Research On Selective Assembly Of Complex Mechanical Product Using Genetic Algorithm

Assembly is the significant step for various parts assembling into products. It playsan important role in production, because it would decide the product’s quality. Intradition, the parts are assembled linearly, at randomly, the tolerance range of theassembled product will be the sum of the part tolerance range. Thus precisionproductions that are manufactured reflect at high manufacturing cost. Selectiveassembly is a technique that allows the increase tolerance of parts, and then achievesassembly accuracy by measuring, marking and mating the parts accordingly. Traditionalselective assembly could only be used in small part type and quantity. So computeraided assembly is researched nowadays for the complex problem. The parts are dividedinto groups and then search the optimal combination by computer. It could improve theutilization and assembly quality effectively.An approach to minimizing surplus parts in selective assembly with geneticalgorithm is proposed. The parts not always follow the normal distribution that wouldlead to surplus parts. So a new grouping method considering the tolerance of parts isproposed, by which, different group numbers are assigned to different parts withdifferent tolerance range. And also a new chromosome structure is proposed in thispaper, it is two-dimensional matrix considering the information of grouping. A ballbearing consisting of three mating parts is taken as the case to introduce the flow ofproposed genetic algorithm. The approach is also improved for the product assemblywhich has more than one dimension chain. A piston-cylinder assembly is taken as thecase.With case study, it is verified that the proposed approach is efficient to improve theassembly success rate and reduce the surplus parts. So it is significant for the actualproduction.