Smart solar powered hydroponics system using internet of things and fuzzy association rule mining

Abstract

Hydroponics is an alternative to limited space agriculture by involving electrical energy for irrigating nutrients for plants. Apart from using an AC power source, hydroponics can also use a DC power source which makes it more portable and can be placed in a location far from the PLN(State Electricity Company) electricity source. DC power sources also have the potential to be supplied by renewable energy sources such as solar power directly without AC / DC conversion. However, solar energy sources have a weakness in the influence of the weather. The study proposes the implementation of the Internet of Things (IoT) to classify the condition of energy sources in its implementation of solar energy powered hydroponics system. The research involves the development of IoT devices to collect energy data generated by solar cells and stored in batteries and environmental conditions, namely light intensity, temperature, and humidity. The proposed IoT system classifies conditions based on the remaining energy in the battery and current weather conditions using the Fuzzy Association Rule Mining (FARM) algorithm. The system also considers the remaining hours of the day, because at night solar energy cannot be produced. The simulation result shows the developed solar-powered hydroponics system capable to keep supplying energy for 24 hours with three 10wp solar cells and a 7Ah battery for the deep flow technique hydroponics with the 24 pot with 1-meter stand and 20watt water pumps. The IoT system capable to collect data and the proposed regression algorithm FARM is capable to do the regression with low error rates (0.06 mean absolute error and 0.24 mean square error) with three days of data. The FARM also capable to extract environmental data that affect the energy generation and usage with a high average of supports (0.29) and confidence (0.32).