Info-Gap Approach to Manage GenCo's Trading Portfolio With Uncertain Market Returns

摘要

An independent generation company (GenCo) secures its future trading position by managing its portfolio among multiple trading options. Future returns of these trading options are not known during decision making and are traditionally estimated using probabilistic or fuzzy methods. Quantifying such uncertainty of market returns by conventional methods does not reflect the information gap existing between estimated and actual market returns. Based on quantification of this information gap, the paper proposes GenCo's portfolio optimization using a non-probabilistic Information Gap Decision Theory (IGDT). This framework comprehensively models GenCo's behavior in deciding its trading strategy. Considering GenCo's risk-averse behavior, the framework provides decisions that are robust towards losses, while considering its risk-seeking behavior the framework offers opportunity to capture windfall gains. The proposed approach has been validated through practical case study of PJM market.