상수도 관망은 지하에 매설되어 있어 노후화로 인한 누수, 파손, 부식 등의 문제가 지속적으로 발생하고 있으며, 상수도 관망에서의 노후도 평가와 정비는 수돗물 안전성과 서비스 안정성을 확보하기 위한 중요한 과제 중 하나이다. 본 연구는 상수도 관망에서의 노후도를 효율적으로 평가하기 위해 모든 간접⋅직접평가 인자를 조사하지 않고도 최소한의 주요 평가 인자만으로 실제 직접평가 등급에 근접한 예측 결과를 효율적으로 도출할 수 있는 머신러닝 기반 상태평가 절차와 모델을 제안하였다. 상수관로 상태평가 모델은 랜덤 포레스트와 SHAP 해석기법을 결합한 머신러닝 기반 상태평가 절차를 제안하였으며, 28개 지자체의 강관과 주철관에 대한 간접⋅직접평가 데이터를 대상으로 SHAP-based feature importance 순위가 낮은 평가 인자부터 단계적으로 제거하여 주요 평가 인자를 도출하였다. 결과적으로 강관의 경우 내부 슬라임 분포면적 비율, 수질부식성, 매설연수, 최대 및 외면 부식깊이 비율, 내⋅외면 도장재 손상 비율, 외면 부식 면적 비율, 파손건수, 토양 산화환원전위 및 함수율로 11개의 평가 인자가 주요 인자로 선정되었고, F1 Score는 0.9273이었다. 주철관의 경우 CML 중성화도, 최대 부식깊이 비율, 내부압력, 외부 토압으로 4개의 평가 인자가 주요인자로 선정되었고, F1 Score는 0.6188이었다. 주요 평가 인자만으로도 직접평가 등급에 근접한 예측 결과를 도출할 수 있었으며, 이를 통해 효율적인 상수관로 상태평가를 위한 머신러닝 기법 적용 가능성을 확인할 수 있었다.

In this study, we developed a technology that can measure key evaluation items in the field for various rehabilitation methods and evaluated its performance. The results are as follows. First, when the spray-on lining adhesion strength is measured after drying at high temperature (60℃) for 4 hours and cooling for 2 hours, the results are equivalent to or higher than the adhesion strength measured at room temperature after 48 hours. Therefore, the time required for measuring the adhesion strength in the field can be shortened to 1/8, and it is expected to be helpful in evaluating the condition of the spray-on lining. Second, depending on close or adhesive to existing pipe of the liner of the close-fit lining or CIPP lining, and the grouting condition of the slip lining, the sound frequency generation pattern, the vibration magnitude, and duration using the impact echo showed different characteristics. Therefore, it is thought that it is possible to evaluate liner lifting or grouting failure through analysis of the acoustic frequency, vibration magnitude, and duration. Third, when water penetrates the back of the liner of the close-fit lining or CIPP lining, it was found that the water penetrated between the liner and the existing pipe acts as a couplant, and a signal is repeatedly generated in which ultrasonic energy is transmitted and reflected to the steel pipe after the liner. Therefore, it is judged that it is possible to check whether water has penetrated the back of the liner using ultrasonic waves. Fourth, the liner tensile strength of the close-fit, CIPP lining was compared with the tensile test and the instrumented indentation test, and it was found that the tensile strength was similar. So it was judged that it will be helpful in evaluating the mechanical strength change without the liner specimen in the future.

Due to the sewer induced ground subsidence, there is an increasing demand for the advanced visual inspection technique for the existing sewer pipe structures. This study aim to develop a visual inspection device and real-time transmission system of inspection data with precisely evaluated structural and operational conditions of underground sewer pipe structures. In this paper, a high-precision image capturing system that automatically extracts cracks in the large-diameter sewer pipes and sewage culverts with a diameter of 1,000 mm or more, a real-time gas detection sensor for investigator safety were studied. By analyzing the module technology of the visual inspection device, the concept design for system integration was derived, and the real time transmission system of the inspection result was developed to establish the technical basis for the commercialized device. Also the crack detection test using crack calibration was carried out for the proposed image capturing camera system, and the position accuracy using L1 grade GPS module was tested in this study. The inspection technique of the existing structure condition using the visual inspection device in this study can be effectively used for various structures types and advanced composite structures in the future.