An overview for emerging control issues in microgrids: Challenges and solutions

CHEN Zhe, WANG Yanbo

Department of Energy Technology, Aalborg University, 9220, Aalborg, Denmark

[Abstract] Control system is one of significant components to ensure reliable, secure and economical operation of microgrids in either grid-connected or stand-alone mode.The development of microgrid technologies has led to new emerging controlconcepts and challenges.This paper presents an overview for emerging control issues in microgrids, and displays the main contributions from this research group in these aspects, including power management among multiple paralleled inverters for islanded microgrids, efficiency improvement of multi-converter microgrid with different power sharing profile and load profile, etc, nonlinear characteristics caused by dead-time nonlinearity of inverter, soft-saturation nonlinearity of filter inductor, long-term operation aging of inverter, etc and long-term operation reliability enhancement.The technique details regarding these emerging control issues in microgrids are discussed, and the relevant research results from this research group are given.Finally, this paper identifies and proposes future trends for potential control issues and solutions.

[Key words]microgrid；control issue；power management；efficiency； nonlinear characteristics；reliability

1 Introduction

With the increasing penetration of renewable energies, the power electronic-fed power systems such as microgrid and virtual power plants, etc.are becoming attractive architectures to integrate renewable energies.These small-scale power systems always integrate various renewable energies, local loads and energy storage (ES) devices, which are able to improve reliability and economic benefits of power supply[1,2].The concept of microgrid was originally introduced in ref.[3,4] as a promising solution to integrate distribution generation.Microgrid can be defined as a small-scale autonomous power system integrating DG units, local loads and ESS to provide reliable and efficient electricity supply at distribution level.Microgrid technology is one of important components in smart grid technology, which is an attractive solution to promote application of renewable energy resources.Microgrid can be operated flexibly either in a grid-connection mode or in an islanded mode according to power system commands.The microgrids can be applied in various occasions such as industrial microgrid, residential microgrid, offshore microgrid and ship microgrid, etc.as shown in Fig.1.

Fig.1 The diagram of the microgrids with different applications圖1 微電網(wǎng)在不同場景應(yīng)用圖

Control system is an essential component to ensure reliable, secure and economic operation of microgrids in either grid-connected or stand-alone mode.There are a number of previous reviews that provide a survey of the state-of-the-art about control strategies in microgrids such as in ref.[5-8].Ref.[5] reviews and categorizes various approaches of power sharing control for islanded operation of AC microgrids.A comprehensive review about active and reactive power sharing control strategies of AC microgrids is given in ref.[6], where various control strategies for reactive power sharing are investigated, and comparative analysis among different control strategies are presented.In ref.[7], the hierarchical control framework of microgrids is developed, where the relationship among primary control level, secondary control level and tertiary is discussed.In ref.[8], the control strategies of microgrid are reviewed and the classifications of control features are discussed in different hierarchical levels, where primary and secondary levels are associated with the operation of the microgrid, and tertiary level pertains to the coordinated operation of the microgrid and the utility grid.Fig.2 shows the diagram of conventional hierarchical control framework of microgrids.

Fig.2 The diagram of conventional hierarchical control framework of microgrids圖2 傳統(tǒng)微電網(wǎng)分層控制結(jié)構(gòu)示意圖

With the increasing penetration of power electronic converters and renewable energy sources, several new emerging issues are challenging the operation performance of microgrids, which are slightly concerned in previous work.In this work, the following aspects in microgrid control are primarily discussed.

1.1 Power management

The power management among multiple paralleled inverters for islanded microgrids is an important issue in microgrid control.

1.2 Energy efficiency

Energy efficiency is one of significant concerns in multi-converter microgrids, which is related with load profiles and power generation profiles, etc.The improvement of energy efficiency is an important issue in microgrid control.

1.3 Nonlinear phenomenon

In practical operation of microgrids, nonlinear phenomena such as dead-time nonlinearity of inverter, soft-saturation nonlinearity of filter inductor, long-term operation aging of inverter, etc.can deteriorate operation performance of microgrids.

