DAILY FORECASTING OF DAM WATER LEVELS USING MACHINE LEARNING

The design and management of reservoirs are crucial towards the improvement of hydrological fields subsequently leading to better Integrated Water Resources Management (IWRM). Different forecasting models used in designing and managing dams have been developed recently. This report paper proposes a time-series forecasting model formed on the basis of assessing the missing values. This is followed by different variable selection to determination to gauge the reservoir's water level. The investigation gathered data from the Klang Gates Dam Reservoir as well as daily rainfall data. The two sets of data are consolidated into a coordinated set formed on the basis of directing it as a research dataset. Furthermore, the proposed model applies a Time Series (TS) Regression Model to develop the forecasting model of the reservoir's water level. The tried results demonstrate that when the Time Series Regression forecasting model is used to select variables with complete variables, it gives a better forecast result than the SVM model