基于CA-CMAC的Q学习截球算法

Intercepting Algorithm Based on CA-CMAC Q-learning

CMAC神经网络的主要优点在于其局部修正权重系数,使每次修改的权重系数极少,因此具有快速学习能力,非常适合于在线实时控制。但是在实际的使用过程中,CMAC算法会产生可信度的分配问题。因此使用CA-CMAC来代替CMAC。Q学习是一种重要的强化学习方法,将Q学习与CA-CMAC网络相结合用到Robocup仿真环境中,使智能体通过学习训练球员的截球能力。通过实际的仿真取得了很好的效果,证明算法是有效可行的。

The main advantage of CMAC neural network is the partial correction of weight coefficient.As change very little weight coefficient to get faster learning ability,CMAC is very suitable for real time control.But in the actual process of using,CMAC often faces the problem occurred by the uneven distribution of units＇ credibility,therefore used CA-CMAC instead of CMAC.Q-learning is an important method of reinforcement learning,combined Q-learning and CA-CMAC neural network uses the algorithm in Robocup simulation for improving the agent＇s ability of intercepting.Good results through the simulation are got which shows that the algorithm is deasible and effective.