Research On Dynamics Of Supply Chains

It is well known that supply chains are complex systems with large scales. Their complex behaviors have been observed by many researchers in recent years, and firms'efforts to manage supply chains have often led to frustration and helplessness. It calls for a model of a supply chain that more accurately reflects its true underlying complexity and dynamism. However, current researches mostly focus on the finding of chaos in supply chains, and it is still an open question that why the supply chains can exhibit such complex behaviours. Based on such a question, the present paper investigates the mechanisms that govern the dynamics of the supply chains by means of nonlinear system theory. In this paper, the dynamic stability of three periodic reviewed supply chains incorporating an order policy named as APIOBPCS(Automatic Pipeline, Inventory and Order Based Production Control System) are investigated. Dynamic models are given by means of state space method. Switched system theory and discrete system control theory are used to analyze the complex behaviour of the system. Management principles are given for each type of supply chains. And the influence of information sharing on the dynamics of supply chain is analyzed. Firstly, the dynamics of an inventory control system with enough supplies is investigated. Two cases of order cancellations policy are analyzed. If order cancellations are permitted, the system is linear. A discrete time model in the form of state space equations is developed and a novel method to analyze discrete systems with constant time delay is proposed, and the general stability criterion of the order policy parameters for a system with any lead time is given by Rouche's Theorem. For the case that order cancellations are not permitted, the inventory control system becomes a switched system with two subsystems. It is shown that it is the switching between the subsystems that induces the complex behaviours of the system. Subsequently, the dynamics of a 2-echelon supply chain is investigated, which consists of one retailer and one distributor. Here, the distributor does not set the safe stock and order cancellations are permitted. The shipment of the distributor in each period has three different expressions according to the on-hand inventory level of the distributor and the demand of the retailer, and this can be termed as the supply uncertainty. With this kind of uncertainty, the supply chain becomes nonlinear and can be modeled as a switched system with three subsystems. The switched rules are formulated and stability of each subsystem is analyzed. It is pointed that there are interactions between the members of the supply chain and improper order policy parameters may lead the system to chaos. This dissertation then extends the supply chain to a 2-stage distribution supply chain with a distributor and two retailers. There is supply uncertainty in such a system too. The switched system model of the supply chain is given, and its subsystems, switching rules and stability are analyzed respectively. The parallel interactions between the members of the system are investigated by simulations. On the basis of the analysis of the 2-stage distribution supply chain, the effect of information sharing on dynamics of a two-stage distribution supply chain with a distributor and two retailers is studied finally. Two cases of information sharing between the distributor and the retailers are considered. In the first case, the distributor obtains information from the retailers about their demand forecasting, and in the second case, the distributor obtains additional information from the retailers about their inventory level. We compare the dynamics between the two cases and then point out that the second case outperforms the first one.