Scheduling Optimization Of JIT Parts Feeding For Mixed Model Automobile Assembly Lines

Considering the impact of the earliness and tardiness of the arrival ofparts to the assembly lines, this study explored the model of a line-feedingsystem which met the philosophy of JIT to minimize the adverse effects.Combining with the characteristics of the Mixed model assembly line,we designed the virtual warehouse jobs, which also allowing the trailers towait for a certain period of time in the warehouse, and made the arrivingtime of the parts close to the actual need time in the assembly line.Compared to the traditional single machine scheduling, the complexity ofproblem is greatly increased.According to the NP-Hard features of this type of problem, based onthe problem structure, we proposed a memetic algorithm (MA) withdifferent versions to solve the problem efficiently.In order to verify the validity of that algorithm, we generatedexamples in four different sizes, also compared the efficiency of thedifferent versions of the memetic algorithm. Computational experimentsshowed that the performance of memetic algorithm was much better thanthe genetic algorithm (GA) and Cplex. Based on the differentcharacteristics of the different versions of memetic algorithm, the companycould select different versions of the memetic algorithm in accordancewith the actual situation.