Training method of tunneling parameter prediction model based on incremental learning and related equipment

The invention provides a training method of a tunneling parameter prediction model based on incremental learning and related equipment. The tunneling parameter prediction precision of a tunneling machine of the model can be improved. The method comprises the steps that an initial tunneling parameter prediction model is constructed based on a bidirectional long-short time memory network, the tunneling parameter prediction model based on incremental learning comprises an input layer, a first hidden layer, an output layer and N second hidden layers, and activation functions of the N second hidden layers are unsaturated activation functions; target input data is input into the initial tunneling parameter prediction model to obtain predicted tunneling parameters, and the target input data is any data in the standardized model input data set; optimizing a hyper-parameter combination of the initial tunneling parameter prediction model based on a quantum particle swarm algorithm, the predicted tunneling parameters an