Research On The Bi-level Programming Model Of Supply Chain And Its Solution Algorithm Based On RFID Investment Incentive

Radio frequency identification technology originated in the Second World War.Because of its good recognition and communication performance,RFID has been widely pursued in the field of supply chain management like science and enterprises.However,the RFID market is not as strong as people had expected.Because the cost of RFID is mostly borne by the manufacturers,while the profits produced are all given to the downstream retailers.This makes the downstream enterprises of the supply chain,especially the large retail enterprises,have great enthusiasm for the RFID integration of the cross enterprise,but the upstream manufacturers are lack of power.In order to improve the enthusiasm of the upstream suppliers to invest in RFID,the downstream retailers need to introduce the corresponding incentive policies,and give a certain amount of reward for the effort of the upstream suppliers to invest in the RFID.So there is a new dynamic game between retailers and suppliers on the level of incentives and efforts.In order to describe the constrained dynamic game problem,we build a bi-level programming model of supply chain based on RFID investment incentive.Bilevel programming problem is a kind of system optimization problem with bilayer recursive structure.It includes the upper and lower two levels of planning,and the upper and lower two levels of constraints and objective functions will be affected by the decision variables of the other side.In all areas of our real life,bilevel programming is everywhere.However,bilevel programming is a NP-hard problem.It is very difficult to solve it even a simple linear bilevel programming problem.In order to solve the model in this paper,we design a hierarchical differential evolution algorithm,which is based on the idea of swarm intelligence algorithm and the cross population differential evolution algorithm based on opposition-based learning.The cross population differential evolution algorithm based on opposition-based learning is the improvement of the standard differential evolution algorithm,through the introduction of chaos dispersion strategy,reverse learning strategy,elite selection strategy and cross population cross method.OLCPDE algorithm expands the search scope of population,improves the global search ability of the algorithm,and enhances the stability of algorithm optimization.Through the experiment of 12 standard test functions,we prove that the OLCPDE algorithm can effectively avoid precocity,and has better convergence speed,optimization precision and robustness.In order to verify the feasibility of the hierarchical differential evolution algorithm,we choose two other algorithms to compare the 2 test models,and the results show the superiority of the hierarchical differential evolution algorithm.Finally,through the modeling,analysis and solution of the actual cases,we get the decision value of each supplier and retailer,which has certain guidance and reference significance for the supplier and retailer on the supply chain.