Forecasting Aquaponic Systems Behaviour With Recurrent Neural Networks Models

Aquaponic systems provide a reliable solution to grow vegetables while cultivating fish (or other aquatic organisms) in controlled environment. The main advantage of these compared with traditional soil-based agriculture and aquaculture installations is the ability produce low water consumption. Aquaponics requires robust control system capable optimizing plant growth ensuring safe operation. To support system, this work explores design process Deep Learning models based on Recurrent Neural Networks forecast one hour pH values small-scale industrial Aquaponics. This implementation guides us through machine learning life-cycle time-series data, i.e. data acquisition, pre-processing, feature engineering, architecture selection, training, model verification.