Improved Multi-objective Fireworks Algorithm And Its Application In Low-carbon Vaccine Cold Chain Distribution

With the improvement in the level of social medical and health care,the rising demand for vaccines leads to the great challenge of vaccine cold chain distribution.The atmospheric environment will be polluted by a large number of carbon emissions in the process of distribution.It is necessary to plan the route of distribution vehicles reasonably,so that the distribution scheme can meet people's demand for vaccines,reduce the transportation cost and negative impact on the environment.Fireworks algorithm,as a swarm intelligence optimization algorithm,has simple structure and high solving efficiency,which is suitable for solving NPhard problems such as logistics path optimization.Based on the above background,this paper studies the multi-objective fireworks algorithm and its application on low-carbon cold chain distribution of vaccines.The main works are as follows:Firstly,this paper analyzes the shortcomings of the existing multi-objective fireworks algorithm,and a local region enhanced multi-objective fireworks algorithm with subpopulation cooperative selection is proposed.The concept of optimal subpopulation spark is introduced into explosion operator and selection strategy,and the local region enhanced explosion operator and the subpopulation cooperative selection strategy are obtained,which make full use of the information of local optimal solution,increase the diversity of fireworks population,and improve the efficiency of the algorithm.The effectiveness of the improved strategy is verified on the standard test set WFG for multi-objective function optimization.Compared with five existing multi-objective meta-heuristics algorithms,the results show that the proposed algorithm has better convergence and diversity on the WFG problem.Secondly,a constrained multi-objective optimization model for low-carbon vehicle routing problem(referred to as low-carbon VRP)is established.In this model,a carbon emission measurement method considering various factors is introduced,which takes the total carbon emission and the longest time of the route as the minimum goal,and the number of available vehicles as the constraint condition.To solve the model,a discrete local region enhanced multi-objective fireworks algorithm is proposed.In order to improve the convergence,diversity and the distribution width of the Pareto front,the partial mapping explosion operator,the hybrid mutation operator for adjusting the route tasks and the objective-driven heuristic extended search strategy are adopted.The conflict of two objectives in the proposed model and the effectiveness of improved strategy are verified on nine low-carbon VRP examples of different scales.Compared with the four existing algorithms,Higher solution accuracy and better solution set distribution performance are obtained by the proposed algorithm,and it is suitable for the proposed algorithm to solving low-carbon VRP of different scales.Thirdly,on the basis of low-carbon VRP,enterprise transportation cost and customer dissatisfaction are taken as optimization objectives,and the low-carbon cold chain distribution problem of vaccines was modeled as a multi-objective optimization problem with the constraints of available vehicle quantity,capacity and time window.A discrete two-archive local region-enhanced multi-objective fireworks algorithm is proposed.With many strong constraint conditions,the proposed algorithm adopts decoding operator to eliminate vehicle constraints.Two external archives of feasible solution and infeasible solution are set,and the feasibility search is implemented for infeasible archive.The experiment results show that the objective setting of the problem is reasonable.The improved strategies can search a set of Pareto nondominated solutions with high convergence accuracy,uniform distribution and wide range in the low-carbon cold chain distribution of vaccines.