Application Of Multi-objective Evolutionary Algorithm In Fixture Layout

In machining,fixture is a widely used clamping equipment.However,with the rapid development of machining technology,the traditional experience-based fixture design has been slowly unable to meet the needs of efficient and precise production.The main factor that affects the high efficiency and precision production is the machining quality,and the displacement of the workpiece is the direct factor that affects the machining quality.In this paper,a multi-objective workpiece position migration model is established for rigid parts,and an improved multi-objective evolutionary algorithm is proposed to optimize it,so as to find a more reasonable fixture layout.The purpose of this paper is to reduce the displacement of rigid parts.The rationality of fixture layout is analyzed,including correct positioning and reasonable constraints.In view of the shortcomings of the traditional workpiece position deviation model,this paper analyzes the relationship between the clamping force,the contact force,the contact force and the local deformation,and the relationship between the local deformation and the workpiece position deviation,so as to establish a multi-objective workpiece position deviation model to meet the different requirements of the workpiece in the X,y,Z three aspects of upward precision,so as to obtain a reasonable fixture layout.In the optimization of fixture layout,the basic principle of MOEA / D algorithm based on decomposition is studied,including the characteristics,basic principle,decomposition strategy and specific algorithm framework of MOEA / D algorithm.There are a lot of combinations of locating points in the fixture,so the algorithm needs to have strong search ability.Therefore,the migration behavior based on Gaussian mutation is introduced to improve the multi-objective evolutionary algorithm based on decomposition,which improves the overall adaptability of the population.Finally,it is proved that the improved method has stronger searching ability and can find better solution in all directions by testing function and milling through groove.In order to further improve the search ability of the algorithm,a quantum particle swarm optimization(QPSO)algorithm with strong global search ability is introduced in Chapter 3.Firstly,the basic principle,algorithm framework,process and advantages of quantum particle swarm optimization are studied.Then,combined with the characteristics of the adjacent subproblem in MOEA / D algorithm,the quantum search method based on the neighbor position is improved,and the global optimal individual is changed to the farthest neighbor position of the global optimal individual when the original method generates the attraction point.The purpose of this is to avoid premature convergence to the current best when it starts to converge at the end of the iteration It can continue to do a small range of local search in a certain range,strengthen the local search ability in the late iteration,and prevent the algorithm from falling into "precocity".And it replaces the first migration method of the third chapter to improve the global search ability of the algorithm.Through the same test function and milling through slot example as the third chapter,it is proved that the improved algorithm has better overall performance.