Process analytics to predict arc loss in an electric arc furnace**

预测电弧炉电弧损耗的过程分析**

• 批准号: 536692-2018
• 负责人: Gopaluni, Bhushan
• 金额: $ 1.82万
• 依托单位: University of British Columbia
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2018
• 资助国家: 加拿大
• 起止时间: 2018-01-01 至 2019-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=652174
• 关键词: Process; analytics; predict; arc; loss

Smelting is an energy intensive operation in metallurgical processes where ore is refined into base metal in units such as an electric arc furnace that can draw up to 80MW of power. Disruptions to the furnace operation, such as loss of the electric arc, can result in expensive costs associated with lost production time and reduced energy efficiency. In this project we aim to combine process analytics, machine learning and large amounts of historical data to develop an inferential sensor that can predict loss of the furnace arc. Specifically, our goal is to study patterns in upstream metallurgical and electrical variables in order to develop a supervised learning classification model that can inform operators when there is a high probability of the onset of arc loss. Operators can then use this information to take corrective actions and perform predictive maintenance to ensure that stable operation and constant production is maintained.

冶炼是冶金过程中的能源密集型操作，其中矿石在电弧炉等装置中被精炼成贱金属，电弧炉的功率可达80兆瓦。炉操作的中断，例如电弧的损失，可能导致与损失的生产时间和降低的能量效率相关的昂贵成本。在这个项目中，我们的目标是结合联合收割机过程分析，机器学习和大量的历史数据，开发一种推理传感器，可以预测电弧的损失。具体来说，我们的目标是研究上游冶金和电气变量的模式，以开发一个监督学习分类模型，可以通知操作员时，有一个很高的发生电弧损失的可能性。然后，操作员可以使用这些信息采取纠正措施并执行预测性维护，以确保保持稳定的操作和恒定的生产。