Skills

Python

80%

Simulink/Matlab

99%

OPRT/DSpace

70%

Experience

 
 
 
 
 
June 2018 – Present
San Jose, CA

Postdoc Reseacher

GEIRI North America

Dr. Duan has been employed at GEIRI North America since June of 2018. In his position, he is responsible for control algorithm developments. Particularly, he is working on artificial intelligence based power grid autonomous control system using the cutting-edge deep reinforcement learning technology from DeepMind. Responsibilities include:

  • AI
  • Control Theory
  • Deep Reinforcement Learning
 
 
 
 
 
May 2016 – August 2016
Worcester, MA

Associate Design Engineer

Conti Corp Inc.

In this intern position, Dr. Duan performed various duties:

  • Collaborated in designing a hybrid PV panel deployment solution and reduced the total cost by over $9 million, with consideration of shading, power balancing and converter matching constraints.
  • Managed and monitored timeline of the program， generated daily status report and presented to the leadership.
  • Supervised and interacted with 12 subcontracting companies to achieve the project milestones.
  • Mapped over 70,000 barcodes including solar panels, inverters, and optimizers into 1,603 subsystem drawings with AutoCAD, and composed placed-in-order letters and safety reports.
 
 
 
 
 
January 2015 – May 2018
Bethlehem, PA

Research Assistant

Lehigh University

In this research position, Dr. Duan performed various duties:

  • Developed an adaptive microgrid management systems. The secondary and primary controllers are designed in a decentralized way to realize proper load sharing and plug-and-play functions with unknown system parameters. The proposed control can guarantee that the desired generation references of tertiary control can be accurately achieved.
  • Reduced the overshoot of transient voltage/current up to 30% for AC/DC microgrids with inverter-interfaced DGs based on advanced control designs. Performed the switch-level real time simulation using RT-Lab.
  • Resolved the impact of pulsed power load in the shipboard power system by introducing and applying the Zero-Sum Game Theory. Accomplished the demonstration with both real time simulation and hardware-in-the-loop experiments.
  • Achieved optimal control of parallel uninterruptible power supply (UPS) system by designing a Neuro Network algorithm to train the unknown system dynamics.

Recent Publications

A novel distributed control scheme is proposed for the inverter-interfaced microgrids in this paper..
Preprint PDF DOI

In this paper, a neural-network-based online optimal control method is proposed for ultra-capacitors energy storage system in hybrid …
Preprint PDF DOI

This paper proposes a resilient-backpropagation-neural-network-(Rprop-NN) based algorithm for Photovoltaic (PV) maximum power point …
Preprint PDF DOI

Considering that conventional periodic control solutions have placed an excessive burden on the cyber infrastructure, in this paper, …
Preprint PDF DOI

This paper presents a novel decentralized output constrained control algorithm for single-bus DC microgrids.
Preprint PDF DOI

This paper presents a distributed control solution for inverter-interfaced microgrids.
Preprint PDF DOI

This paper targets at wide-area damping control of WAPS under both physical and cyber uncertainties.
Preprint PDF DOI

This paper presents a novel fault location method for single-phase microgrids.
Preprint PDF DOI

Projects

Grid Mind

Grid Mind is a novel autonomous control framework developing for secure operation of power grids based on cutting-edge artificial intelligence technologies.

Tags

Contact

  • jiajunduan
  • 250 W Tasman Dr STE100, San Jose, California, 95134, USA
  • Monday 10:00 to 13:00
    Wednesday 09:00 to 10:00
    Otherwise email to book an appointment