Wireless multi-base-station positioning network optimization method integrating reinforcement and joint learning

The invention discloses a wireless multi-base-station positioning network optimization method integrating reinforcement and joint learning, which comprises the following steps: S1, collecting common feature data and personalized feature data of different optimized base stations before and after state transformation, fitting a linear model by using a partial least square method, and carrying out cross validation until the maximum cross covariance is reached, performing iterative optimization to obtain a common feature linear model; s2, selecting similar base stations according to the optimized base station deployment parameters and the optimized base station parameters, and constructing a base station common model; and S3, introducing the public feature linear model into a base station public model, judging whether a personalized feature model needs to be introduced to other base stations according to positioning terminal performance data of other base stations, if the positioning terminal performance data is