Multi-objective Optimization Of Mixed-model Assembly Line Balancing Problem In Z Automobile Company Based On Multi-population Genetic Algorithm

With the diversification of customer needs,increasingly product variety,there is a severe challenge to the production and operation of modern manufacturing.Mixed-model production has thus become more and more popular as a form of production organization,relying on the traditional industrial engineering techniques to solve the multi-species mixed model assembly line balancing problem is in urgent need for a more efficient way.Z Motor Company's production line is mixed of T and X models,and there is an optimization issue of mixed-model assembly line balancing.Traditional solution of assembly line balancing problem is complex,lack of systematic,and most will involve equipment,layout and secondary tools design and re-investment.This paper presents an improved multi-population genetic algorithm,and applies it to optimize assembly line balancing problems,defined as the assembly line design optimization.At present,industrial engineering solution for such problems is more common to intelligent optimization method,and in the intelligent optimization method,genetic algorithm is often used,suitable for solving NP problems in combinatorial optimization problem because of better match for assembly line balancing problem.But the current domestic research about assembly line balancing GA mostly concentrated in a single population genetic algorithm,population diversity is only guaranteed by the mutation operator,which is in the initial design phase of the algorithm planted the possibility of local optima,in recent years,double-populations genetic algorithm for this problem has been improved in the initial population generation stage to improve the diversity of the population.Based on this,multi-populations genetic algorithm and re-inserted between the operator and the sub-species migration strategy to is used for further improved genetic algorithm,so that in operation effectiveness,efficiency,and avoid local optima aspect has been improved.This article will address the assembly line balancing multi-objective optimization,and put forward the principles and methods for multi-objective scale transformation coefficient setting,gives multi-objective problem-solving process theoretical basis,the use of genetic algorithms to achieve goals while addressing multiple purposes.This article constructs an assembly line mixed-model multi-objective optimization model(MALBP-MOO),and introduces principles,methods and design process of a multi-population genetic algorithm,and finally developed a GUI graphical user interface through Matlab programming.And evaluate program effectiveness and reliability through experimental data validation and analysis.Finally,the software is applied to Z Motor Company interior line mixed-model multi-objective optimization problem.