US 12,294,219 B2
Battery energy storage-based controller for improving microgrid power quality
Jaber Ali Alshehri, Dhahran (SA); Muhammad Khalid, Dhahran (SA); and Ahmed Hussain Alzahrani, Dhahran (SA)
Assigned to KING FAHD UNIVERSITY OF PETROLEUM AND MINERALS, Dhahran (SA); and King Abdullah City for Atomic & Renewable Energy (K.A. CARE), Riyadh (SA)
Filed by KING FAHD UNIVERSITY OF PETROLEUM AND MINERALS, Dhahran (SA)
Filed on Mar. 12, 2021, as Appl. No. 17/199,647.
Claims priority of provisional application 62/990,723, filed on Mar. 17, 2020.
Prior Publication US 2021/0296902 A1, Sep. 23, 2021
Int. Cl. H02J 3/32 (2006.01); G05B 6/02 (2006.01); G05B 13/02 (2006.01); H02J 3/38 (2006.01)
CPC H02J 3/32 (2013.01) [G05B 6/02 (2013.01); G05B 13/027 (2013.01); H02J 3/381 (2013.01); H02J 2203/20 (2020.01); H02J 2300/10 (2020.01); H02J 2300/24 (2020.01)] 13 Claims
OG exemplary drawing
 
1. A microgrid system, comprising:
a synchronous generator configured to convert mechanical power into electric power;
an energy storage system configured to store and supply electric power;
a controller configured to control operation of the energy storage system; and
a point of common coupling bus connecting the synchronous generator and the battery energy storage system, wherein
a controller parameter of the controller is determined based on a level of a disturbance using a trained artificial neural network (ANN) in response to occurrence of the disturbance in the synchronous generator,
wherein
the controller includes a proportional-integral (PI) controller and a hybrid DEO-ANN unit,
the controller parameter is a controller parameter of the PI controller, and
the trained ANN is obtained by the hybrid DEO-ANN unit using differential evolution optimization (DEO) and subsequent ANN training with input and output patterns of the DEO,
the hybrid DEO-ANN unit includes a DEO module, an ANN module, and a data storage module,
the data storage module stores therein a mathematical model representing the microgrid system,
the DEO module acquires the mathematical model from the data storage module and obtains a plurality of optimized controller parameters of the PI controller by applying a plurality of levels of disturbances, which are input patterns of a DEO algorithm, to a part of the mathematical model corresponding to a mechanical part of the synchronous generator and optimizing, for each level of disturbance, the controller parameter of the PI controller using the DEO algorithm in such a manner as to restore a system frequency and a system voltage of the MG system to preset normal operating limits, and
the ANN module trains an ANN using the plurality of levels of disturbances as input and the plurality of optimized controller parameters as output and stores a trained ANN in the data storage module,
wherein
the energy storage system is a battery energy storage system (BESS) including a rechargeable battery and a voltage source converter, the voltage source converter performing a bidirectional conversion between DC voltage and AC voltage,
the controller compares a system frequency of the microgrid system to a preset reference frequency and determines mismatch between the system frequency and the preset reference to generate an error signal that is input into the PI controller, wherein the PI controller outputs a control signal based on the inputted error signal, and
the PI controller whose controller parameter is optimized controls the voltage source converter with the control signal through a modulation index and a phase angle, and in response to receiving the control signal, the BESS supplies or absorbs at least active power to perform quick damping of oscillations to minimize the mismatch in the microgrid system.