Rough Set Theory In The Diagnosis Of Common Citrus Diseases

Citrus fruits are the most important fruits in the world,it plays an important role in the economy and trade at home and abroad.However,the occurrence of disasters still threatens the healthy development of the citrus industry.The Diagnosis and prediction of disease prevention and control is the premise of Citrus Diseases,only the timely and accurate diagnosis of the occurrence of the disease type and predict disease,can we take preventive measures to against,thereby reducing the cost of planting,improve the yield and quality of citrus fruit,maintain the market competitiveness of citrus industry and increasing farmer income.We selects three common plant diseases in citrus: scab(Sphaceloma fawcetti Jenk),Citrus canker(Xanthomonas Campestris pv.citri),and Citrus anthracnose(Colletotrichum gloeosporioids Penz.)as main research subjects,introduced the occurrence and control methods of the three diseases mentioned above,and used the knowledge of Rough set theory to diagnose the extent of the above-mentioned diseases in Jiangxi Province in the past 7 years,Compared with the degree of disease in the existing data,the result is accurate;then predicted the degree of occurrence of the above three kinds of diseases in 2018.And the drawing of the result of the experiment is finished.The specific work is as follows:To quickly and accurately mine useful information and improve computational efficiency,this article has introduced relevant theoretical knowledge of attribute reduction,analyzed the impact of each factor on the degree of occurrence of diseases,and removed those factors that had no or lesser impact,and retained the main influencing factors.For example,the listed four possible causes of disease occurrence: average annual temperature,average annual minimum temperature,number of clear days(sun hours),the number of rainy days(rainfall),“the number of clear days” the degree of importance of this influencing factor relative to the incidence of "scar lesions" is 0,which means that this attribute has no influence on the occurrence of scar lesions,and this attribute may not be considered in the process of diagnosis.For data sets with missing data,a solution is giving through the combination of incomplete information systems theory and relevant knowledge of probabilistic Rough sets.For example,the“<-5℃ longest continuous time” property value of the “Ningdu-Shishang”orchard was missing.Under the circumstances,we can still diagnose the "heavier" degree of frost damage in the orchard,the diagnostic results are consistent with the absence of data missing.The diagnosis and application of the Rough set theory to citrus diseases makes the diagnosis of citrus disease more informational and intelligent,and hopes to provide an effective theoretical basis for the study of citrus diseases.