MPC Based Centralized Voltage and Reactive Power Control for Active Distribution Networks

Hoa Minh Nguyen, Jose L. Rueda, Aleksandra Lekic, Hoan Van Pham

Research output: Contribution to journalArticleScientificpeer-review

2 Downloads (Pure)


The paper presents an approach for online centralized control in active distribution networks. It combines a proportional integral (PI) control unit with a corrective control unit (CCU), based on the principle of Model Predictive Control (MPC). The proposed controller is designed to accommodate the increasing penetration of distributed generation in active distribution networks. It helps in continuously satisfying the reactive power requirements of the transmission system operators (TSOs), while maintaining an acceptable voltage profile in the active distribution network, and simultaneously minimizing the total active power losses. The controller also ensures compliance to operation requirements of distribution network operators (DNOs). By replacing the full load flow (LF) calculation with sensitivities, derived from a linearized model of the network, the controller can work in real-time applications. Moreover, the computational burden of the proposed controller is reduced since the CCU is activated only when a voltage violation or considerable change of operation condition occurs. The performance of the proposed controller is demonstrated on a 11-kV test network with 75 buses and 22 distributed generators.

Original languageEnglish
Number of pages11
JournalIEEE Transactions on Energy Conversion
Publication statusPublished - 2021


  • Active distribution networks
  • Artificial neural networks
  • artificial neural networks
  • distributed generation
  • Distribution networks
  • model predictive control
  • Optimization
  • Reactive power
  • reactive power management
  • Sensitivity
  • smart grids
  • Voltage control

Fingerprint Dive into the research topics of 'MPC Based Centralized Voltage and Reactive Power Control for Active Distribution Networks'. Together they form a unique fingerprint.

Cite this