44. Joint optimization of assembly line and on-site renewable energy implementation under uncertainty
Invited abstract in session MA-15: Applications to Logistics and Supply Chain Management, stream Combinatorial Optimization.
Monday, 8:30-10:00Room: Esther Simpson 1.08
Authors (first author is the speaker)
| 1. | Yuchen Li
|
| Economics and Management, Beijing University of Technology |
Abstract
Utilizing the on-site renewable energy (RE) resource (e.g.,photovoltaic (PV) and wind power (WP)) is one of the significant strategies among the initiatives that address the challenges of decarbonization. For example, Ford announced that the existing 4.64MWh per year solar installation at its Almussafes plant in Valencia, Spain, would be augmented by a further 3.76MWh per year.
In this paper, we aim to provide a decision framework for a manufacturer’s energy plan and production scheduling for its production system, namely, assembly line (AL) system. The decisions are to find the number of renewable energy apparatus (wind and solar), assign robots and tasks to the workstations. The objectives are the cycle time, total carbon footprint, the total cost. Given the output of each renewable energy apparatus is uncertain, the problem is formulated mathematically. A customized deep Q-learning algorithm (DQA) , which features an experiment replay and a fixed-Q target treatment, is proposed to solve the problem. A real-life case study concerning engine production is conducted and some management insights are obtained.
Keywords
- Combinatorial Optimization
- Multi-Objective Decision Making
- Manufacturing
Status: accepted
Back to the list of papers