Date of Award

5-2016

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Automotive Engineering

Committee Member

Dr. Laine Mears, Committee Chair

Committee Member

Dr. Joachim Taiber

Committee Member

Dr. Kumar Venayagamoorthy

Committee Member

Dr. Paul Venhovens

Abstract

World energy consumption has continued increasing in recent years. As a major consumer, industrial activities uses about one third of the energy over the last few decades. In the US, automotive manufacturing plants spends millions of dollars on energy. Meanwhile, due to the high energy price and the high correlation between the energy and environment, manufacturers are facing competing pressure from profit, long term brand image, and environmental policies. Thus, it is critical to understand the energy usage and optimize the operation to achieve the best overall objective. This research will establish systematic energy models, forecast energy demands, and optimize the supply systems in manufacturing plants. A combined temporal and organizational framework for manufacturing is studied to drive energy model establishment. Guided by the framework, an automotive manufacturing plant in the post-process phase is used to implement the systematic modeling approach. By comparing with current studies, the systematic approach is shown to be advantageous in terms of amount of information included, feasibility to be applied, ability to identify the potential conservations, and accuracy. This systematic approach also identifies key influential variables for time series analysis. Comparing with traditional time series models, the models informed by manufacturing features are proved to be more accurate in forecasting and more robust to sudden changes. The 16 step-ahead forecast MSE (mean square error) is improved from 16% to 1.54%. In addition, the time series analysis also detects the increasing trend, weekly, and annual seasonality in the energy consumption. Energy demand forecasting is essential to production management and supply stability. Manufacturing plant on-site energy conversion and transmission systems can schedule the optimal strategy according the demand forecasting and optimization criteria. This research shows that the criteria of energy, monetary cost, and environmental emission are three main optimization criteria that are inconsistent in optimal operations. In the studied case, comparing to cost-oriented optimization, energy optimal operation costs 35% more to run the on-site supply system. While the monetary cost optimal operation uses 17% more energy than the energy-oriented operation. Therefore, the research shows that the optimal operation strategy does not only depends on the high/low level energy price and demand, but also relies on decision makers’ preferences. It provides not a point solution to energy use in manufacturing, but instead valuable information for decision making. This research complements the current knowledge gaps in systematic modeling of manufacturing energy use, consumption forecasting, and supply optimization. It increases the understanding of energy usage in the manufacturing system and improves the awareness of the importance of energy conservation and environmental protection.

Share

COinS