Research And Implementation Of Human Resource Scheduling In Mail Distribution Center

As an important link of postal express network,postal distribution center is an operation place for sorting and processing the goods delivered by the public,which occupies an important position in the whole postal logistics operation chain.However,the post arrangement of the staff inside the distribution center is relatively fixed,and the subjectivity of personnel scheduling is relatively large.The problems such as unreasonable personnel allocation and chaotic working procedure arrangement often occur in the mail processing process,which affects the mail processing efficiency and human cost of the whole distribution center.Based on mail processing process and organization structure on the basis of the analysis of operational staff,the human resource scheduling problem faced by hub mathematical model transformation,abstract and concrete algorithm is studied,and the multiple projects under the environment of human resource scheduling provides decision support,and developed many project human resource scheduling management system.The main research work of this paper is as follows:Firstly,the flow problem of operator on multiple mail processing lines is abstracted into a kind of multi-project human resource scheduling problem according to the relationship between mail processing process and operator.The problem is further divided into singleobjective multi-project human resource scheduling problem and multi-objective multi-skill human resource multi-project scheduling problem,and the corresponding mathematical model is established.Secondly,an improved discrete cuckoo search algorithm and a reinforcement learning algorithm combining multiple meta-heuristic algorithms are proposed to solve the scheduling problem of single objective human resource and multiple projects.In the improved discrete cuckoo search algorithm,according to the characteristics of the problem,a Levy flight scheme based on task position movement,improved inversion variation and task list reorganization,and a random walk scheme based on adaptive repair exchange variation were designed to improve the optimization accuracy of the algorithm.At the same time in order to combine the cuckoo search algorithm stronger global search ability and good convergence,particle swarm algorithm to design the fusion many kinds of heuristic algorithm of reinforcement learning algorithm,the reinforcement learning technology combined the two algorithm,according to the evolution of the population status,choose the best performance,guiding algorithm update solution space more effectively.Experimental results on MPLIB data set show the improved effect of the algorithm and the effectiveness of reinforcement learning,and verify the good optimization performance of the two algorithms.When dealing with practical problems,these two algorithms can also give faster scheduling results than the original scheme,which reflects the effectiveness of the algorithm and its engineering significance.Thirdly,an improved multi-objective imperialist competitive algorithm and a multiobjective imperialist competitive algorithm based on reinforcement learning are designed to solve the multi-objective and multi-skill human resource multi-project scheduling problem.In the improved multi-objective imperialist competition algorithm,a hybrid priority rule is designed to generate the initial population to increase the diversity of the initial population.Three search operators,colony revolution based on insertion mutation,empire enhancement based on local search and interempire communication based on single point crossover,are added to make the algorithm not easy to fall into local optimization and improve the global search ability of the algorithm.The multi-objective imperialist competitive algorithm based on reinforcement learning adaptively selects five different search operators in imperialist competitive algorithm by means of reinforcement learning model to improve the search efficiency and optimization accuracy of the algorithm.The experimental results of the two algorithms on i MOPSE data set verify the effectiveness of the improved algorithm and the necessity of using reinforcement learning mechanism,and prove that the search results of the two algorithms are better and more dispersed.In addition,these two algorithms can also quickly provide a set of mail processing scheduling scheme with the minimum balance of total time and labor cost when dealing with the corresponding practical problems,which reflects the practical value of the algorithm.Finally,the postal hub project human resource scheduling management system is designed and developed.The job scheduling model and four kinds of algorithm is blended in among them,in the form of tables and gantt chart shows the scheduling scheme,in the form of news to push to improve the efficiency of the distribution center information bulletin,and use the Redis cache,RBAC authorization management,and other technology improve the performance and security of the system.