强化学习一词来自于行为心理学，这门学科把行为学习看成反复试验的过程，从而把环境状态映射成相应的动作．在设计智能机器人过程中，如何来实现行为主义的思想、在与环境的交互中学习行为动作？ 文中把机器人在未知环境中为躲避障碍所采取的动作看作一种行为，采用强化学习方法来实现智能机器人避碰行为学习．Ｑ－学习算法是类似于动态规划的一种强化学习方法，文中在介绍了Ｑ－学习的基本算法之后，提出了具有竞争思想和自组织机制的Ｑ－学习神经网络学习算法；然后研究了该算法在智能机器人局部路径规划中的应用，在文中的最后给出了详细的仿真结果 The term reinforcement learning comes from behavioral psychology, which considers behavioral learning as a process of trial and error, mapping the state of the environment into action. In the process of designing intelligent robots, how to realize the idea of ​​behaviorism and learn to act in the interaction with the environment? In this paper, the action taken by robots to avoid obstacles in an unknown environment is taken as a kind of behavior, and the reinforcement learning method is adopted to realize intelligent robot avoidance behavior learning. Q-learning algorithm is a kind of reinforcement learning method which is similar to dynamic programming. After introducing the basic Q-learning algorithm, a Q-learning neural network learning algorithm with competitive ideas and self-organizing mechanism is proposed. Then, The application of the algorithm in the local path planning of intelligent robots gives the detailed simulation results