1.4 Long-term operation reliability

The reliability of microgrids under long-term operation has become essential concerns.In practical operation, active and passive components of distributed generators are sensitive to various ambient stressors such as vibration, humidity and temperature, etc., which can decrease operation performances of microgrids under long-term operation.

The aim of this work is to investigate the potential new challenges in microgrid control, and review the relevant research efforts including the work from this research group.This paper is distinct from the previous overview papers with reviewing several new challenges and solutions, including nonlinear phenomena, operation reliability.The rest of this paper is organized as follows.In Section II, the key control issues and solutions in power management are reviewed.In Section III, the discussion about energy efficiency of microgrid is given.The control issues and solutions about nonlinear characteristics in microgrid are reviewed in Section IV.Reliability issue of microgrid is addressed in Section V.The future trends and solutions are discussed in Section VI.

2 Key control issues and solutions in power management

2.1 Power sharing control

Microgrid can be operated either in grid-connected mode or islanded mode.During the islanded mode, output powers are assigned by DG inverters themselves.Droop control methods are frequently employed to automatically perform proportional power sharing according to power rating of inverters without using critical communication facilities as shown in Fig.3, which can improve reliability and flexibility of electricity services.The classical active power-frequency (P-ω) and reactive power-voltage (Q-V) droop control strategy[9,10]is given as：

(1)

Fig.3 The diagram of droop-based power controller圖3 下垂功率控制器框圖

(2)

Where:ωmaxandωminare maximum and minimum values of allowable angular frequency.VmaxandVminare maximum and minimum values of allowable voltage amplitude.Pmax,iandQmax,iare active and reactive power ratings ofi-th inverter.

It is well-known that the accuracy of reactive power sharing can be degraded by mismatched feeder impedance, local reactive power loads or voltage sensor scaling errors[11-13].A review of active and reactive power sharing strategies in hierarchical controlled microgrids has been presented in ref.[5], where the different droop schemes are compared and summarized for active and reactive power sharing.Also, a comprehensive review about reactive power sharing control has been given in ref.[6], where various control strategies for reactive power sharing are investigated, and comparative analysis among different control strategies are provided.The prevalent reactive power sharing methods are classified into communication-based methods, droop-based methods[14]and virtual impedance-based methods[15].Communication-based methods can eliminate reactive power sharing error with immunity to mismatched line impedance, local reactive power loads as well as sensor errors[10,15].However, the implement of communication facilities will cause additional costs and may not be available for rural or remote microgrids.Reactive power sharing accuracy can also be improved by simply increasingQ-Vdroop slope under conventional droop control strategy.However, it will cause an undesirable voltage deviation.Therefore, improved droop control methods are developed to mitigate reactive power sharing error.Various improved droop control schemes have been developed to enhance the accuracy of reactive power sharing control.Also, signal injection-based droop schemes are developed in ref.[16], where a small AC harmonic voltage is injected to output voltage and the frequency of injected signal is formulated as a function of output reactive power of DG.However, the current is deteriorated and power loss of transmission lines is increased because of harmonic current.The reactive power sharing error is estimated by injecting small real power distribution to droop control loop.However, the synchronization signal from a central controller is indispensable.In addition, virtual impedance based methods are proposed to enhance reactive power sharing accuracy by compensating the feeder impedance mismatch via inserting virtual impedance in voltage control loop[17,18].However, the application of virtual impedance tends to cause a significant voltage drop, deteriorating voltage quality and limiting output power capability of DGs is introduced in ref.[19] to facilitate adaptive virtual impedance control by real-time calculation.However, feeder current sensing and physical feeder impedance information are necessary.Various methods are proposed in ref.[19,20] to find optimum virtual impedance value with consideration of power/voltage constraints, transient response, and stability.However, it is difficult to calculate an optimum virtual impedance value, especially under microgrids with complex feeder network.Communication-based virtual impedance design methods are also developed in ref.[21].However, the application of these methods can result in extra cost and low reliability.

