Structure key area control parameter early warning method based on recurrent neural network

Abstract

The invention discloses a structure key area control parameter early warning method based on a recurrent neural network, and the method comprises the following steps: arranging sensors in a key area of a structure, and monitoring the displacement and strain parameters of the key area of the structure; converting data measured by the arranged sensors into structure control parameters; carrying outnormalization preprocessing on the obtained structure control parameters group by group; constructing a recurrent neural network structure; respectively training each layer of weight of the recurrentneural network by using the obtained multiple groups of data at different time intervals to obtain a group of models; monitoring data of the structure in a recent period of time is taken and subjectedto normalization and down-sampling processing to serve as initial input of the network for prediction. The future health state of the structure can be predicted, compared with an existing structure health monitoring method, more reaction time can be gained for a maintenance unit, the safety of the structure is maintained more effectively, and casualties and property losses caused by sudden damageof the structure are greatly reduced.

Description

technical field [0001] The invention relates to the technical field of interaction between civil engineering and artificial intelligence, in particular to a method for early warning of control parameters in key structural areas based on a cyclic neural network. Background technique [0002] With the rapid development of my country's infrastructure construction in recent years, the civil engineering industry has developed rapidly, and a large number of bridges, tunnels, and houses have been constructed. After long-term use of these civil engineering structures, structural deformation, concrete cracking and other damage will occur, which will affect the actual use of the structure. safety in the process. The current structural maintenance methods include structural detection and monitoring. The structural detection and evaluation cycle is relatively long. What the structural maintenance unit provides is the historical health status of the structure. The structural health monito...

AI Technical Summary

Problems solved by technology

However, for some structures that are in poor health but are still in use, or some particularly important structures, only knowing the real-time monitoring status of the structure cannot reserve sufficient response time for the maintenance unit to avoid casualties and injuries caused by sudden structural damage. property loss

Images