Agricultural Diseases And Pests Forecasting Based On Rs Theory And Fuzzy Reasoning

Rice is one of the major agricultural crops in china, and the occurrence of rice diseases and pests directly affect the national food security, Therefore, providing timely and accurate forecasting information of pest has became a main task in the process of rice production. Diseases and pests prediction is a comprehensive discipline, integrating agriculture、meteorology、mathematics and many other subjects.It monitor the pest situation and predict its future development. This paper developed an intelligent and integrated forecasting system, including data preprocessing,attribute reduction, prediction model, pest forecasting and diagnosis.According to the inaccurate and incomplete characteristics of pest data, we proposed a preprocessing scheme based on the rough set theory, which aims to solve different problems with different solutions. If data is incomplete, we will choose the attribute reduction algorithm under incomplete decision tables. If the value of attributes extents from one element to multiple elements, then we use the attribute reduction under set-valued information system. Moreover, in order to improve the effect of algorithm, this paper improved the attribute reduction algorithm based on single attribute information quantity which measures condition attribute significance from the information quantity of single condition attribute, to reduce the storage space, we improved the algorithm based on linked lists under incomplete decision tables which store the discernibility matrix elements by linked list nodes and improved the algorithm based on attribute significance in Set-valued Information System.The preprocessed data have been used in this system to predict the plant diseases and pests, This study introduced four models, including the development process of forecasting model,the multiple linear regression model, the model based on rough set theory and so on. If the symptoms of rice disease is unidentifiable, the fuzzy reasoning model would be adopted for disease diagnosis.The final purpose of this system is to predict the occurrence period of pest, the occurrence amount of pest, and to translate the result to clear and concise electronic maps and reports for the users.