<![CDATA[The Gajah Mungkur Reservoir is one in all the most largest dams or reservoirs in Java which is categorized as a multipurpose reservoir. With various benefits, it is necessary to forecast the inflow discharge to avoid excess or shortage of water in the reservoir as well as errors in water disposal. A common mistake is the release of water which can cause flooding in areas lower than the reservoir. Discharge forecasting can also be used to plan water allocations such as power generation and irrigation. Changes in inflow discharge that occur are always fluctuating, so from these problems forecasting inflow discharge is needed to overcome the large amount of water discharge that comes out. The data used is inflow discharge from January 2009 to December 2019 and the method used is Extreme Learning Machine (ELM) because it has fast learning speed and good generalization. The test results obtained are the optimal number of features as many as 7, the optimal number of hidden neurons as many as 9, with the percentage of training data 80% and 20% of the test data producing RMSE 28.7303, MAD 21.8002 with a runtime of 0.0272s. With an RMSE value that is far from zero, the error rate obtained is high and bad. And also with a runtime that is in seconds or less than seconds, this research also confirms that ELM has advantages in fast learning speed.]]>
Copyrights © 2021
