SAFEGUARDING HYDRAULIC STRUCTURES USING MACHINE LEARNING

Abstract

Across the globe, dams are constructed and monitored for the safety of livelihood associated with them and for the purpose they serve. The main purpose of this study is to introduce emergency preparedness for safeguarding hydraulic structures and preventing potential disasters. In the growing world as technologies are advancing, so to use of the concept of machine learning in the field of dam will help to make the studies faster and more accurate. Machine learning-based water level prediction can play a significant role in enhancing the monitoring and management of water levels, allowing authorities to take proactive measures to mitigate risks. Various techniques of deep learning have been taken into account like boosting and bagging using standard algorithms such as Random Forest, Decision Tree, Support Vector Machine, Boosted Regression trees, and long short-term memory (deep learning) models with consideration parameters like precipitation, evaporation, sent out, and inflow. However, in times of emergency, responsibility lies on the dam owner, to make a faster decision-making process, forecasting of water affluence becomes a key responsibility for managing dams and other hydraulic structures. Since early warning systems become most important in flood-prone areas and for the safety of livelihood residing downstream of dams. Therefore, by considering the output from the study, will help authorities to easily vacate the flood-prone areas downstream and for the lives residing around the catchment of the dam’s reservoir. Moreover, from the study, it has been found that the Long Short-term memory-based model gives convincing results compared with other algorithms coefficient of determination of 0.93.