论文部分内容阅读
讨论了连续动作的强化学习系统实现及学习方法 .首先介绍了连续动作的强化学习系统的组成原理 ,讨论了采用神经网络实现强化学习系统的方法 ,然后 ,介绍了强化学习机制在智能机器人避碰行为学习系统中的应用 ,并给出了系统的仿真结果 .仿真结果表明机器人具有较好的避碰能力 .
The realization mechanism and learning method of continuous learning system are discussed.Firstly, the composition principle of continuous learning system is introduced, and the method of using neural network to strengthen learning system is discussed. Then, Behavior learning system and gives the simulation results of the system.The simulation results show that the robot has a good ability to avoid collision.