2.2 Power control considering intermittency and fluctuation of renewable energy resources

One of the most important features of microgrids is the capability of integrating various generation sources including synchronous generators, renewable energy sources (RESs), energy storage systems (ESS), etc.RESs are commonly controlled as current-controlled sources to perform maximum power point tracking (MPPT).However, current-controlled RES fails to participate in frequency regulation of microgrids, which significantly restrict the penetration of RES[22].Therefore, the control strategies considering power fluctuation of RES have been developed[22-24].A modified droop control strategy is proposed in ref.[22] for PV generation system, where MPPT can be ensured at steady state and frequency support is performed during load step.A novel control strategy is presented in ref.[23] by combiningVg/Vdcdroop andP/Vgdroop control methods in order to perform MPPT control and voltage regulation.A quasi-master-slave control frame is proposed in ref.[24], where RESs are controlled as current controlled voltage sources, so that MPPT and frequency/voltage regulation can be performed.In addition, rating-dependent power sharing under conventional droop control strategy fails to capture optimum economic dispatch due to various operation costs of different DGs.Communication-based control methods have been developed to perform optimum economic operation.However, requirement of communication network tends to result in relative high installation cost and low reliability.

3 Control issues about efficiency of microgrids

Energy efficiency is one of significant concerns for DG inverters in microgrids.The efficiency improvement for individual inverter has been frequently developed in ref.[25-28], including efficient modulation methods[25], topologies design and passive component optimization[26,27], etc.In multi-converter microgrids, energy efficiency is associated with power sharing ratio, length of transmission lines and load profiles[28-29]so that the control strategies for efficiency improvement are complicated.Furthermore, the optimal efficiency points are time-varying as variation of load profiles as shown in Fig.4.

Fig.4 The efficiency points of multi-converter microgrids as variation of load profiles 圖4 不同負(fù)荷曲線下多逆變器微電網(wǎng)最優(yōu)效率點的變化曲線

To improve energy efficiency of multi-converter microgrid, the different control strategies have been presented in ref.[30,31].A dynamic module-dropping strategy is presented to improve operation efficiency of multi-converter microgrid under light-load condition in ref.[30], where the number of operated inverters is determined by evaluating load conditions.The rest of inverter is sequentially activated once previous modules reach their maximum output power.Further, the efficiency can be improved by optimizing the number of operated inverters, which has been addressed in ref.[31,32].In ref.[33], a game theoretic-based optimization approach is presented to investigate optimum number of paralleled converters and optimum power sharing ratio in DC microgrids.A forward-backward sweep algorithm is proposed to calculate the optimum power sharing and the optimum switching point for the system with identical inverters in ref.[34].However, the application of centralized communication facilities mitigates reliability and increases system hardware cost.

In addition, system efficiency can be improved by regulating power sharing ratio among paralleled inverters.A hierarchical control strategy is developed to improve efficiency of DC microgrids in ref.[35],where genetic algorithm is applied to capture optimum efficiency points, and an adaptive virtual resistance (VR) controller is developed to track the optimum efficiency point.A smart control strategy is proposed to improve energy efficiency of paralleled inverters under light loads in ref.[36], where particle swarm optimization algorithm is implemented to calculate current references of paralleled inverters for optimum system efficiency.An exhaustive optimization-based control strategy is developed in ref.[37] to improve operation efficiency of wind energy conversion system.However, centralized communication channel is required to implement the optimization procedure, which undoubtedly increases operation cost, and mitigates reliability of microgrid due to side effects such as data drop-out and time delay.To deal with the drawbacks, a distributed communication-based control strategy is proposed to improve operation efficiency of microgrid by adjusting power sharing ratio in ref.[38], where a dynamic consensus algorithm is applied to solve optimization problem according to information from neighbors.However, the optimum sharing ratio is obtained by real-time online computation, which tends to increase computational burdens.In previous work of our research group[39-41], an efficiency-prioritized droop control strategy is proposed in ref.[39], where time-varying optimum operation conditions of system efficiency model are derived and adopted in droop controller to improve system efficiency without using communication facilities.Further, an efficiency modelling and analysis method for multi-bus microgrids with consideration of transmission network is presented in ref.[40], where the load-related optimum conditions are derived according to the established efficiency model by Lagrange Multiplier Method.In addition, a new perspective on power control for microgrid considering operation cost and efficiency simultaneously is presented in our recent work[41], where a self-optimization control strategy with subject to optimal operation conditions is proposed to improve the overall operation performance.However, it is not easy to adaptively ensure these optimum conditions by decentralized control method in a multi-bus microgrid.

