Integration Agricultural Knowledge and Internet of Things for Multi-Agent Deficit Irrigation Control

Abstract

Technology related to the Internet of Things and Big Data is developing rapidly, favoring agricultural management and cultivation. External environmental parameters affect crop growth. Irrigation operations involve multi-conditional decisions. Artificial intelligence and image processing technology are used to control watering, process environmental data, and use the processed data are then used to train models to predict soil moisture. The results thus obtained are used to divide soil into that with high, normal, and low moistures. A camera commonly used to capture images of soil, which are imported into a convolutional neural network for feature extraction. Next, the category of soil is predicted following deep learning. The results of the two predictions are sent to a fuzzy control system and the soil moisture interval is defined based on the category of crop that is being planted in it. The rule base is constructed using the defined parameters, and determines the judgment rule from the agricultural professional, And external conditions lead to the fuzzy controller. Defuzzified results are imported into controlled a control device. This work uses the IoT to serialize information from the fields in a farm and uses artificial intelligence techniques to analyze the state of plants. Image recognition technology is to provide more precise and comprehensive control to improve the accuracy of irrigation.