Wireless access point version upgrading method and system based on reinforcement learning

The invention belongs to the technical field of wireless communication, and particularly relates to a wireless access point version upgrading method based on reinforcement learning. The method comprises the following steps: S1, starting to initialize parameters, and assuming that N APs are accessed to the same AC; s2, receiving an AP upgrading request by the AC, and setting a set variable K, a counting variable Count and a cumulative counting variable Total; s3, judging whether the number of the APs to be upgraded is equal to M or the number of the remaining APs is smaller than M, and skipping to the step S4; otherwise, skipping to the step S2; s4, the AC obtains a strategy according to a reinforcement learning algorithm, a version upgrading file is issued to the AP according to the strategy, and the AP is restarted to complete the upgrading after the version upgrading is completed; s5, AP upgrade selection and a bandwidth allocation strategy are carried out; s6, judging that all APs are upgraded, and ending