PREDICTIVE MODELLING OF PH LEVELS FOR OPTIMIZING WATER QUALITY IN SHRIMP FARMING

Water quality is a critical factor in shrimp farming, directly influencing the growth, reproduction, and survival of shrimp. pH is one of the key parameters that affect water quality, with deviations from the optimal range (5.5–8.5) leading to stress, weakened immune responses, and potential infections in shrimp. This research presents the development of an automated pH monitoring and forecasting system aimed at improving water quality management in shrimp farms. The system uses a moving average algorithm to predict future pH levels based on real-time data collected by a pH sensor. The predicted and real-time values are transmitted to a cloud database, and farmers receive alerts via the Line application if pH levels deviate from the acceptable range. The system's performance was evaluated through six experiments, using different data collection intervals and durations. The most accurate forecasting results were achieved with 10-minute data collection intervals over a 2-hour period, yielding a mean squared error (MSE) of 0.003050 and a root mean square error (RMSE) of 0.038628. The system also demonstrated its ability to send real-time alerts to the farmer, ensuring prompt corrective action in the event of critical pH values.