Building agents for power trading agent competition (TAC)

• Main Author: Nguyen Le Hoang
• Other Authors: Bo An
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/148103
• Institution: Nanyang Technological University

Smart grid is an efficient and sustainable model that aims to integrate both suppliers and consumers of the energy supply chain. In order to design the smart grid for operations on large scale, high fidelity simulations for such a system must be studied to understand market and customer interactions. PowerTAC simulation platform aims to provide a way for researchers to study the mechanism of smart grids and strategies that maximizes efficiencies in the grid. In PowerTAC, brokers compete to generate profits through activities in the tariff, balancing and, wholesale market while maintaining the overall balance and stability of the grid as far as possible. Brokers who cause disruption in the grid’s operation are punished and those that help maintain the grid are rewarded. In this paper, we introduce BBNTU, an agent that adopts a disruptive tariff strategy that abuses the transmission capacity cost to push competitors to unsustainable tariffs. We price our tariffs at cost and encourages other brokers to price their tariffs lower, we then deliberately give away customers, earning early contract cancellation fees and only maintain a strict profitable market share. In the balancing market, we target hybrid storage consumer customers to earn from their consumption activities and storage capabilities. In the wholesale market, we adopt a seasonal autoregressive integrated moving average (SARIMA) to predict customer demand profile. The novelty lies in the modification which introduced localization weights on top of differencing which is modelled by an exponential distribution.