Hierarchical Control for Optimal and Distributed Operation of Microgrid Systems

Research output: Book/ReportPh.D. thesisResearch

Abstract

The distributed generation, storage and consumption, as well as the sustainability consideration prompt a revolution to the existing electric power grid. Microgrids (MG) concept has been proposed to liberate the operation of each distribution system fraction, forming in that way a flexible and sustainable grid. To achieve autonomous operation for MGs, hierarchical control was proposed with primary, secondary and tertiary control levels differentiated. In conventional hierarchical scheme, primary control is issued for power sharing, secondary control takes care of power quality, and tertiary control manages the power flow with external grids, while the economic and optimal operation of MGs is not guaranteed by applying the existing schemes.

Accordingly, this project dedicates to the study of real-time optimization methods for MGs, including the review of optimization algorithms, system level mathematical modeling, and the implementation of real-time optimization into existing hierarchical control schemes. Efficiency enhancement in DC MGs and optimal unbalance compensation in AC MGs are taken as the optimization objectives in this project. Necessary system dynamic modeling and stability analysis are also conducted in order to ensure safe operation during the optimization procedure.

In addition, as the secondary and tertiary controls require global information to perform the functions, they are usually implemented in centralized fashion. In this sense the communication links are required from the central unit to each local unit, a single point of failure in the central controller may jerpodize the safety of the whole system, and the flexibility of the system is limited. Consequently, this project proposes the application of dynamic consensus algorithm (DCA) into existing hierarchical control scheme aiming at achieving fully distributed secondary and tertiary controls. The primary purpose of DCA is to allow a number of distributed units to obtain global information that is necessary for the control layers. Also by applying DCA the communication links are only needed between nearby units. Based on these features, this project proposes distributed control schemes for power quality regulation in three-phase AC MGs, as well as voltage/current control in DC MGs.

Moreover, in order to provide analytical method for evaluating the system stability with such kind of distributed control scheme, a discrete-time domain modeling method is proposed to establish an accurate system level model. Taking into account the different sampling times of real world plant, digital controller and communication devices, the system is modeled with these three parts separately, and with full consideration of the underlying communication features (sampling time, topology, parameters, etc.). System dynamics and sensitivity analysis are conducted based on the proposed model.

A MG central controller is also developed based on the experimental system in the intelligent MG lab in Aalborg University for providing a comprehensive platform for MG related study purposes. LabVIEW software is used as the programming language, UDP/IP based Ethernet communication links are built between the central controller and each setup. System performance is shown by experimental examples.

Finally, to verify the effectiveness and performance with the proposed control schemes and modeling methods, experimental and hardware-in-the-loop simulation studies are conducted in the intelligent MG lab. The successful realization of online optimization and distributed control functions is expected to be able to provide guidance for real world implementation of similar approaches. The generalized discrete-time modeling method, with verified correctness and accuracy, can give insight view of the distributed control scheme and the impact of communication part on system dynamics.
Close

Details

The distributed generation, storage and consumption, as well as the sustainability consideration prompt a revolution to the existing electric power grid. Microgrids (MG) concept has been proposed to liberate the operation of each distribution system fraction, forming in that way a flexible and sustainable grid. To achieve autonomous operation for MGs, hierarchical control was proposed with primary, secondary and tertiary control levels differentiated. In conventional hierarchical scheme, primary control is issued for power sharing, secondary control takes care of power quality, and tertiary control manages the power flow with external grids, while the economic and optimal operation of MGs is not guaranteed by applying the existing schemes.

Accordingly, this project dedicates to the study of real-time optimization methods for MGs, including the review of optimization algorithms, system level mathematical modeling, and the implementation of real-time optimization into existing hierarchical control schemes. Efficiency enhancement in DC MGs and optimal unbalance compensation in AC MGs are taken as the optimization objectives in this project. Necessary system dynamic modeling and stability analysis are also conducted in order to ensure safe operation during the optimization procedure.

In addition, as the secondary and tertiary controls require global information to perform the functions, they are usually implemented in centralized fashion. In this sense the communication links are required from the central unit to each local unit, a single point of failure in the central controller may jerpodize the safety of the whole system, and the flexibility of the system is limited. Consequently, this project proposes the application of dynamic consensus algorithm (DCA) into existing hierarchical control scheme aiming at achieving fully distributed secondary and tertiary controls. The primary purpose of DCA is to allow a number of distributed units to obtain global information that is necessary for the control layers. Also by applying DCA the communication links are only needed between nearby units. Based on these features, this project proposes distributed control schemes for power quality regulation in three-phase AC MGs, as well as voltage/current control in DC MGs.

Moreover, in order to provide analytical method for evaluating the system stability with such kind of distributed control scheme, a discrete-time domain modeling method is proposed to establish an accurate system level model. Taking into account the different sampling times of real world plant, digital controller and communication devices, the system is modeled with these three parts separately, and with full consideration of the underlying communication features (sampling time, topology, parameters, etc.). System dynamics and sensitivity analysis are conducted based on the proposed model.

A MG central controller is also developed based on the experimental system in the intelligent MG lab in Aalborg University for providing a comprehensive platform for MG related study purposes. LabVIEW software is used as the programming language, UDP/IP based Ethernet communication links are built between the central controller and each setup. System performance is shown by experimental examples.

Finally, to verify the effectiveness and performance with the proposed control schemes and modeling methods, experimental and hardware-in-the-loop simulation studies are conducted in the intelligent MG lab. The successful realization of online optimization and distributed control functions is expected to be able to provide guidance for real world implementation of similar approaches. The generalized discrete-time modeling method, with verified correctness and accuracy, can give insight view of the distributed control scheme and the impact of communication part on system dynamics.
Original languageEnglish
PublisherDepartment of Energy Technology, Aalborg University
Number of pages58
ISBN (Print)978-87-92846-66-2
Publication statusPublished - Oct 2015
Publication categoryResearch

    Research areas

  • Microgrids, Hierarchical control, Optimization, Distributed control, Dynamic consensus algorithm, Power quality, Efficiency, System modeling, Microgrid central controller

Download statistics

No data available
ID: 220600049