Field Soil Data Clustering Method Based On Fusion Of Feature Distance And Information Entropy

Agriculture is the foundation of national modernization. And it is the key and difficult point of building comprehensive well-off society. At present, China is still in a critical period from traditional agriculture to modern agriculture. Because of the large data quantity of farmland soil information and complex structure, the introduction of data clustering method is applied to effectively mining inner link sensing data, which provides a feasible scheme for filtering the redundant data and optimizing the deployment of sensors.The research in this paper has been integrated into the national science and technology support program of China. Aiming at the problem of large data redundancy and sensor deployment aliasing, field soil data clustering method based on fusion of feature distance and information entropy is studied. The main research contents are as follows:1、Data efficiency analysis and clustering algorithm support for field operations. The characteristics of air and soil in the field work were studied. This paper offered the research on data sensing ability of Bluetooth, RF, Zigbee and other sensor networks in field operation, and data analyzed ability of BIRCH, STING, DENCLUE.2、Field soil data clustering method based on fusion of feature distance and information entropy. Order to making the clustering algorithm suitable for field soil related data and applied to the sensor deployment, selecting the initial clustering is optimized by the Euclidean distance control factor based on aggregation density sparse degree and Multi-objective clustering approach based on dynamic cumulative entropy in this paper. And then I put forward the non-uniform sparse data clustering cascade algorithm based on dynamic cumulative entropy. The algorithm is applied to optimize the sensor deployment scheme, and put forward the field soil data clustering method based on fusion of feature distance and information entropy.3. This paper builds the field moisture prediction system to test the performance of the algorithm. The data collected from changing Ge of He Nan experiment field is used to cluster, and the sensor deployment is guided according to the clustering results. To predict soil moisture using the initial field data and optimization field data. The result shows that prediction of soil moisture based on the algorithm about the deployment has less errors, so the algorithm is feasible and effective.