Variable-scale teaching and learning algorithm-based tolerance allocation multi-objective optimization design method

The invention relates to a variable-scale teaching and learning algorithm-based tolerance allocation multi-objective optimization design method. Through analysis, the manufacturing cost, the quality loss and the processing process energy consumption are determined as objective functions, and a cost, quality and energy consumption-based tolerance allocation multi-objective optimization function model is built; model constraints are processed by using a violation function method, initial populations is grouped and an optimal solution in groups is calculated; in a teaching stage, local search isenhanced for the optimal solution in the groups by adopting a DFP variable-scale method, so that students are ensured to learn from teachers; in a learning stage, the students are guided to learn fromoptimal students in the groups by adopting a directed learning strategy, so that the learning has a certain directivity; and finally a group of tolerance allocation multi-objective solving preferencedecisions are obtained by