论文部分内容阅读
利用模糊数学进行植物病害流行学方面的研究可能具有较大的潜力。在已有的多组可靠数据上,把与流行事件有较大关系的因子作为论域的元素,并建立隶属函数,按实际情况将流行事件划分为若干等级并组建各等级的代表子集,然后,利用已建立的隶属函数和代表子集,利用模型识别方法判断该事件在某时刻的可能情况,在理论上是可行的。利用稻瘟病菌附着胞形成率和小麦赤霉病流行两实例对本法进行验证,准确度都较高。
The use of fuzzy mathematics for plant disease epidemiology may have greater potential. On the existing multiple sets of reliable data, the factors that have a greater relationship with popular events are taken as the elements of the discourse domain, and the membership functions are established. According to the actual situation, the popular events are divided into several levels and the representative subsets of each level are set up. Then, using the established membership function and representative subset, it is theoretically feasible to use the model identification method to determine the possible situation of the event at a certain moment. This method was validated by two examples of the formation of appressorium of M. grisea and the prevalence of wheat head blight. The accuracy of the method was high.