The Improved Coordinated Control Method of Islanding and Grid-Connected Operations for DC Microgrids

The DC microgrid plays an essential role in realizing smart power supply systems for the DC loads. It combines various distributed generations and transmits the power to the loads or the utility grid through the DC bus. It attaches great importance to energy management because of the load and photovoltaic power fluctuations. This paper proposes an improved coordinated control based on the characteristics of various distributed generations and utility grids. The proposed method can realize the stabilization of bus voltage when the system switches between multiple operating modes in grid-connected and island mode due to the change of energy. Compared with the present ways, the proposed method provides a further optimization on the design of charging for energy storage units in the grid-connected mode and the current sharing with the backup storage units in the islanded mode which is conducted by SoC balancing droop control. Finally, the DC microgrid simulation platform is built to verify the correctness and feasibility of the proposed method.


Introduction
Due to the continuous innovation of power electronic technology, solar and wind energy, as the representatives of RESs, are gradually replacing fossil energy [1].Microgrids (MGs) can effectively integrate various RESs, energy storage systems (ESSs), loads and power electronic devices.Moreover, due to the increasing number of DC sources and the wide application of DC loads, intelligent control and DC technology in DC microgrids (DCMGs) have become more and more popular worldwide [2].The typical structure of DCMG is shown in Figure 1.IOP Publishing doi:10.1088/1742-6596/2625/1/012056 2 This regional power supply system needs to switch between island mode and grid-connected mode, so power management is essential for the DCMGs [3], coordinating all the units reasonably in each operation mode while maintaining the bus voltage in a stable area is of great importance to ensure the stability of the DCMGs.
Due to the differences in application backgrounds for the systems, several coordinated control methods for DCMGs have been proposed in recent years.In 2011, Sun et al. [4] proposed coordination control based on DBS, but it generally applies to the DCMGs with a simple structure.A power management strategy is proposed in [5], which optimizes the transient performance of the DCMGs, but the strategy just investigates the grid-connected mode while ignoring the islanded mode.Kumar et al. [6] proposed an algorithm for coordinated control in both grid-connected and off-grid operations.By analysing the perspective of the control structure, a decentralized control method aiming at realizing the generationstorage coordination was proposed in [7] by Xia et al.For the DCMGs with multiple energy storage units (ESUs) access, Zhi et al. [8] gave an SoC-based control method to realize coordination control, which can both enhances the transient stability and reduces SoC differences between ESUs in all operating modes.All of the above methods are focusing on the research of certain operational modes, but lack the plan of coordination between gird-connected modes and islanded modes.Meanwhile, the consideration of ESSs ought to be considered more fully to maintain the efficiency and operation health of energy storage devices [9].
In this paper, an improved coordination control method is proposed.Firstly, the control methods of each component in DCMGs are described.Then, based on the control methods, this paper proposes improved ideas to make a coordination of the power flow between the RESs, ESSs, loads and utility grids.In addition, the proposed method provides a further optimization on the design of charging for ESUs in the grid-connected mode and the current sharing with the backup storage units in the islanded mode which is conducted by SoC balancing droop control compared to existing methods.Finally, the simulation results in Matlab/Simulink are given to verify the feasibility of the proposed method.

The control methods for PV component
As one of the most representative RESs, PV can realize power delivery to other components with boost converters.To take full advantage of solar energy, PV components operate with maximum power point tracking (MPPT) in most cases.However, if the other voltage regulation converters are in standby mode while the required power is less than that from the PV, so in order to sustain a stable bus voltage, the PV component has to operate at constant voltage control (CVC).
Figure 2 shows the structure and control diagram of the PV component.In this paper, the MPPT for PV is carried out by the perturbation and observation method (P & O), while the CVC is performed with the voltage and current double closed-loop structure.

