the bridge to obtain the bridge health level.

Description

technical field [0001] The invention belongs to the technical field of bridge health level assessment, in particular to a bridge health level determination method based on semi-supervised error correction learning. Background technique [0002] In the context of a large number of roads and bridges in our country, the maintenance and management of roads and bridges is a huge task. In line with the development direction of "standardized evaluation and scientific decision-making" of roads and bridges maintenance management in my country, we will further strengthen the management of roads and bridges maintenance and improve the quality of roads and bridges. It is of great significance to the economic development and social stability of our country to improve the level of maintenance management and ensure the quality of highway and bridge maintenance. The assessment of the overall technical status of the bridge, that is, the health level of the bridge, will directly determine the ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a bridge health level determination method based on semi-supervised error correction learning to solve the problem that the accuracy of the evaluation results is not accurate due to bridge data with uncertain labels in the current supervised learning bridge health level determination method. High, affecting the maintenance and management of highway bridges

Images