Research On Balancing And Sequencing Problems Of Stochastic Assembly Line

As a production organization mode used widely in automotive industry, assemble lineplays extremely important role in the automotive course of development.However, most ofresearch literatures about assembly line balancing and sequencing are based on strictpremises nearly half a century, definite task times. Thus, many research achievements aregenerally difficult to apply directly in production practice. Aim at research status onprogramming design of assembly line, stochastic assembly line balancing and sequencingproblems have been studied in depth in this paper.At first, stochastic assembly line balancing problem has been studied. On the groundof small lot sizes product variety, a mathematical model of stochastic assembly linebalancing problem was established aim at balancing rate, smoothing index and completionprobability. Through taking examples from idea of chromosomes crossover operator andmutation operator, basic particle swarm optimization algorithm was improved. At the sametime, chromosomes encoding and decoding method based on task rule were designed tosolve stochastic assembly line balancing problem.Second, stochastic assembly line sequencing problem has been studied. A simulationmodel of stochastic assembly line was built and an improved particle swarm optimizationalgorithm was proposed which two objectives were optimized simultaneously:consumpution rate of parts and the total completion time of product. On account oftraditional method solved product sequencing problem, a method based on particle swarmoptimization algorithm combined with simulation analysis was put forward. Therefore,particle swarm optimization algorithm is integrated into eM-plant simulation software.The method has two advantages: high-efficiency search capability of particle swarmoptimization algorithm and convenient for dynamic profiling of eM-plant software.Thus,we can avoid putting up simulation analysis through sequencing scheme typed manually.Aslo, the efficiency of searching of algorithm can be improved greatly and less work mustbe done by designers. In the meantime, all the searching processes could be showeddynamically.At last, rear axle assembly line balancing layout and design of some enterprise is takenfor example to study assembly line balancing and sequencing problems,and they weresolved through software platform of Matlab and eM-plant. Availability of the methodmentioned in the paper is verified to solve large-scale stochastic assembly line balancing and product sequencing problem by research results. Aslo, it has theoretical meaning andgood application value of popularization for modern manufacturing enterprises.