Method for monitoring and controlling abnormal gushing of goaf gas into working face

The invention discloses a method for monitoring and controlling abnormal gushing of goaf gas into a working face, which comprises the following steps: firstly, acquiring ground atmospheric pressure, underground atmospheric pressure, adjacent goaf air pressure, ground temperature, underground temperature, adjacent goaf temperature and return air oxygen concentration data, and establishing an oxygen concentration LSTM neural network prediction model; the parameters serve as a model training input data set, and a model output variable is the working face oxygen concentration; the model is trained, so that the model obtains the corresponding relation between the parameters and the oxygen concentration of the working face; during follow-up monitoring, after real-time data of all parameters are regularly collected and input into the model, the model can output the O2 concentration predicted value of the working face, meanwhile, the O2 concentration of the working face is graded, corresponding treatment measures are