Research On Data Fusion Method Of Crop Growth Environmental Information

Using the method of data fusion to obtain information on crop growth environment, it canimprove the detection accuracy of the system. Redundant information between different sensors,which makes the system has better fault tolerance capability, provide accurate data to theenvironmental control systems of crop growth.With the rapid development of science and technology and the decrease of arable landmakes the plant factory has been greatly developed. Especially in the21century, with theimprovement of living quality, people growing awareness of food safety, green, organic fooddemand is increasing, especially the reef islands, cold high altitude outposts and submarinesoldiers, do not eat fresh vegetables all the year round,"family plant factory" concept came intobeing. On the basis family plant factory, design a smart plant growth cabinet factory. The mainresearch results were as follows:Design of a fusion model, the local fusion uses a probabilistic algorithm in time, the globalfusion mainly uses the fusion of BP neural algorithm to the three main environmental factors ofthe crop growth: temperature, humidity and CO2concentration.For multi-sensor data acquisition and preprocessing, the preprocessing is mainlyto eliminate divorced value, using the distribution display method for divorce values remove,which can eliminate half of divorced value, and this method is not affected by the distribution ofthe data, robustness.The research of this paper is data fusion of the crop growth environment information in theintelligent plant growth cabinet, the main use of Bayesian data fusion method for data fusion tothe growth cabinet temperature, increasing temperature acquisition precision, but also caneffectively eliminates errors due to sensor failure brought. And using the real-time changes of BPneural network data fusion method of environmental factors simulation, the results showed it canquickly track the changes. Control method using fuzzy PID to control, guidance on the use ofknown expert experience knowledge, on the growth of plants in the system’s multi-sensorstemperature, humidity, CO2concentration information were fusion, the final results provideaccurate data to the modeling and control for crop growth environment.