王晓丹 汪惠芬 柳林燕

摘要：由于注塑行业成型系统较为复杂并且对环境较为敏感，注塑成型加工过程中的不稳定因素很容易导致产品不良的发生，造成经济损失。由于过程的复杂性，不可预见的干扰，设备故障以及原料成分的变化，一批批次中止时的产品质量可能与规格相差甚远，所以对注塑成型大数据进行分析，来感知这些不可见的干扰因素，然后通过分析建模解决甚至避免现场痛点问题。制造业一直面临着几个挑战，包括可持续性、性能和生产质量。制造商试图通过在制造过程层面融合物联网（IoT）和ICT（信息与通信技术），通过实施CPS（信息物理系统）来提高企业的竞争力。而CPS平台（或者说智能工厂）是根据制造企业的特点，由不同类型的数据采集/处理方法、决策规则和功能组成。本文根据注塑工艺的特点、构成框架的模块及其具体功能，提出了一种基于实时制造数据的智能注塑系统框架。希望本文能作为指导，提高注塑行业的市场竞争力，支持智能工厂的建设，为工业4.0做准备。

Abstract： Because the molding system of injection molding industry is more complex and sensitive to the environment， the unstable factors in the process of injection molding are easy to lead to the occurrence of defective products and economic losses.Unforeseen interference， because of the complexity of the process， equipment failure and the change of raw material composition， a number of batches to suspend the quality of the products may far and specification， so the analysis of injection molding large data， to perceive these invisible interference factors， and then through the analysis of modeling to solve the problem even pain points on site.The manufacturing sector has been facing several challenges， including sustainability， performance and production quality.Manufacturers are trying to improve their competitiveness by integrating the Internet of Things （IoT） and ICT（Information and Communication Technology） at the manufacturing process level， and by implementing CPS（Information Physical System）.The CPS platform （or smart factory） is composed of different types of data acquisition/processing methods， decision rules and functions according to the characteristics of the manufacturing enterprise.In this paper， an intelligent injection molding system framework based on real-time manufacturing data is proposed according to the characteristics of injection molding process， the modules that constitute the framework and their specific functions.It is hoped that this paper can serve as a guide to improve the market competitiveness of the injection molding industry， support the construction of intelligent factories and prepare for Industry 4.0.

關键词：注塑产品;质量控制;智能注塑系统;人工神经网络;数据驱动

Key words： automobile injection molding;quality control;smart factory;artificial neural network;data-driven

中图分类号：U471.14                                    文献标识码：A                                  文章编号：1674-957X（2021）14-0173-04

1  简介

随品质控制的新时代，“零缺陷”是各行各业所追求的终极目标，强调“第一时间把事做对”的零缺陷制造（ZDM）已成为制造业目前正在面对且想要克服的最大挑战之一。

不同于互联网行业在跟随移动、社交、电商互联的浪潮进行了一系列的数据挖掘和分析探索，大量的工业企业日积月累的过程数据、监测数据还有质检数据并没有充分地得到利用和探索，这些未被充分发掘的数据价值对于工业企业来说是一笔潜在的财富[1]。……