学习能力是认知无线电区别于现有电台的最主要特征,研究基于SOAR架构的案例学习功能在已有认知引擎平台中的设计与实现。首先介绍了人工智能开发工具SOAR的基本原理与概念,进而在认知无线电原型系统平台上,设计并实现了基于SOAR架构的,具备案例学习能力的认知学习模块。具备案例学习能力的认知引擎使认知无线电系统具备从以往经验中获取知识的能力,提高了其在未来任务中的决策性能。
Learning capability is the key characteristic of a cognitive radio distinguished from a traditional radio.In this paper,the design and realization of SOAR-based episodic learning function in the available cognitive engine platform are studied emphatically.The basic principle and concepts of SOAR are introduced,which is a widely used in development architecture for artificial intelligence system.The design and realization of SOAR-based episodic learning module are elaborated.By applying episodic learning to cognitive engine,the cognitive radio system could acquire knowledge from experiences and thus improve decision-making capability in its future tasks.