A Prediction Method of Tunnel Leakage Rate Based on Neural Network

Abstract

A tunnel leakage rate prediction method based on neural network, constructs a tunnel leakage rate prediction model based on convolutional neural network and long short-term memory network, takes three-dimensional laser scanning images of tunnels as the data source, and uses training data sets and test data sets. Train the tunnel leakage rate prediction model, input the prediction data set into the trained tunnel leakage rate prediction model, and obtain the leakage water velocity corresponding to the tunnel image. The invention uses three-dimensional laser technology to detect tunnel diseases, has high operation efficiency and rich results, and builds a convolutional neural network and a long-short-term memory network tunnel leakage water velocity prediction model to predict and analyze the leakage rate.

Description

technical field [0001] The invention relates to the field of tunnel disease detection, in particular to a tunnel leakage rate prediction method based on a convolutional neural network and a long-short-term memory network. Background technique [0002] Due to changes in natural conditions (groundwater, materials, strata, freezing and thawing, etc.), various variation phenomena (such as cracking, dislocation, etc.) occur in the tunnel structure, resulting in the surrounding rock groundwater or surface water directly or indirectly leaking or gushing out. The form enters the tunnel, causing tunnel leakage disease, eroding the tunnel structure, and affecting the normal operation of the tunnel and the use of equipment in the tunnel. [0003] At present, the investigation of leakage diseases in tunnels mainly adopts the methods of manual inspection, photographing and on-site recording, focusing on the area of ​​seepage and the rate of leakage. The workload is large, the efficiency ...

AI Technical Summary

Problems solved by technology

[0006] The above-mentioned patents all use machine learning methods to detect and study tunnel water leakage diseases. Although the automatic identification of water leakage diseases is realized, the rate of water leakage is not predicted and analyzed, and the risk prediction of tunnel structural safety cannot be carried out.

Images