Research And Simulation On Storage Allocation And Task Scheduling Algorithm In Warehouse Management System

The logistics system is nowadays more and more complex,and there newly appears a new kind of high-density tridimensional warehouse based on shuttle and cruiser.This new kind of warehouse has more space utilization than the traditional ones,such as stacker based warehouse and forklift truck based warehouse.Therefore,the storage and retrieval allocation problem and the task scheduling problem in such new warehouse need to be considered.This paper studies the two problems in the shuttle and cruiser based high-density tridimensional warehouse.The storage allocation algorithm proposed in this paper can allocate storage positions in a dynamic way,and the strategy of storage allocation can be easily changed by modifying the penalty score strategy.This shows both extensibility and flexibility which can satisfy different business scenarios.Then the result of storage allocation can be parsed into a task sequence with dynamic probability matrix and the method of task sequencing.Finally a suboptimal solution can be get by using the Monte Carlo method based heuristic algorithm with dynamic probability matrix proposed in this paper.The works of this paper are:1.Research on storage allocation algorithm: This paper declare the constraints and optimization goals for storage allocation at first.And secondly a map accessibility scanning algorithm and a cargo way entrance selection strategy are proposed.Then this paper abstracts and maps the storage allocation problem into to knapsack problem.Then an algorithm based on penalty score strategy is proposed,using the thought of dynamic programming.Finally,real cases are used to prove that the dynamic programming algorithm is feasible.2.Storage task sequencing: At first,the task scheduling problem is constrained and simplified in this paper.A simple storage task can be parsed as a detailed task with a specific routine by using router tables.Then the storage tasks are analyzed,decomposed,abstracted and reorganized as pre-part and post-part.And a time compression method is proposed for this kind of reorganized task sequence.Finally,a task sequences generating method is proposed,which provides a research basis for task scheduling algorithm.3.Research on task scheduling algorithm: This paper proposes a Monte Carlo method based heuristic algorithm with dynamic probability matrix to solve the task scheduling problem.And the result of real case simulation shows that the heuristic algorithm can converge rapidly.And the performance of the proposed heuristic algorithm is good compared with the optimal solution in small-scale task scheduling problems.Compared with heuristic algorithms,the storage allocation algorithm in this paper has the advance of being able to obtain one of the optimal solution.Meanwhile,the time complexity of the algorithm proposed in this paper can also be lower than heuristic algorithm with a good performance does in a real-case scale.And Compared with heuristic algorithms such as genetic algorithm,ant colony algorithm and particle optimization algorithm,the task scheduling algorithm proposed in this paper is easier.In addition,within the real problem scale,it has high computing efficiency and does not fall into a local optimum,so it can be better applied to the automatic storage and retrieval system.