Rough Set Theory In Agricultural Disasters

Agriculture is the basic industry in China. There are many types of agriculturaldisaster，such as: droughts, floods, landslides, mudslides, forest pests and many other types.Three types of disasters, which seriously restrict the development of agriculture, areaddressed in this paper. They are droughts, floods and forest pests. We introduce theirstatus, feature and other basic situation. Agriculture disaster has become a main factorrestricting agricultural development in China. So it is necessary to do a deep analysis ofagricultural natural disaster, analysis of the effect factors of disaster, forecast of agriculturaldisaster.With the development of information technology, modern agricultural work becomesmore informationization, intelligent and networking. This paper applies rough set theory toanalysing and forecasting factors of agricultural disasters, hoping to provide theoreticalmethod for the analysis of agricultural disasters. It plays an important role in predictionand control of agricultural disasters, reducing disaster influence of agriculture, promotingagricultural modernization, informationization.In order to explore fast and accurately useful information from the complex data, theattribute reduction theory is introduced in this thesis. We remove the unnecessary factorsfrom the effect factors and retain the major role. For example, in analysing of effect factorson forest diseases and pests, we select four factors as rainfall, degrees accumulatedtemperature more than0, degrees accumulated temperature more than10and sunshine.After attribute reduction, we remove degrees accumulated temperature more than0,reducing the burden and complexity of following analysis.The collected data may be incomplete in the process of disaster data collection. Suchas, when using remote sensor access to remote sensing data, we may encounter sensorfailure or data transmission barriers or other issues. Sometimes, information cannot beobtained in time and accuratly. We use the incomplete information system to deal withthese incomplete data and achieve the desired results. We hope the conclusions of thisstudy can provide theoretical guidance and technical support for the prediction andprevention of agricultural disasters.