The control methods for the ESS component
ESSs can compensate or buffer power fluctuations of the RESs promptly with the Bi-DC/DC converters, so they can integrate discontinuous and random sources into a continuous, stable energy supply.
Figure 3 shows the structure and control block diagram of the ESSs.The Bi-Buck/Boost converter is adopted to realize the energy flow between the energy storage devices and the DC bus, and considering the long-term operation of the system, the battery pack is chosen as the major source in the ESS.In the islanded mode, the ESS operates at constant voltage mode to sustain a stable DC bus voltage with average current control, and the PWM generator is adopted the complementary PWM mode which realizes the autonomous bidirectional flow without the extra judgment of charging or discharging.In addition, the ESS can operate at constant current mode to charge the battery pack when connecting to the grid.To reduce the pressure of heavy power load on the ESS in the island mode, a method of adding reserve energy storage units to the system is proposed.The objective can be realized by adaptive droop control and adding it into the CV mode control as the reference value Vo * for the voltage loop.
The adaptive droop control proposed in this paper can be obtained by improving the SoC balancing method in [10], and Figure 4 shows the control block diagram.compensation circuit where v dc is the rated reference voltage; SoC i is the state of charge of the battery; Ks is the adjustment coefficient; R d is the virtual impedance; i oi represents the output current of each converter.The SoC is generally calculated as _0 1 ( ) ( ) where SoC i_0 represents the initial value at the beginning of estimation; C e represents the rated capacity; i bati represents output current of the battery pack.
Assuming that there are n ESUs (n = 1, 2, 3, …) in the ESS operating at CV discharging mode after adding the backup energy storages and neglecting the voltage drop of each converter to the DC bus, it yields where vbus is the DC bus voltage.Moreover, combined with Equation ( 1), the output current relationship of n converters in the ESS can be expressed as where the δV in Equation ( 4) is defined as Based on the relationship in Equation ( 4), the output current value of each converter in ESS is determined by the SoC in the batteries.The ESU with a higher value of SoC will provide more power (or output current) than it with a lower value.Also, the ESU with a higher SoC value will show a higher rate of SoC change than the ESUs with a lower value, and the values of SoC in all batteries would converge after a period of operation to achieve the objective of SoC balancing in the case of backup energy storages adding.

The control methods for grid-connected component
Different from the island mode, the DC bus and the utility grid have been connected to realize the bidirectional power flow when the grid connected, and the single-phase voltage type Bi-AC/DC converter is adopted to connect the two in this paper.Moreover, the Bi-AC/DC converter can operate in a rectifier or inverter state according to the value of the power difference between the AC and DC side.
In this paper, the DQ current decoupling control is used as the control method of the Bi-AC/DC converter, and it can convert AC quantity into direct current through a synchronous rotating coordinate system.The control block diagram of the Bi-AC/DC converter is shown in Figure 5.

Priority setting in proposed coordinated control
The highest priority is to maintain bus voltage stability within the rated range, thus, in any case, the converters for stabilizing the bus voltage are always available in the system.The second priority is the rationality of each equipment parameter in the system.The SoC of each battery in the ESS must be maintained in the range of 20% to 90% at all times, and once this range is exceeded, the ESU will be shut down.The maximum charging power of the battery in the ESS should be no less than the maximum rating output power of the PV, and hence hardware configuration of Bi- DC/DC converter should be matched to this setting.Besides, the transmission power of the Bi-AC/DC converter should meet the maximum capacity requirement of the load on the DC side and the maximum power requirement of the PV component.

