Build software agents for power trading agent competition

• Main Author: Tan, Zhi Xiang
• Other Authors: Bo An
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Computer science and engineering
• Online Access: http://hdl.handle.net/10356/62610
• Institution: Nanyang Technological University

Due to the increase in population with higher demand for energy, sustainable energy systems for the future are being researched on. Energy balancing is critical to ensure proper functioning of such system. Power Trading Agent Competition (PowerTAC) is a risk-free simulated environment that runs and tests agents aiding in the research of automation technologies to increase the efficient. This research aims to create an agent, agent NTU, which will act as a broker to trade energy. Several design considerations will be discussed with proposed solutions, which include the use of moving average and Markov Decision Making Process to design the agent while ensuring its adaptability to the changing environment. Detailed implementations of the agent are documented in this report.