University of Groningen
Distributed Control, Optimization, Coordination of Smart Microgrids
Silani, Amirreza
DOI:
10.33612/diss.156215621
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2021
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Silani, A. (2021). Distributed Control, Optimization, Coordination of Smart Microgrids: Passivity, Output Regulation, Time-Varying and Stochastic Loads. University of Groningen.
https://doi.org/10.33612/diss.156215621
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
C
H
A
P
T
9
Conclusions and Future Research
This thesis provides a framework to design controller schemes and energy man-agement strategies to address the problem of time-varying or stochastic loads and renewable generations. In this chapter, we discuss the main conclusions and find-ings presented of the thesis. We also provide suggestions for future research in this chapter.
9.1
Conclusions
Motivated by the inadequacy of the existing control for power systems affected by time-varying or stochastic uncontrolled power injections such as loads and the increasingly widespread renewable energy sources, this thesis proposes control schemes to address the frequency regulation problem for AC network, the voltage regulation problem for DC networks, an energy management strategy with stochastic loads and a microgrid optimal control problem considering the social behavior of the EV drivers. In the following, we summarize the conclusions of each chapter.
In Chapter 3, stochastic dynamics for the impedance (Z), current (I) and power (P) components of ZIP loads in DC power network have been considered. More precisely, each component of the ZIP load has been modeled as the sum of an unknown constant and the solution to a stochastic differential equation describing the load dynamics. Then, the stochastic passivity of the considered system has been verified. Indeed, sufficient conditions for the stochastic passivity of the open-loop
system have been presented, facilitating the interconnection with passive control systems. Then, a distributed control scheme has achieved average voltage regulation and current sharing and guaranteed the asymptotic stochastic stability of overall system. Therefore, asymptotic stochastic stability of the overall network including stochastic and unknown constant loads has been guaranteed and average voltage regulation and current sharing have been achieved.
In Chapter 4, time-varying dynamics for the load components of a DC power net-work have been considered. Indeed, each load component has been described as the output of a large class of nonlinear dynamical exosystem. Next, the voltage control in DC networks including time-varying loads has been formulated as a standard output regulation problem. Then, a proposed control scheme designed based on the output regulation methodology has achieved voltage regulation and guaranteed the stability of the overall network. Thus, the stability of overall network affected by time-varying loads has been guaranteed and the voltage regulation has been achieved.
In Chapter 5, time-varying dynamics with constant uncertainties for the load components of a DC power network have been considered. Indeed, superposition of time-varying and uncertain constant ZIP load components has been considered where the time-varying components of loads are the output of dynamical exosystems. Next, the voltage control problem in DC networks with time-varying and uncertain constant loads has been formulated as a robust output regulation problem. Then, proposed control schemes designed based on the robust output regulation methodology have achieved voltage regulation and guaranteed the local robust stability of the overall network including time-varying and uncertain constant ZIP loads, where the local nature of the controllers is due to use of linearization for control design. Moreover, a proposed control scheme has achieved voltage regulation and ensured the global robust stability of the overall network including time-varying and uncertain constant ZI loads. Hence, the local and global robust stability of the overall network affected by time-varying and uncertain constant loads have been guaranteed and voltage regulation has been achieved.
In Chapter 6, time-varying dynamics for the uncontrolled power injections (the difference between the power generated by the renewable energy sources and the one absorbed by the loads) of an AC power network have been considered. Indeed, the uncontrolled power injections have been described as the output of a large class of nonlinear dynamical exosystem. Then, the output regulation theory has been used for the design and analysis of control schemes for nonlinear power networks affected by time-varying renewable energy sources and loads. More precisely, a proposed controller designed based on the classical output regulation theory has provably regulated the frequency deviation to zero even in presence of time-varying uncon-trolled power injections. Then, besides merely controlling the frequency deviation, a
proposed controller designed based on the approximate output regulation has solved the approximate OLFC problem in presence of time-varying uncontrolled power injections while ensured the stability of the overall network. Therefore, the stability of the overall network affected by time-varying renewable energy sources and loads has been ensured and LFC and approximate OLFC have been achieved.
