Stochastic Demand Inventory Routing Problem Optimization

Inventory routing problem is a comprehensive optimization problem that integrates inventory management,vehicle routing and transportation decision.Inventory management or transport paths are optimized individually,can not reducing the cost of the whole supply chain.Inventory routing problem take into account constraints such as customer inventory,transport vehicle capacity,and in a logistics system of individual suppliers and multiple geographically dispersed customers,the inventory strategy is jointly optimized with the distribution routing to achieve the goal of overall cost reduction.The inventory routing problem under stochastic demand needs to consider the influence on inventory distribution solution caused by the change of demand.Therefore,on the basis of ensuring inventory stability,the joint optimization of inventory strategy and distribution routing is the focus of this paper.Firstly,the single-good inventory routing problem which does not consider the backlog in the limited plan period under stochastic demand has been studied,and establishes a linear mixed integer programming model with the goal of minimizing the total cost of inventory distribution,and solves it directly.The validity of the model and algorithm is verified by an example,and the necessity of considering robustness and infinite planning period is introduced through the influence of external demand change on the results of inventory distribution strategy.Secondly,this paper studies the problem of single-good inventory routing problem considering the backlog in the infinite plan period,and uses Lyapunov optimization method to combine the stability of inventory queue with the optimization goal of minimize inventory distribution cost,and establishes a mixed integer programming quadratic model.The algorithm transforms the global optimization problem under the original infinite plan period into an iterative optimization problem under a single plan period,and theoretically proves the equivalence of the transformation.Through the example and theoretical analysis,the validity,robustness and adaptive characteristics of the change of the external demand of the model are illustrated.Finally,the problem of multi-goods inventory routing problem considering the backlog in the infinite plan period is studied,and the application range of the algorithm is extended.The branch and bound method is used to solve the example of the mixed integer programming quadratic model,and the validity and low computational complexity of the algorithm are further verified.