【摘要】 随着汽车售后服务市场竞争愈发激烈,为了提高市场占有率,配件代理商不仅要向客户提供优质的服务,还要减少库存成本。因此制定准确有效的配件销量预测模型对配件代理商至关重要。本文以汽车价值链业务协同过程中产生的配件销量数据为基础,提出了一种BP-GRU组合预测模型,用于配件的销量预测。该模型利用BP网络对数据的特征进行初步提取,接着利用GRU捕获时间长期依赖性,将BP和GRU的预测结果进行加权融合。最后采用了包括平台某配件代理商的销售数据在内的3种数据集进行了5组对比实验。实验结果表明,该组合模型的预测效果比起其他单一模型的效果更好。更多还原

【Abstract】 With the increasingly fierce competition in the automotive after-sales service market, in order to increase market share,parts agents must not only provide customers with high-quality services, but also reduce inventory costs. Therefore, it is very important for the accessories agent to formulate accurate and effective parts sales forecasting. Based on the parts sales data generated during the business collaboration of the automotive value chain, a BP-GRU combined forecasting model is proposed for the forecasting the sales of parts. The model uses the BP network to initially extract the characteristics of the data, and then uses the GRU to capture the long-term dependence of time, and then weights the BP and GRU prediction results. Finally, three data sets including the sales data of an accessory agent on the platform were used to conduct five sets of comparative experiments. The experimental results show that the prediction effect of the combined model is better than that of other single models.更多还原