In Chapter 7, a distributed energy management strategy has been proposed with stochastic loads. In this approach, the underlying power distribution network and its constraints have been considered while existing methods have ignored the underly-ing power and assumed that there is a perfect prediction of load. More precisely, the EMS problem in microgrids has been formulated as a nonconvex optimization prob-lem taking into account the loads, power flows, and system operational constraints in a distribution network such that the costs of the DGs, DSs and energy purchased from the main grid are minimized and the customers’ demanded loads are provided where the loads are considered stochastic generated by a time-homogeneous Markov chain. Then, the nonconvex constraints have been relaxed to obtain a convex optimization problem according to the conditions provided in [110, 111] for the exactness of this convexification. Next, the centralized optimization problem has been decomposed into a distributed problem via the Predictor Corrector Proximal Multiplier (PCPM) method proposed by [122] in order to handle the customers’ privacy, communication challenges, and high computational burdens of centralized optimization. Indeed, the LC of each unit optimizes the cost function of it and sends an optimal schedule to the MGCC and the MGCC optimizes the total cost function of grid based on the optimal schedule received from the LCs and its constraints. Also, to produce stochastic load, random data have been generated based on statistical analysis and transformation process of the uncertainty over time by using Markov chain rule. Hence, a distributed energy management strategy considering underlying power distribution network and its constraints with stochastic loads has been proposed.
In Chapter 8, a dynamical BtG framework composed of a TSO, DSO networks, and buildings has been utilized where the EVs in each building can follow or unfollow a smart charging contract. Since humans make the decision to follow or unfollow the smart charging contract, a data set of EVs’ social behavior to model such behaviors in the power network has been used. The social behavior of EV drivers, i.e., the extent to which EV drivers are willing to use smart charging, is a vital factor in smart charging which we have taken into account. Indeed, we have investigated whether EV drivers are willing to have their batteries used as storage devices and which factors affect their willingness. Then, in order to regulate the frequency and minimize the costs of different units in the power network, an optimal control problem has been introduced. Finally, this optimal control problem has been tackled through a MPC based method. Thus, the frequency control has been achieved, the costs of
different units have been minimized and the effects of the willingness of EV drivers to use smart charging in the power network have been investigated.
9.2
Future Research
This thesis provides a framework to design controllers achieving frequency regulation in AC networks and voltage regulation in DC networks with time-varying or stochastic uncontrolled power. Also, it provides an optimization framework to design an energy management strategies with stochastic loads and investigating the social behavior of EV drivers in power networks. However, there are some possibilities to extend the results of this thesis discussed as follows:
• The main control objective in Chapters 4 and 5 are the regulation of the voltage in the DC power network with time-varying load components (see Objective 4.1). However, due to different generation capacities, it is reasonable to require that the total load demand of the microgrid is fairly shared among all the different generation units, i.e., achieving current sharing objective (see Objective 3.1). Since achieving current sharing objective, does not generally allow to perform also voltage regulation, the second objective can be defined as average voltage regulation (see Objective 3.2). Thus, the output regulation and robust output regulation methodologies introduced in Chapter 4 and Chapter 5, respectively, can be investigated for use in obtaining current sharing and average voltage regulation objectives in the DC network addressing the problem of time-varying load components.
• In Chapter 6, a classical output regulation and an approximate output regula-tion method have been used to obtain load frequency control (see Objective 6.1) and approximately optimal load frequency control objectives (see Objective 6.2), respectively. However, the robust output regulation methodologies introduced in Chapter 5 can be investigated for use in obtaining optimal load frequency control objective addressing the problem of time-varying and uncertain uncon-trolled power injections (i.e., the difference between the power generated by the renewable energy sources and the one absorbed by the loads). Although, the latter methodology requires the solution to the regulator equation (6.16) which is generally challenging, considering some assumptions on the exosys-tem model (6.5) can be investigated for obtaining an analytical solution to (6.16). Also, an approximate solution to the regulator equation can be obtained via the approximation methods proposed for instance in [77, Chapter 4], [133, 134]. Then, the approximate solution can be investigated for use in the robust output
regulation methodologies introduced in Chapter 5 to address optimal load frequency control problem.
• In Chapter 7, in order to convexify the problem (7.29), we have relaxed the constraint (7.25) to (7.30). The represented relaxation is exact if the injected power is not large and the bus voltage is near its nominal value [110]. How-ever, this relaxation may make the solution suboptimal. Therefore, nonconvex optimization methods can be investigated for use in solving the nonconvex problem (7.29) without such relaxation.
• In Chapter 8, the willingness of EV drivers to use smart charging can be investi-gated for use in modeling probability distribution functions. Then, a large-scale human-in-loop stochastic hybrid MPC method can be investigated for solving the problem (8.34), where the the ON or OFF mode of EVs in the simulation can be generated by Markov chain. Moreover, the data set used in Chapter 8 contains the knowledge of the range anxiety, environmental self-identity and willingness of EV drivers to have the EV battery used. Using the questionnaire results, these items can be investigated for use in modeling probability distribu-tion funcdistribu-tions. Then, their effects on the stability of the power network can be investigated via a stochastic hybrid MPC method solving the problem (8.34).
