Study On Planning Models And Dynamic Coordination Mechainsms For Multi-stage Spare Parts Support Under Demand Uncertainties

Spare parts support plays an import role, because using spare parts to replace the destroyed components is the fastest manner to repair damaged weapons during wartime. In order to meet the requirements of Precision Support, spare parts support must be planned deliberatively prewar and be executed flexibly to deal with the various demand uncertainties during the wartime.Based on the features of spare parts support, Anticipative Decision and Adaptation Decision are defined which are combined in the decision framework for multi-stage spare parts support under demand uncertainties. The framework embodies the multi-stage character of spare parts support well; which on one side can realize object oriented long term planning, on the other side can achieve flexible supply based on dynamic states. The models and methods in the framework are established to answer how to select types and to forecast demands, how to deploy spare parts in the multi-echelon structure and how to distribute them during the wartime. The main innovative contents of the thesis are outlined as follows:(1)The valid type selection and demand forecasting methods are given. There are many factors need to be counted when selecting spare types, some of which are hard to quantified, and therefore Fuzzy Comprehensive Evaluation (FCE) is adopted. In order to ensure the effectiveness of the method, House of Quantity (HoQ) in Quality Function Deployment (QFD) is introduced to establish the evaluation indexes and their weights. Owing to the shortage of demand records, a new Fuzzy Inference Based Demand Forecasting (FIBDF) method is presented, in which expert forecasting value and Markov forecasting value are combined effectively. The method is able to handle various forecasting inputs, and can be widely used in demand forecasting under uncertainties.(2) Planning models are established; in wchich both cost and benefit are considered strategically to decide how to deploy spares. As the courses are not the same for different spare types, Stage Expected Fill Rate (SEFR) and Stage Expected number of BackOrders (SEBO) are selected as the performance parameter respectively, accordingly the chance constrain planning model for unrepairable unit is established and the optimization deployment model for repairable unit under different repair condition is given. When optimizing different spares, simulation based optimization methods are used to weigh different type of spares in which the optimization problem is taken as a combination one.(3)Dynamic coordination mechanisms are given and tested based on ABMS. During the process of muli-stage support, the spare parts should be distributed effectively in the multi-echelon structure to maximize the benefit. Based on the similarity between the spare parts support system and the multi-agent system, Agent based modeling and simulation methods are adopted to investigate the dynamic coordination mechanisms. The order decision model is established, in which the temporal restriction of Precision Support is reckoned seriously. Take the Reinforcement Learning (RL) as the coordination mechanism, the supply way is illuminated.Not only the effective models and methods are presented, but also Precision Support is stated innovatively from the combination of Anticipative Decision and Adaptation Decision. The researches our made are valuable exploration to realize military Precision Support, which will strengthen wartime spare part supports and enrich wartime logistic theory.