The Society of Convergence Knowledge TransactionsISSN:2287-8920(Print)융복합지식학회

Recently smart agriculture is improving the agricultural competitiveness of productivity and product quality throughout all stages of the agricultural value chain (Value Chain) such as seeds, production, harvesting, distribution, and consumption by incorporating communication technologies (ICTs) such as the Internet of Things (IoT), artificial intelligence (AI), big data, and robots. However, abnormal climatic conditions such as extreme weather events, changes in temperature and precipitation patterns have a serious impact on agriculture, causing various agricultural problems. In this paper, we propose an AI-based efficient smart agricultural management technique to optimize autonomous sensor information so that it can actively respond to abnormal climate changes. The proposed technique employs an asymmetric processing method to prepare the data set for machine learning and AI models, ensuring balanced autonomous sensor information. The proposed technique ensures continuous learning of large-capacity real-time autonomous sensor information using the DNN model to cope with abnormal climatic conditions (water shortage, extreme temperature, seasonal change, humidity, etc.). In addition, the proposed technique plays the role of a distributed record storage system by grouping and processing real-time large-capacity autonomous sensor information into a blockchain so that the DNN service can be handled efficiently. The proposed technique may maintain and verify synchronization between autonomous sensor information by accumulating probability values at regular intervals for weights of autonomous sensor information.