The improved Q learning algorithm was suggested because of the traditional algorithm has limitations of slow and partial constringency.

• 传统的Q学习存在收敛速度慢和容易导致局部收敛的矛盾，为此提出一种改进的Q学习算法。