基于证据加权不确定推理构建鸡病诊断与信息管理系统  

Construction of auxiliary diagnostic and information management system of chicken diseases based on uncertain reasoning of weighted evidence

作　　者：李建新[1] 肖建华[1] 秦宏宇[1] 高翔[1] 王欢[1] 王洪斌[1] LI Jianxin XIAO Jianhua QIN Hongyu GAO Xiang WANG Huan WANG Hongbin(College of Veterinary Medicine, Northeast Agricultural University, Harbin 150030, China)

机构地区：[1]东北农业大学动物医学学院,黑龙江哈尔滨150030

出　　处：《畜牧与兽医》2017年第9期92-97,共6页Animal Husbandry & Veterinary Medicine

基　　金："十二五"农村领域国家科技计划课题(2015BAD12-B05)

摘　　要：针对鸡常见群发病,该文采用Visual Basic.NET 2008、Asp.NET等技术,以产生式规则进行知识表示,建立基于证据加权不确定推理机制,并根据角色存取控制进行用户及后台维护,开发疾病诊断与信息管理决策支持系统以提高鸡病诊断正确率及疾病信息管理的效率。构建了丰富的鸡病知识库,推理机制中引入加权可信度,提出3种诊断方式可对91种鸡常发病进行辅助诊断,系统提示相应的治疗及预防方案,并提供直观的病例统计图表。通过800例确诊病例资料验证,系统辅助诊断符合率为95.50%,基本符合基层兽医诊断要求,通过不断地维护及更新系统知识库,不断丰富系统知识库,提高诊断符合率,使该系统逐步推广应用。By collecting and analyzing the diagnosis and treatment experience from the chicken disease experts,establishing the disease diagnosis and treatment data model,and importing the weighted uncertain reasoning,an auxiliary diagnostic expert system of chicken diseases based on Microsoft. Net and SQL Server 2008 was developed. According to the characteristics of chicken diseases,this study established an chicken disease knowledge base,and introduced weighted credibility to inference mechanism. The system could diagnose 91 common chicken diseases and provide three kinds of diagnostic methods,and also provide suggestions on appropriate drugs and treatment options. Users could check the medical record statistical charts. The performance test showed that diagnostic accuracy of the system was up to 95. 50% in 800 cases,which can basically meet the requirement of the subject. By constantly maintaining and updating the knowledge base,and enriching the knowledge base,the system can definitely foresee a wide application in the countrywide.

关 键 词：鸡病 诊断 信息管理 证据加权 

分 类 号：S858.3[农业科学—临床兽医学]