融合社会学因素的模糊贝叶斯网技术预测模型  

A technology prediction model based on fuzzy Bayesian networks with sociological factors

作　　者：张豪 李子凌 刘通 张大伟[1] 陶建华[1] ZHANG Hao;LI Ziling;LIU Tong;ZHANG Dawei;TAO Jianhua(National Laboratory of Pattern Recognition,Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China)

机构地区：[1]中国科学院自动化研究所模式识别国家重点实验室,北京100190

出　　处：《山东大学学报（工学版）》2023年第2期23-33,共11页Journal of Shandong University（Engineering Science）

基　　金：中科院战略性先导科技专项项目(C类)(XDC02050000)。

摘　　要：为研究诸如无人机等新兴信息科学的未来发展趋势,更好地把握这些技术发展与应用情况的动态并及时调整发展战略,在专家系统框架下提出一种基于模糊贝叶斯网的技术趋势预测方法,有效预测无人机技术在未来10 a的发展趋势。在构建预测方案的过程中,结合领域专家的知识,设计若干影响技术发展的维度用作预测参数,不仅包括技术型指标,也包括社会型指标,从而融合更加丰富的信息,使预测结果更加专业可信。提出基于模糊贝叶斯网的技术预测模型,分别对两种类型指标的影响程度进行综合计算形成推理机,使预测结果更加直观精细,具备一定的解释性。对预测模型的结果进行详细分析,并结合其他模型进行对比评价,结果与已有的专业预测结果相符合。试验结果表明,对无人机技术未来10 a发展趋势的预测中,基于模糊贝叶斯网的技术预测模型能够获取影响因素之间的关联关系,具有更好的预测效果。A prediction method based on fuzzy Bayesian networks was proposed under the framework of expert system in order to study development tendency of new information science and technology like unmanned aerial vehicle technology,better grasp the current developments and applications trends of those technologies and timely adjust the development strategy.This model could effectively predict development tendency of unmanned aerial vehicle technology in the next decade.During constructing the prediction solution,some dimensions which could influence the development of technology were designed to be served as parameters.Based on expert knowledge,those dimensions included not only technical indexes but also social indexes which could fuse more abundant information and make prediction result more professional and credible.A technology prediction model based on fuzzy Bayesian networks was proposed to calculate the influence degree of the two types of indexes for constructing an inference engine,which could make the prediction result more intuitive and interpretable.The results of the proposed approach were analyzed in details through comparing with other model,and it was roughly in line with the existing professional results.The experimental results showed that the proposed fuzzy Bayesian networks based prediction model could acquire the information about relationship among influence factors in predicting development tendency of unmanned aerial vehicle technology.Hence the proposed model had better performance.

关 键 词：技术预测 模糊贝叶斯 专家系统 多映射 社会学因素 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]