Shield construction parameter multi-objective optimization method based on machine learning and improved genetic algorithm

The invention discloses a shield construction parameter multi-objective optimization method based on machine learning and an improved genetic algorithm. The method comprises the following steps: acquiring various shield construction parameters, horizontal and vertical deviation correction errors of a shield attitude and settlement data of ground surface deformation caused by shield tunneling; establishing two SVM (Support Vector Machine) prediction models of the shield construction parameters, the shield horizontal attitude and the shield vertical attitude, and an LGBM prediction model of the shield construction parameters to the final surface settlement to obtain a regression prediction function; taking horizontal and vertical attitude deviations of a shield and final settlement of ground settlement as control targets, taking a regression prediction function as a fitness function of an improved NSGA-III algorithm, setting constraint conditions of shield construction parameters, and constructing an LGBM-SVM-N