大型装备制造企业期量标准智能生成系统研究与开发

Research and practice on the preiod and quantity standard generation system of large-scale equipment manufacturing enterprises

生产提前期、生产批量等期量标准是MRP子系统依据的核心参数,由于大型装备制造企业具有产品规模大、结构复杂、通常按订单生产、产品变型设计频繁等特点,其产品期量标准的觚定存在很大的复杂性,期量标准准确性差大大影响了ERP的实施效果。针对以上问题,本文提出了期量标准的智能化解决方案。利用BP神经网络及其变形网络"识别"历史数据中最相似的"零件模型",并对新型零件的提前期进行"预测",其核心是网络的学习方法——BP算法。以此为理论依据,我们提出了详细设计方案,开发出了相应的计算机系统,运用BP神经网络结合梯度下降法来对变型零件的期量标准进行估算,达到了很好的效果。

The period and quantity standards, such as production lead time and production batch, are the core parameters MRP system relied on. In large-scale equipment manufacturing enterprises, the products are so large and complex. They are made to order and need to be redesigned usually. These particularities bring more complexity to generate the period and quantity standards. The imprecise data influences the application effect of ERP. In order to solve this problem, an intelligent generation solution of the period and quantity standards is introduced in this paper. It uses BP neural network and its transmutation network to identify the nearest part model in the history data, and estimate the lead time of the redesigned parts. The core of this solution is the network learning algorithm, BP algorithm. Based on this theory, we put forward a detailed design proposal, develop corresponding computer programmes. The effect is good to use BP neural network and gradient decent method to estimate the period and quantity standards of redesigned parts.