Application of decision trees to the determination of the year-end level of a carryover storage reservoir based on the iterative dichotomizer 3

This study presents a new method for deriving an operational model (rules) for the carryover storage reservoir water level at the end of year (WLE), which considered all of the operational information using an optimal model. To identify carryover storage reservoir WLE operation rules for a cascade reservoir system, the best approach is to exploit large databases using a data mining tool such as decision trees. Thus, we applied decision trees based on the iterative dichotomizer 3 (ID3) to the Wu Jiang cascade reservoir system to derive the carryover storage reservoir WLE operation rules. The rules derived can be used to guide optimal real-time operation in the long term. The results demonstrated that the solution derived using tree-based rules performed better than the existing design rules in terms of the power generation benefit, which satisfied the output of the cascade reservoir system.