Research On Lot-sizing Problem In Supply Chain Based On Chance Constrained Programming

Supply chain of lot-sizing problems have made a great contribution in managinginventory and cost savings. Due to a large number of factors, such as market demandand other conditions affecting environmental fluctuations, there will be all kinds ofunpredictable change, and it often can not achieve precise expectations. Then, fuzzychance constrained programming with joint replenishment problem emerged. In themodel, the demand was treated as fuzzy variables, and it can simplify the whole issueinto a mathematical model to solve, then it became a central issue to be addressed inthis model. Algorithm for fuzzy chance constrained programming with JointReplenishment Problem directly impact on the problem of the treatment effect, andultimately affect the actual value of the issue.Firstly, fuzzy chance constrained programming with joint replenishment modelfor multi-vendor demand conditions with fuzzy variables, identifies to minimizeinventory costs as the objective function. Through analysis of the model of theproblem, and then fuzzy chance constrained programming with joint replenishmentmodel need to convert to its equivalent clear model, in order to easy to solve.Secondly, according to the fuzziness of the variables and constraintcharacteristics in the model of fuzzy chance constrained programming with jointreplenishment, compare and select genetic algorithm and particle swarmoptimization algorithm as objects of study. The genetic algorithm requireschromosome encoding for decision-making variables, set chance-constraints and thefitness function and then handle the chromosome crossover, mutation and selection,then the best chromosome as a set of optimal solution. Under the same conditions,the particle swarm algorithm is used for the corresponding test results. At first, thesolution space need to be defined, set the constraints and the fitness function and thenthe particle speed is updated, which involving themselves and social learning, thenthe best of a group of particles as a set of optimal solution. The conclusions of the genetic algorithm and particle swarm algorithm are compared. Then, compare twodifferent algorithms performance of the optimization results to analyze, theevaluation of the performance in dealing with the fuzzy chance constrainedprogramming with joint replenishment problems between particle swarmoptimization and genetic algorithm is given and summarize its underlying causes.Finally, the algorithm discussed in this article need to apply to practice, sodevelop and design an adaptive inventory management system, and discuss thewhole system overall planning and implementation of technology-related, and verifythe application of fuzzy chance constrained programming model.