4 Control issues considering nonlinear characteristic in microgrids

Nonlinear characteristics may cause the variation of operation points, which have significant effects on operation performances of microgrids either in islanded mode or in grid-connected mode.The effects of nonlinear characteristics such as soft-saturation nonlinearity of filter inductor[42], dead-time effect of inverter[43]and nonlinear control command[44]on operation performances of microgrids have been paid intensive concerns.During islanded operation, the effects of nonlinear characteristic of filter inductor have significant influence on power control performance of microgrids[45].In our previous work[45], the effect of nonlinear inductor on control performance of multi-converter microgrids is quantitatively analyzed by establishing average model of inductor, where a robust droop control method is also developed to mitigate the effect of inductor nonlinear characteristics on power control performance.Further, the sensitivity analysis is performed to evaluate the robustness of power controller with immunity to nonlinear inductor.Fig.5 shows the experimental results about multi-converter microgrids considering nonlinear characteristics of inductor.It can be seen from Fig.5(a) that the output current of paralleled inverters is deteriorated due to soft-saturation nonlinearity of filter inductor.Once the robust droop control scheme is activated, the nonlinear characteristic can be mitigated and the accurate current sharing of paralleled inverters can be performed as shown in Fig.5(b).

In grid-connected mode, nonlinear characteristics have important effects on stability of microgrids, including soft-saturation nonlinearity of filter inductor[46], dead-time effect of inverter[47].The effect of nonlinear inductor on control performance of grid-connected PV inverter is originally investigated in ref.[48], where a self-learning algorithm is adapted to calculate nonlinear inductance.A systematicLCLfilter design method for grid-connected inverter considering nonlinear inductor is presented in ref.[49], where the effect of nonlinear inductor on resonance phenomenon is analyzed.Considering soft-saturation nonlinearity of filter inductor, a direct digital control method of grid-connected inverter is proposed in ref.[50], where a parameter space approach is developed to investigate system stability.A virtual impedance-based stability analysis method is also developed in ref.[51] to analyze stability of grid-connected inverters with nonlinear filter inductor.Fig.6 shows the research results about the oscillation phenomenon caused by soft saturation nonlinearity of inductor in grid-connected inverter.Recently, our research group presents a time-varying modelling and stability analysis method for gird-connected inverter based on Lyapunov energy function method[52], which is able to investigate the effects of time-varying operation points on long-term stability.However, the existing works mainly focus on the effect of nonlinear inductor in an individual inverter.The effects of nonlinear characteristics of inductor on microgrid with multiple grid-connected inverters are slightly concerned.

Fig.5 The experimental results about multi-converter microgrids considering nonlinear characteristics of inductors in islanded mode圖5 孤島(離網(wǎng))模式下考慮電感非線特征下多逆變器微電網(wǎng)動作試驗結(jié)果

Fig.6 The verification results about oscillation phenomenon caused by soft saturation nonlinearity of inductor in grid-connected inverter圖6 并網(wǎng)逆變器中電抗器軟飽和非線性引起的振蕩現(xiàn)象驗證

5 Control issues about reliability of microgrids

Microgirds can be utilized to provide the reliable electricity supply in safe-critical systems, such as ship microgrid, aircraft power system, and remote microgrid, etc.Reliability and its enhancement strategies are becoming essential concerns in microgrids dominated by critical loads.It is well-known that operating temperatures of power devices have key impacts on long-term reliability of power electronic system, where over-temperature and temperature fluctuations caused by power semiconductors losses are the critical factors of power modules failures[53].Fig.7 shows the typical junction temperature profile of power device of DG inverter.Once power devices are continuously operated at a junction temperature above the maximum temperature (Tj,max), the failure of power devices may happen[53].

