火力与指挥控制2025，Vol.50Issue(4)：160-168,9.DOI:10.3969/j.issn.1002-0640.2025.04.022
基于知识和数据联合驱动的目标意图识别方法
Target Intention Recognition Method Based on the Joint-driven of Knowledge and Data
摘要
Abstract
A target intention recognition method based on the belief rule base and deep learning is proposed.This method integrates the advantages of data-driven methods and expert system methods,possessing both powerful learning ability and maintaining the interpretability of the decision-making process.In order to ensure that the parameters of each module of the model can be fully trained,a collaborative optimization approach using backpropagation and differential evolution is proposed to tune the model.When this method is applied to the task of airborne target intention recognition,the prediction accuracy can reach over 90%,indicating that the expert knowledge embedded in the belief rule base can provide additional information,thereby enhancing the performance of the deep learning model.关键词
置信规则库/神经网络/协同优化/目标意图识别/空战Key words
BRB/DNN/collaborative optimization/target intent recognition/air combat分类
航空航天引用本文复制引用
杨光飞,郑轩,饶世钧,王苛芸,崔家连,刘连..基于知识和数据联合驱动的目标意图识别方法[J].火力与指挥控制,2025,50(4):160-168,9.基金项目
国家自然科学基金资助项目(42071273) （42071273）
中央高校基本科研基金项目(DUT24YG147) （DUT24YG147）
