摘要
伤亡性质鉴定意见是影响军人伤亡保险理赔的重要因素,伤亡性质鉴定错误将直接侵害伤亡军人、风险共担军人、以及承保军人伤亡附加险的保险公司利益。新作战环境下长期艰苦的备战使得军人因战、因公、因病的伤亡性质趋于复杂化和模糊化,而现行以军内各级卫生、政治部门出具的伤亡鉴定意见直接定性的方式极易产生口径不一、鉴定错误的情况。文章尝试结合模糊数学和人工神经网络的思想,构建具有自学习能力、容错能力、以及处理非线性关系的模糊BP神经网络模型,解决新作战环境下军人伤亡性质鉴定的模糊性、复杂性问题,从量化的角度提高伤亡性质鉴定的准确性。研究发现,通过伤亡时点等客观环境类和行为属性等主观行为类8项指标可以充分衡量军人伤亡事件的性质;采用模糊隶属度的转换,构建8×10×1的模糊BP神经网络结构,可以完成军人伤亡性质的精准鉴定;伤亡性质的模糊隶属度可以量化军人病伤、公伤或战伤的属性程度。
Casualty nature appraisal opinion is an important factor affecting the settlement of military casualty insurance claims.The mistake of casualty nature appraisal will directly infringe on the interests of military casualties,the risk-sharing military personnel,and the insurance companies that underwrite the additional insurance for military casualties.The long-term hard preparation under the new combat environment makes the nature of casualties of soldiers due to war,duty and illness tend to be complicated and blurred.However,the current method of directly determining casualties based on the medical and political departments at all levels in the army is easy to produce incidents of different standards and wrong identification.This paper attempts to combine fuzzy mathematics and artificial neural network to construct a fuzzy BP neural network model with good self-learning ability,fault-tolerant ability and nonlinear relationship to solve the fuzzy and complex problems in the identification of military casualties,and to improve the accuracy of the identification of casualties from the quantitative point of view.It is found that the nature of military casualties can be well measured by eight indices in objective environment and subjective behavior,such as the time point of casualties.The fuzzy membership degree can be used to construct the 8×10×1 structure of fuzzy BP neural network,which can accurately identify the nature of casualties.The fuzzy membership degree of casualties can quantify the degree of attributes of military casualties,public casualties or war casualties.
作者
刘辉
李仁传
Liu Hui;Li Renchuan
出处
《保险职业学院学报》
2019年第1期55-61,共7页
Journal of Insurance Professional College
关键词
军人伤亡保险
伤亡鉴定
模糊
BP神经网络
The military casualty insurance Casualty nature appraisal
Fuzzy
BP neural network