The Study On Dynamic Pricing And Ordering By Censored Information Updating For Multi-period Demand

Demand information censored (DIC) means if demands exceeds the available inventory, the demand information can not be observed for the reason of stock-out, then we callthe situation as partly observed information or censored information. DIC is an important character of the market.which is a very common situation in nowadays inventory management. With the rapid development of global economy and the coming-up of buyers’market, more challenges will flood to the companies:the uncertainty of the market, the diversities of product requirement, the contract of product life cycle and order lead time which need the control of inventory more flexible. But because of censored information, companies will underestimate the size of the market and cannot make the accurate strategy approach to the real market. And the previous researches mainly focus on fixed demand or the discrete demand distribution. Because if continuous demand distributions have computable difficulties in multi-period issues. So the study in this article is meaningful both in theory and in reality.We study the dynamic pricing and ordering under the DIC situation with Bayes information updating. We will formulate two different models:perishable inventory problem and nonperishable inventory problem. In the article we show Weibull density is the only member of newsvendor distributions for which the optimal solution can be expressed in scalable form. Consequently, scalability yields sufficient dimensionality reduction so that we can derive the optimal solution and profit in easily computable simple recursions for the Weibull demand and in explicit closed form for the exponential demand. In the third part of the paper, Three nonperishable inventory models are formulated under two periods situation, and make comparisons among censored information updating, full information updating, and no information updating. In the fourth part of the paper, a perishable inventory model under multi-period situation is formulated, and also we give an algorithm to compute dynamic prices and orders, this method can be the decision support used in companies. Then we show the numerical analysis and sensitivity analysis to get some insight results.