将Qos的网络参数映射成用户体验QoE与网络能耗以及网络的切换损耗之差作为每次网络选择的即刻回报值， 应用Q-learning算法逐步动态根据网络QOE参数进行网络选择。(Mapping Qos network parameters into user experience QoE and network energy consumption and the switching loss of the network as the immediate return value of each n