一种基于操作条件反射原理的学习模型-控制与决策.PDFVIP

第29 卷 第6 期 控 制 与 决 策 2014 年 6 月 Vol. 29 No. 6 Control and Decision Jun. 2014 文章编号: 1001-0920 (2014) 06-1016-05 DOI: 10.13195/j.kzyjc.2013.0522 一种基于操作条件反射原理的学习模型 a a,b b a 阮晓钢 , 黄 静 , 范青武 , 魏若岩 (北京工业大学a. 电子信息与控制工程学院，b. 实验学院，北京100124) 摘 要: 针对认知机器人的自主学习问题, 提出一种基于操作条件反射原理的学习模型(OCLM). 该模型采用状态 空间、操作行为空间、概率分布函数、仿生学习机制、系统熵等进行描述, 给出状态的“负理想度”的概念, 定义了取 向函数的计算方法. 运用模型对机器人避障导航问题进行仿真实验, 并对参数设置进行了讨论. 实验结果表明, 基于 OCLM 模型的机器人能通过与环境的交互获得认知, 成功避障到达目的地, 具有一定的自学习能力, 从而表明了模型 的有效性. 关键词: 学习模型；操作条件反射；自学习；仿生；避障 中图分类号: TP273 文献标志码: A A learning model based on operant conditioning principles RUAN Xiao-ganga , HUANG Jinga,b , FAN Qing-wub , WEI Ruo-yana (a. College of Electronic Information and Control Engineering，b. Pilot College ，Beijing University of Technology, Beijing 100124 ，China ．Correspondent ：HUANG Jing ，E-mail ：mymailhj@) Abstract: Inspired by Skinner’s operant conditioning theory, an operant conditioning learning model is presented to deal with the autonomous learning problem in cognitive robotics. The model is described by nine elements, including the space set, the action set, the bionic learning function and the system entropy etc. A notion “negative ideal rate” is defined to compute the orientation function. The OCLM is applied to solve obstacle avoidance and navigation problems for mobile robots. The experiment results show that the robot based on the model can autonomously learn how to arrive at the goal in a collision-free way through interaction with the environment, and show the effectiveness of the proposed model. Key words: learning model ；operant cond