Rapid inversion method for logging-while-drilling information based on transfer learning

The invention discloses a transfer learning-based logging-while-drilling information rapid inversion method, which comprises the following steps of: constructing a logging-while-drilling azimuth electromagnetic wave logging forward model, and simulating logging-while-drilling response of a multi-component logging-while-drilling electromagnetic wave logging instrument under different stratum parameter conditions by utilizing the logging-while-drilling azimuth electromagnetic wave logging forward model; the method comprises the following steps: constructing a logging-while-drilling response database, combining a convolutional neural network with an LSTM network to construct a logging-while-drilling azimuth electromagnetic wave logging information inversion model, training and optimizing logging-while-drilling response data in the logging-while-drilling response database to obtain a source domain network model, and constructing a target domain network model based on transfer learning. After the target domain net