Lunar soil drilling temperature real-time prediction method based on spatial-temporal feature fusion

The invention relates to the technical field of computer science, artificial intelligence and lunar exploration and drilling, in particular to a lunar soil drilling temperature real-time prediction method based on spatial-temporal feature fusion. Comprising the following steps that drilling state parameters and drilling temperature data are collected, then data preprocessing is carried out, the provided model optimization method for early stop-Bayesian hyper-parameter optimization is added in the training process, hyper-parameters of the model are effectively adjusted, finally, a test set is input into the trained model, and the drilling temperature is obtained. According to the method, the real-time prediction of the drilling temperature is realized, in addition, a practical application effect test is carried out, good robustness and generalization ability are shown, a deep learning technology is applied, a Bi-LSTM (Bi-directional Long Short-Term Memory) and a self-attention mechanism are combined, character