Fig.7 The typical junction temperature profile of power device of inverter圖7 逆變器功率器件典型結(jié)論分布圖

In microgrids, thermal operating points of paralleled inverters may be different due to differences of power devices and aging effects under long-term operation[54].The unequal thermal stress distribution among paralleled inverters may be caused by the difference of inverter parameters, which thus mitigates reliability of microgrids under long-term operation.Thermal management is becoming an important aspect of power electronic systems with the increasing demands of high reliability[55].Active thermal control strategies by reducing or redistributing load power have been introduced to increase the average lifetime of inverters.To improve long-term reliability of power converters, an active thermal control-based current sharing scheme for paralleled inverters is presented in ref.[54], where the effects of parameters perturbation on temperature is investigated and load current is redistributed among paralleled inverters according to their temperature difference.A dynamic electro-thermal model is established to estimate transient junction temperature of semiconductor devices in ref.[55], and then power sharing among paralleled inverters is facilitated according to estimated junction temperature.The proposed method is able to improve overall system efficiency and reliability.However, the communication channels are required to support the proposed control method.

In previous work[56-58]of our research group, the thermal management framework toward reliability enhancement of microgrids is developed.In ref.[56,57], a lifetime-oriented droop control strategy is developed to implement the active thermal sharing among multiple paralleled inverters.The proposed method is able to redistribute thermal dissipation and mitigate the thermal stresses, which also preserves the inherent advantage of conventional droop control.In our recent work[58], an enhanced hierarchical control framework of microgrids is developed to perform thermal management and improve operation efficiency, which is an enhanced complementary for conventional hierarchical control of microgrids.

6 Discussions and future trends

This paper presents an overview for emerging control issues in microgrids, and displays the main contributions of the research group in these aspects, including power management, efficiency improvement, nonlinear characteristics and reliability enhancement.The future trends and solutions are discussed as following.

1) Power management.The power management of paralleled inverters in microgrids is the primary concern in microgrid control.The existing works mainly focus on the accuracy of power sharing control.The power management strategies considering intermittency and fluctuation of renewable energy sources should be further addressed.

2) Energy efficiency.Energy efficiency is becoming an important concern in multi-converter microgrids.Different from individual inverter, the efficiency of multi-converter microgrid is related with power sharing profile, load profile, transmission line, etc.Therefore, the optimal efficiency points are always time-varying.The dynamic identification of optimal efficiency points under different operation conditions is critical for efficiency improvement.Also, the adaptive efficiency improvement strategies without using communication channels are of importance.

3) Nonlinear phenomenon.In practical operation, nonlinear phenomena may deteriorate operation performance of microgrids.During islanded operation, the nonlinear characteristics of inverters may deteriorate the power sharing performance.In grid-connected mode, the nonlinear characteristics of inverter such as soft-saturation nonlinearity and dead-time nonlinearity of inverter can cause the instability phenomenon.With the increase of operation time, the system parameters degradation may cause aging nonlinear characteristics, which further deteriorates the operation performance of microgrids.The effects of nonlinear characteristics on operation performance of microgrid should be further investigated, and the mitigation control strategies for nonlinear phenomena should be further developed.

4) Long-term operation reliability.With the increase of operation time, the system parameters can be perturbed.For instance, the active and passive components of distributed generators are sensitive to various ambient stressors such as vibration, humidity and temperature, etc., which further deteriorates the operation reliability of inverters within the lifespan.The reliability mechanism and reliability enhancement strategies of microgrids under long-term operation should be further addressed.