论文部分内容阅读
提出一种在非平稳需求以及区间预测需求情况下的基于置信规则推理的库存控制方法.该方法不依赖于需求的分布模型,区间预测需求利用能够处理多种定性和定量不确定性信息的ER(证据推理)框架进行表达,领域专家知识可以用来构建和初始化置信规则库,历史需求信息可以用来训练置信规则库,以得到更加可信的推理.给出了一个汽车4S店库存-销售实例,证实了该方法的可行性及其相对于传统方法的优越性.
This paper proposes an inventory control method based on confidence rule reasoning in the case of non-stationary demand and interval forecasting demand.The method does not depend on the distribution model of demand, and the interval forecasting needs to use the ER which can handle a variety of qualitative and quantitative uncertainty information (Evidence Reasoning) framework, field expert knowledge can be used to construct and initialize the confidence rule base, and historical demand information can be used to train the confidence rule base to get more credible reasoning.An automobile 4S shop inventory - sales The examples prove the feasibility of the method and its superiority over the traditional methods.