Data-driven Modeling and Optimization for Energy-Smart Manufacturing

点击翻译按钮获取中文摘要

The concept of energy-smart manufacturing is to deliver customized products while simultaneously optimizing energy consumption, product performance (e.g., product functionality, quality and process variability) and equipment maintenance cost. Many previous attempts with similar goals focus on one objective at a time. In reality, because energy consumption, product performance, and equipment maintenance are correlated, decisions related to one aspect will often affect other aspects. This award will support fundamental research to discover the interactions among energy efficiency, product performance, and equipment maintenance. The knowledge thus gained will be used to optimize the manufacturing processes and maintenance operations for high energy efficiency and low cost. The methodology is intended to be widely applicable to many different types of manufacturing operations. The research results will be broadly disseminated to equip the current and future manufacturing engineers with the new methodologies through joint workshops with an industrial collaborator, technical training sessions, and case studies. Summer outreach workshops will be organized to engage high school students from underrepresented groups. The objective of this research is to customize data-driven modeling and optimization methodology to achieve high energy efficiency, excellent product performance, and low maintenance cost in manufacturing. A data-driven decision-making framework will be developed with the following intellectual merits: (1) dynamic models will be developed to quantify the product performance by considering equipment degradation effects and the change of product types; (2) a degradation index will be constructed from multivariate degradation measurements and a cumulative damage model will be used to predict equipment degradation; and (3) energy efficiency and maintenance cost will be optimized through customization of optimization algorithms at both the manufacturing system and the enterprise levels. These methodologies will be validated in a plasma spray coating process in the aero-engine manufacturing industry, and it will be designed to be broadly applicable to other high-energy-consumption manufacturing operations.