基于语义匹配的作战体系仿真想定生成方法  

Combat system-of-systems simulation scenario generation approach based on semantic matching

作　　者：曾广迅 龚光红 李妮[1] ZENG Guangxun;GONG Guanghong;LI Ni(School of Automation Science and Electrical Engineering, Beihang University, Beijing 100191, China)

机构地区：[1]北京航空航天大学自动化科学与电气工程学院,北京100191

出　　处：《系统工程与电子技术》2021年第8期2154-2162,共9页Systems Engineering and Electronics

基　　金：装备预研基金(61400010112)资助课题。

摘　　要：针对传统的想定建模方法耗时长、成本高的问题,为了实现作战体系概念模型数据到仿真想定的数据映射重用,提出一种基于语义匹配的作战体系仿真想定生成方法。首先,基于网络爬虫技术获取武器装备的参数信息,构建了作战领域知识库。在此基础上,提出了新型的混合式语义匹配方法以提供仿真想定的数据来源,根据自然语言数据库WordNet开展调参实验确定算法参数,通过对比人工打分结果有效提高了算法的可信性。在确定了概念模型与仿真想定的数据映射关系后,给出了仿真想定的生成方法。最后,通过编队突防仿真应用案例,验证了该方法的可行性。The combat system-of-systems simulation scenario is the basic basis and data source for combat simulation.But the traditional scenario generating method have the problem of time-consuming and costly.In order to achieve the data mapping and reuse of the conceptual model data of the combat system-of-systems to the simulation scenario,a method for generating combat system-of-systems simulation scenario based on semantic matching is proposed.Firstly,information of weapons and equipment were obtained based on the web crawler technology,and a knowledge base in the combat domain was constructed.Then,a new hybrid semantic matching method is proposed to provide data sources for simulation scenario,and the algorithm parameters are determined by parameter tuning experiments based on the natural language database WordNet,and the credibility of the algorithm is effectively improved by comparing the results of manual scoring.After determining the data mapping relationship between the conceptual model and the simulation scenario,the method of generating the simulation scenario is given.Finally,the feasibility of the method is certified by using the application case of the formation raid defense system-of-systems.

关 键 词：语义匹配 作战体系仿真想定 网络爬虫 作战领域知识库 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]