system identification methods have been employed to estimate the power system inertia using the operational PMU data (with no external excitation signal) [22, 23].
1.2.1 Frequency Control
Preliminary efforts in the field of power grid frequency regulation are reported in [24]. Subsequently, an IEEE working group prepared some standard definitions of significant terms and concepts on power system frequency control [25]. Considering the physical constraints and to cope with the advances in technologies and the changed system environment, dynamic modeling developments, security constraints, and communication delays, as well as modifications on the frequency control definitions, have been discussed over the years [26–30]. A comprehensive survey and exhaustive bibliography on frequency control up to 2014 are given in [31, 32].
Frequency control analysis, frequency response modeling, nonlinearity and uncertainty presentation, specific applications, frequency bias calculation, control performance standards, load characteristics impacts, and parameters identification are presented in several documents [2932–42]. A Considerable research on the time‐delayed system is contained in [4, 30]. In addition, regarding parametric uncertainty, several self‐tuning, adaptive, and robust control strategies are widely applied for power grid LFC system synthesis over the years [4,43–51].
Dynamic impacts of intermittent DGs and high penetration of RESs on power grids frequency response are discussed in [3252–56]. A low inertia can negatively affect the grid frequency dynamic performance and stability. A number of recent works have suggested the application of inverter‐based virtual inertia emulators to improve frequency stability and frequency response performance [57–61]. Furthermore, numerous research works have been recently focused on the use of DGs, RESs, MGs, electric vehicles, and storage devices to provide frequency control supports in the power grids [62–69]. Providing frequency control support via controllable loads and smart load technologies using the concepts of demand response (DR) is discussed in [41, 42,70–76]. Two recent works in this area are [77, 78], that discuss the impact of a high integration of MGs on the frequency control of power systems, and propose a decentralized stochastic frequency control of MGs.
PMU‐based/data‐driven online tuning frequency control approach is not addressed in the abovementioned worldwide published works. In most cases, the secondary frequency control is designed using conventional frequency response model, which is very difficult to realize in a modern power grid with a highly variable structure and penetration of DGs/RESs.
1.2.2 Voltage Control
Since 1990s, supplementary control of generator excitation systems, static var compensator (SVC), and high voltage direct current (HVDC) converters is increasingly being used to solve power system oscillation problems [7]. There has also been a general interest in the application of power electronics‐based controllers known as flexible alternating current transmission system (FACTS) controllers for the damping of system oscillations [79]. Following several power system collapses worldwide [80–82], in 1990s, voltage stability has attracted more research interests.
Recently, following the development of PMUs, communication channels, and digital processing, wide‐area power system stabilization and control have become areas of interest [83, 84]. A typical generic of different voltage control levels is discussed in [85]. Optimal voltage control has long been successfully implemented in power systems, including the three‐level hierarchical automatic voltage control in Europe [86–88], and the adaptive zone division method in China [89].
A supervisory voltage control strategy for large‐scale solar photovoltaic (PV) integration in power network is proposed in [90, 91] to enhance the voltage stability. A survey of methods, mostly based on PMU data, for long‐term voltage instability detection is given in [92]. In [93], a two‐stage distributed voltage control scheme is proposed. The first stage is the local control of each DG based on sensitivity analysis, and the second stage acquires reactive power support from other DG units. In [94], a consensus‐based cooperative control is proposed to regulate voltage by coordinating electric cars and active power curtailment of PVs. In [95], a distributed voltage stability assessment considering DG units is developed based on distributed continuation power flow. Coordinated voltage control is a technique which provides voltage control by means of adjusting, sequencing, and timing various kinds of controllers within a system. Some relevant works are reported in [96–98].
As mentioned above, several PMU‐based voltage control methodologies have been reported worldwide; however, mostly presented a voltage recovery approach in an off‐normal or emergency condition. Among existing three hierarchical levels of voltage control (primary, secondary, and tertiary controls), only few works are mainly focused on optimal supervisory on secondary voltage control, which is required to coordinate adjustment of the set‐points of the existing voltage controllers. In this regard, the online adaptive tuning of available voltage control systems in a power grid with high integration of DGs/RESs is not well addressed. Furthermore, the overlap between voltage dynamics and frequency/active power as well as rotor angle dynamics in a modern power grid has not been highlighted in the published reports.
1.2.3 Oscillation Damping
Traditionally, the power system oscillations are damped through the generator local controllers, such as the exciter and governor, which are designed to ensure only the local stability of the generator (1–2 Hz). In order to increase the stability of the system, PSSs and power electronic converter‐based FACTS are added into the grid [99–101]. In a broader context, the power system oscillation problem has also been related to voltage stability. The control interaction is discussed in [102, 103]. The exploitation of the wide‐area measurements, provided by PMUs, for monitoring and controlling the power system led to the introduction of the wide‐area monitoring and control (WAMC) systems [104]. The advent and application of synchronized measurement technology has enabled the detection and observation of poorly damped oscillations (such as the inter‐area modes) and became the backbone for more development of the WAMC systems [105]. Inter‐area oscillations are characterized by low frequency (0.2–1 Hz) and occur when generators of one group swing against generators of another group [106]. Integration of RESs into the WAC scheme for damping power oscillations is discussed in [107, 108]. The utilization of a networked control system model for the WAC design, according to linear matrix inequality techniques, is proposed in [111]. Furthermore, Ref. [110] presents a WAC design, based on particle swarm optimization, for improving the performance of the power system through the control of wind farms.
More specifically, WAC aims to utilize the synchronized phasor measurements in order to provide coordination signals to the local controllers, making them capable of damping effectively all the inter‐area oscillations [100]. In the literature, various works deal with the development of a WAC system. The proposed WAC schemes are segregated mainly according to the components of the power system that the WAC is intended to coordinate [3, 110]. Multiple control methodologies have been developed for damping the inter‐area oscillations deploying a WAMC. In [83], a decentralized/hierarchical architecture for wide‐area damping control using PMU remote feedback signals was discussed. References [100, 111] proposed the design of wide‐area damping controllers that provide supplementary damping control to synchronous generators (SGs). A networked control system model for wide‐area closed‐loop power systems is applied in [109]. A power oscillation damping controller is introduced in [112] based on a modal linear quadratic Gaussian methodology. A combination of controlling SGs and renewable sources in order to increase the overall damping capability of the system is shown in [101, 107, 108, 113]. Few LPV control solutions to power oscillation damping are proposed that use either a low‐order first principle model of the system [114] or a reduced‐order parametric LPV identified model [115].
In comparison of frequency and voltage control, a higher number of reports have been published in PMU‐based oscillation damping (rotor angle control) field. However, most of the reported approaches
