Comparative Experimental Study on the Evaluation of the Unit-water Content of Mortar According to the Structure of the Deep Learning Model

딥러닝 모델 구조에 따른 모르타르의 단위수량 평가에 대한 비교 실험 연구

The unit-water content of concrete is one of the important factors in determining the quality of concrete and is directly related to the durability of the construction structure, and the current method of measuring the unit-water content of concrete is applied by the Air Meta Act and the Electrostatic Capacity Act. However, there are complex and time-consuming problems with measurement methods. Therefore, high frequency moisture sensor was used for quick and high measurement, and unit-water content of mortar was evaluated through machine running and deep running based on measurement big data. The multi-input deep learning model is as accurate as 24.25% higher than the OLS linear regression model, which shows that deep learning can more effectively identify the nonlinear relationship between high-frequency moisture sensor data and unit quantity than linear regression.