项目名称: 基于工业大数据挖掘的复杂产品总完工时间动态预测
项目编号: No.51505357
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 机械、仪表工业
项目作者: 常建涛
作者单位: 西安电子科技大学
项目金额: 20万元
中文摘要: 复杂产品的总完工时间预测直接影响到制造企业生产计划编制的合理性,也是工业4.0背景下制造企业迈向“预测型制造”的重要环节。本项目拟突破以往总完工时间研究局限于车间工序级尺度的限制,开展基于工业大数据挖掘的复杂产品尺度的总完工时间动态预测研究。首先,对工业大数据进行预处理,在此基础上运用大数据挖掘方法进行复杂产品总完工时间影响因素、约束条件的挖掘,探索总完工时间与影响因素之间的关联机理;基于影响因素及约束条件数据和总完工时间历史数据,建立复杂产品总完工时间预测模型,探索将动态特性融入预测模型的规律,进行模型的动态改进和优化;最后,运用动态预测模型进行复杂产品总完工时间的求解和误差分析,并在企业进行验证。通过本项目的研究,能够使得企业更加合理地编制复杂产品生产计划,为企业向“预测型制造”转型提供技术借鉴。
中文关键词: 总完工时间预测;大数据挖掘;预测型制造;工业大数据
英文摘要: Total completion time prediction of complex products can directly affect the rationality of production planning of manufacturing enterprises, and it is one of the key technologies for manufacturing enterprises’ predictive manufacturing under industry 4.0 environment. We intend to break through the restriction of workshop process level and conduct the research of dynamic prediction for total completion time of complex products based on industrial big data mining. First, industrial big data will be pretreated, the factors and constraints for total completion time prediction of complex product will be mined and analyzed using big data mining method, the relationship mechanism between Total completion time and the factors will be explored. Prediction model will be proposed using artificial neural networks or deep learning algorithm based on history industrial big data, the law of dynamic characteristics merged into prediction model will be explored, and the prediction model should be improved and optimized. Finally, the total completion time for complex products will be calculated using prediction model, the errors of prediction values will be analyzed, and the prediction model will be applied into a manufacturing enterprise. The goal of this project is to make the production planning more reasonable and provide technologies for manufacturing enterprises’ predictive manufacturing.
英文关键词: Total completion time prediction;Big data mining;Predictive manufacturing;Industrial big data