Arowana cultivation water quality monitoring and prediction using autoregressive integrated moving average

Abstract

Decorative fish is a fish that humans keep for amusement. There are many decorative fish that exist in this world, one of them is known as the Arowana fish (Scleropages Formosus). This fish is known around Asia including in Indonesia. However, to ensure the Arowana is living well is not easy. The water quality inside a farm must follow a strict balance. The pH of the water must not exceed or below 7 pH. Meanwhile, the total dissolved solid (TDS) salt must not exceed 1000 parts per million. If the balance collapsed, the Arowana fish will not grow. Thus, the owner must monitor the water to make sure that the water is ideal. There were many approaches including internet of things (IoT) solutions. However, they have weaknesses with prediction. Because of this reason, this study designed pH and TDS monitoring with autoregressive integrated moving average (ARIMA) as the algorithm. To achieve the solution, this study used experiment methodology as the research fundamental from top to bottom. According to the evaluation, this study found that the accuracy of ARIMA model is 98.12% for pH and 98.86% for TDS. On the contrary, the seasonal autoregressive integrated moving average (SARIMA) model has an accuracy of 98.52% for pH and 99.89% for TDS.