Intelligent Agricultural Forecasting System Based on Wireless Sensor

In order to provide more precise decision-makings for agricultural producers, an intelligent forecasting system is proposed for agricultural applications in this paper, which processes data stream collected from agricultural environment applying WSN (Wireless Sensor Network) technology. Sensor data stream is different from other traditional stream, the former is characterized by real-time, sequential, missing and imprecise. In the proposed system, the sliding window is used to model the sensor data stream, some fuzzy rules are constructed based on expert knowledge, and the fuzzy inference is used to collect different environmental parameter data stream to provide intelligent services for guiding irrigation control, disease prevention or other applications. The proposed forecasting system takes the sensor data stream as research objects and provides quantitative prediction services instead of vague judgments, which is the most useful for agricultural manager.