Study On Modeling And Algorithm Of Stock Control With Dynamic Multi-Interval In Retail Trade

Every corporation needs to manage stock in reality, it is hard to decrease cost through the process of producting with drastic changing of commercial circumstance, however there are much profit to pursue in the process of stock and transportation, especially for retail trade that deal with more complex problems of stock than others's. So the process of stock and transportation become the resource of increasing benefits. The problem of stock in this thesis is abstracted from real process of stock about retail trade.How much stock the corporation take directly influence the level of stock and current speed of stock, ulteriorly result in different cost of logistics and profit of corporation. Meanwhile, supplier can offer some favourable terms to influence the stock project of tradesman in order to get benefit each other. So the manager must use scientific ways to control stock. Firstly, need to satisfy the need of customer, secondly, choose the favourable terms rightly, the aim is to keep the quantity of stock to get most profit.In this paper, major work is as followed:1. The basic theories and some common ways about stock control are introduced, then the limitations are pointed out when they are applied to retail trade.2. According to the limitations of simple genetic algorithms, design genetic algorithms based on particle swarm.3. The mathematical models are derived.4. The results of experiment show that genetic algorithm based on particle swarm is more powerful than simple genetic algorithm on many aspects, and the data of experiment explain discrete model can control stock scientifically, get maximal profit.