loading page

Deep Learning Assisted Operation Model for Interconnected Networks Based on State Logic
  • +5
  • Olatunji Adeyanju ,
  • Manohar Chamana ,
  • Stephen Bayne ,
  • Luciane Canha ,
  • Pierluigi Siano ,
  • Mahtab Murshed ,
  • Mauricio Sperandio ,
  • Ayda Demir
Olatunji Adeyanju
Texas Tech University

Corresponding Author:[email protected]

Author Profile
Manohar Chamana
Author Profile
Stephen Bayne
Author Profile
Luciane Canha
Author Profile
Pierluigi Siano
Author Profile
Mahtab Murshed
Author Profile
Mauricio Sperandio
Author Profile
Ayda Demir
Author Profile

Abstract

This study presents a deep learning-assisted operation model for interconnected autonomous networks based on state logic to address dynamic local and global objectives simultaneously. The interconnected independent power systems depend on each other through common power pool sharing to maintain a reliable power delivery to the end users. The usual practice is to pursue a global objective for the entire interconnected systems. However, intermittent energy resources create unique local operation challenges for each system, compelling their operators to solve dynamic local objectives in addition to the global objective to enhance their operations. Consequently, a novel distributed operation approach incorporating Deep Learning and Mixed Integer Non-Linear Programing is formulated in this study to address both dynamic local and global objectives for the entire system. The resulting model is solved in two stages considering each systemâ\euro™s state logic, demand response, load management, and market situation. The performance of the developed model is demonstrated on a modified IEEE 30 bus system with two interlinked autonomous networks. The results presented show that the novel model is more effective in reducing the overall system operation cost and amount of load shedding when compared to its centralized counterpart benchmark model.