一种基于情感智能的机器人自主趋光行为研究  

Research on the autonomous phototaxis behavior of a robot based on emotion intelligence

作　　者：阮晓钢[1] 庞涛[1,2] 张晓平[1] 王尔申[2] 

机构地区：[1]北京工业大学电控学院,北京100124 [2]沈阳航空航天大学电信学院,辽宁沈阳110136

出　　处：《智能系统学报》2015年第1期97-102,共6页CAAI Transactions on Intelligent Systems

基　　金：国家973计划资助项目(2012CB720000);国家自然科学基金资助项目(61375086;61101161);北京市自然科学基金资助项目/北京市教育委员会科技计划重点资助项目(KZ201210005001)

摘　　要：针对移动机器人的自主趋光行为问题,提出了一种基于情感智能的内发动机仿生学习机制。该学习机制以生物体感觉运动系统的学习机制为基础,包括评价环节、行为选择环节和取向环节,采用模糊神经网络构建情感模型,情感模型的输出作为评价环节的内部奖赏信号。该学习机制能够使机器人在未知环境下通过自主的学习和训练逐渐形成、发展和完善趋光行为技能,通过情感智能的作用可以增加试探成功次数和减小学习步数,仿真实验证明了该方法的有效性。In this paper,a kind of bionic learning mechanism of intrinsic motivation on the basis of emotion intelligence is proposed for the autonomous phototaxis behavior of a mobile robot. The learning mechanism is on the basis of the learning mechanism of the sensorimotor system of a living body,including the links of evaluation,behavioral choices and orientation. The emotion model is constructed by the fuzzy neural network and the output of emotion model is taken as the internal reward signal of the evaluation link. The learning mechanism may make a robot gradually form, develop and perfect the phototaxis skills by autonomous learning and training in an unknown environment. The emotion intelligence may increase the number of exploratory successes and reduce the learning steps. The simulation results demonstrated the effectiveness of this method.

关 键 词：认知机器人学 内发动机 移动机器人 趋光技能 情感智能 感觉运动系统 模糊推理方法 取向性 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]