Multi-target prediction and optimization method for rotary abrasive water jet polishing

The invention discloses a multi-target prediction and optimization method for rotary abrasive water jet polishing, which comprises the following steps: S1, acquiring processing parameters and processing results to obtain an experimental data set; s2, shuffling the experimental data set, and dividing a training set and a test set according to a proportion; s3, using a Bayesian optimization algorithm to optimize hyper-parameters of the machine learning integration model; s4, inputting the shuffled and divided experimental data set into the optimized machine learning integration model for training and prediction; s5, selecting an optimal regression prediction model according to R2 and RMSE evaluation indexes; s6, performing multi-objective optimization by taking the optimal regression prediction model as an objective function of a multi-objective optimization algorithm; s7, selecting an optimal multi-objective optimization algorithm through the hyper-volume evaluation index; and S8, selecting an optimal solution