The description of the proposed coordinated control
The proposed coordinated control divides into the 6 operation modes beside for the two different cases of island and grid connection based on the control methods of each component and the consideration of the control priorities.
When it is in islanded operation situation, and the power management is just for the RESs, ESSs and loads.
Mode Ⅰ: The PV component operates in MPPT mode, while the energy storage unit in the ESS is controlled in CV mode.(In which, if the PV cannot provide enough power for the load, the ESS discharges to compensate for the remaining power.On the contrary, if the output power of the PV is too large to be consumed by the load, the battery will absorb the excess energy.And the battery can charge or discharge in the condition of 20% ≤ SoC ≤ 90% without additional switching judgment.) Mode Ⅱ: The PV component is in CVC mode, while the energy storage is in standby mode.(In which, the battery has been charged for a long time in mode Ⅰ and its SoC has reached the critical point of 90%, so it is shut down on this occasion.And the battery would be put back into operation in discharging state only when the system requires it to provide power to the loads.) Mode Ⅲ: The PV component is in MPPT mode, while the backup energy storages are added, operating in CV mode at the same time.(In which, to avoid a single battery with excessive output current under high load demand where its SoC drops rapidly to the critical condition of 20%, the backup energy storages can alleviate this problem.) In this operation, the utility grid is connected through Bi-AC/DC converter.As a result, power management involves RESs, ESSs, loads and the utility grid.Besides, the PV component operates in MPPT mode in the following three modes.
Mode Ⅳ: The battery operates at constant current charging mode and the Bi-AC/DC converter is operating at rectification providing power to the DC bus.(In which, the battery has not reached the maximum critical value.So, to ensure that the battery has enough energy to be used again, the battery should be charged by the PV component while the utility grid provides power to the loads.) Mode Ⅴ: The battery is in standby mode and the Bi-AC/DC converter is operating in rectification mode delivering power to the DC bus.(In which, the SoC of the battery has already reached its maximum limit and does not need to be recharged, and the power generated from PV and the utility grid meets the demand of the loads.) Mode Ⅵ: The battery is in standby mode and the Bi-AC/DC converter is operating in the inverted state to deliver power to the utility grid.(In which, the battery has already reached its SoC maximum limit and does not need to be recharged, if the output power of the PV is too large to be consumed by the load, and the remaining power is directly integrated into the utility grid.) Based on the control priority, the judgment conditions are designed for mode switching among the above 6 modes to ensure reasonable operation of the system in various conditions.
The detailed mode switching and operation flow are shown in Figure 6.

Simulation results
The simulations for verifying the feasibility of the proposed control method are implemented by Matlab/Simulink.The DCMG simulation platform consists of a PV unit, a grid-connected unit, an ESS in which only one battery is configured, and some backup energy storage units.

Coordination of the islanded and grid-connected
The simulation results of the DCMG operation in islanding and grid-connected modes are represented in Figure 7.According to Figure 7 (a), the system operates in mode Ⅰ first and switches to grid-connected mode at the moment t = 1 s to connect to the utility grid.Starting from the moment t = 1 s, the system operates in mode Ⅳ, and the battery has not reached its maximum capacity limit.Until the moment 1.9 s, the battery reaches the SoC maximum limit and the system operates in mode Ⅴ or mode Ⅵ based on the power relationship among each component.After that, the system is disconnected from the utility grid at t = 4 s and switches to the mode Ⅱ of the islanding.Besides, Figure 7

Conclusion
This paper proposed an improved coordinated control method for the DCMGs, which can be applied to both islanding and grid-connected operation.According to the control methods of each component in the system, the proposed control method provides six control modes and novel control logic for the coordinated operation.In addition, compared to existing methods, the proposed method provides a further optimization on the design of charging for ESUs in the grid-connected mode and the current sharing with the backup storage units in the islanded mode.Finally, the feasibility of the proposed method is verified by the DCMGs simulation platform.

Figure 2 .
Figure 2. The structure and control diagram of the PV

Figure 3 .
Figure 3.The structure and control block diagram of the ESSs

Figure 4 .
Figure 4.The control block diagram of the adaptive droop control And the reference value Vo * of each converter in ESS after adding backup energy storages can express as

Figure 5 .
Figure 5.The control block diagram of the Bi-AC/DC converter 3. Proposed improved coordinated control method

Figure 6 .
Figure 6.The detailed mode switching and operation flow of the DCMGs

Figure 7 .Figure 8 .
Figure 7.The coordination of the islanded and grid-connected