Q-learning Trained Virtual Inertial Control Strategy to Improve Battery Life in Microgrids Powered by Wind Turbines

This paper discusses the impact on the battery life of providing virtual inertia in a standalone microgrid, which is powered exclusively by wind turbines and battery units. In this microgrid batteries are responsible for providing virtual inertia through inverters to regulate the system frequency within the desired ranges. However, the energy spent to generate such virtual inertia could potentially increase the battery depth of discharge (DoD) and shorten its life. A Q-learning training process is designed for the battery energy storage system (BESS) to develop a virtual inertial control strategy so that the battery life is less affected. A simulation model of the microgrid is built in MATLAB and tested using the control strategy. According to the results, the trained virtual inertial control strategy could prevent battery lifetime loss and provide a sufficient portion of virtual inertia.

[1]  K. Mandal,et al.  Optimal Sizing of Battery-Ultracapacitor Hybrid Energy Storage Device in a Standalone Photovoltaic System , 2018, 2018 International Conference On Advances in Communication and Computing Technology (ICACCT).

[2]  Junwei Lu,et al.  Optimisation of battery energy storage capacity for a grid-tied renewable microgrid , 2017, 2017 IEEE Innovative Smart Grid Technologies - Asia (ISGT-Asia).

[3]  Alexis Kwasinski,et al.  Energy storage sizing for effective primary and secondary control of low-inertia microgrids , 2015, 2015 IEEE 6th International Symposium on Power Electronics for Distributed Generation Systems (PEDG).

[4]  Kankar Bhattacharya,et al.  Optimal Sizing of Battery Energy Storage Systems for Microgrids , 2014, 2014 IEEE Electrical Power and Energy Conference.

[5]  Alexis Kwasinski,et al.  Effects of Hurricanes Isaac and Sandy on Data and Communications Power Infrastructure , 2013 .

[6]  Naehyuck Chang,et al.  Optimal control of a grid-connected hybrid electrical energy storage system for homes , 2013, 2013 Design, Automation & Test in Europe Conference & Exhibition (DATE).

[7]  A. von Jouanne,et al.  A methodology to enable wind farm participation in automatic generation control using energy storage devices , 2012, 2012 IEEE Power and Energy Society General Meeting.

[8]  Jin Lin,et al.  Sizing of battery and supercapacitor in a hybrid energy storage system for wind turbines , 2012, PES T&D 2012.

[9]  Cody A. Hill,et al.  Development of a real-time testing environment for battery energy storage systems in renewable energy applications , 2011, 2011 IEEE Power and Energy Society General Meeting.

[10]  Charles J Mozina,et al.  Wind-Power Generation , 2011, IEEE Industry Applications Magazine.

[11]  K. C. Divya,et al.  Battery Energy Storage Technology for power systems-An overview , 2009 .

[12]  X.Y. Wang,et al.  Determination of Battery Storage Capacity in Energy Buffer for Wind Farm , 2008, IEEE Transactions on Energy Conversion.

[13]  Jan Pierik,et al.  Inertial response of variable speed wind turbines , 2006 .

[14]  Z.A. Styczynski,et al.  Benchmark for an Electric Distribution System with Dispersed Energy Resources , 2006, 2005/2006 IEEE/PES Transmission and Distribution Conference and Exhibition.

[15]  Peter Dayan,et al.  Q-learning , 1992, Machine Learning.

[16]  James F. Manwell,et al.  Lifetime Modelling of Lead Acid Batteries , 2005 .

[17]  Michael I. Jordan,et al.  On the Convergence of Stochastic Iterative Dynamic Programming Algorithms , 1994, Neural Computation.