TY - JOUR
AU - Zhang, Jian
AB - Abstract The hybrid energy storage system can compensate the bus power fluctuation caused by the output power and load variation of the generator set in the Direct Current (DC) microgrid. In the current control strategy, the voltage droop method is used to control the non-high-frequency components of the battery to suppress the power fluctuation of the bus and the integral droop method is used to control the components of the supercapacitor to suppress the high-frequency power fluctuation. The effect of this strategy is similar to that of a filter, which automatically divides the ripple power into high frequency and non-high-frequency ripple components. However, when the voltage drop method is adopted, the difference of battery State of charge (SOC) in the parallel hybrid energy storage unit is not considered and the non-high-frequency fluctuation components subjected to by the battery are not redistributed, which is not beneficial to the avoidance of battery overcharge and over discharge. It is considered that the voltage drop coefficient will be affected by SOC, the voltage drop coefficient is divided by the n power of SOC to get a new voltage drop coefficient, and then the improved voltage drop method related to the voltage drop coefficient SOC is obtained, and then a new coordinated control strategy is obtained by combining the integral drop method. Simulation and experimental results show that the proposed control strategy can reallocate the non-high-frequency power according to the SOC of the battery and suppress the bus power fluctuation at the same time, so as to realize the SOC balance of different energy storage cells, and the bus voltage fluctuation is small during the whole process. 1. INTRODUCTION In order to realize the energy structure transformation and respond to the sustainable development strategy of energy, distributed generation represented by wind energy and solar energy has quickly become the focus of experts and scholars [1]. Distributed generation is random and intermittent, and its stability is too poor to directly access the large grid [2, 3]. In order to solve the problem of integration of distributed power and large grids, the concept of microgrid came into being. The microgrid is divided into Alternating Current (AC) microgrid and DC microgrid [4]. Although the former currently occupies a major position, the DC microgrid structure is simpler and does not need to consider factors such as frequency, reactive power and phase, which is more advantageous than AC microgrid. Therefore, DC microgrids have received people’s attention in recent years. The bus voltage directly reflects whether the power between the DC microgrid power supply and the load is balanced and whether the system can operate stably [5]. Because of the randomness of the output power variation of the distributed power generation unit and the uncertainty of the load change, the hybrid energy storage system (HESS) plays a crucial role in the regulation of the DC microgrid bus voltage. The battery and the supercapacitor have a strong complementarity in their inherent characteristics [6, 7] and use them to form an HESS, making full use of the advantages of both. Therefore, the study of the power distribution strategy of HESS is the key to ensure the stable operation of DC micro-electricity. Figure 1 Open in new tabDownload slide Structure diagram of DC microgrid with hybrid energy storage Figure 1 Open in new tabDownload slide Structure diagram of DC microgrid with hybrid energy storage Yang et al. [8] improve the accuracy of the current distribution but do not consider the SOC and cannot perform power distribution based on the capacity of the energy storage unit. Zhang et al. [9] divide the operating mode according to the bus voltage information and use droop control for the photovoltaic array or the battery module of the electric vehicle to achieve power distribution in the reference. Li et al. [10] improved the SOC exponent of the voltage droop method during charge and discharge of the energy storage unit and improved the resolution of the SOC. Through this droop control method, the system bus voltage was stabilized and the system quickly converged to the SOC equilibrium state. Lin et al. [11] propose a novel integral sag method. The integral sag method is used to control the supercapacitor to quickly compensate the high-frequency fluctuation components of the power fluctuation. The traditional voltage sags method is used to control the battery to stabilize the non-high-frequency fluctuation components of the power fluctuation. The coordinated control of the two drooping methods achieves the power distribution of the fluctuating power between the energy storage units. In order to solve the overcharge and over-discharge of the battery in the energy storage unit, it is necessary to redistribute the non-high-frequency fluctuation component that the battery bears. Based on the traditional voltage droop method, this paper improves the voltage droop coefficient and then combines the improved voltage droop method with the integral droop method to obtain a coordinated control strategy. Simulation and experimental results show that the proposed control strategy can suppress the bus voltage fluctuations and can achieve the battery SOC equilibrium in the energy storage unit. 2. DC MICROGRID HESS Figure 1 shows a schematic diagram of a DC microgrid with a typical HESS. The grid-connected operation mode is not discussed in this paper. Here, only the hybrid storage power allocation strategy when the microgrid is operating in islanding mode is studied. There are many kinds of connection modes between the storage battery and the supercapacitor in the HESS, and often used are direct parallel connection, single DC/DC parallel connection and dual DC/DC parallel connection [12, 13]. The disadvantage of direct parallel connection is obviously that the voltage output from the supercapacitor and the battery is always the same, resulting in inflexible control of it. There are two situations in which a single DC/DC is connected in parallel. The first is that the battery is first connected to the DC/DC converter and then in parallel with the supercapacitor. This method can guarantee the service life of the battery and reduce the requirements on the battery [14]. However, if a single supercapacitor is directly connected to the busbar, its voltage is difficult to reach the DC bus voltage level. When multiple supercapacitors are connected in parallel, there is no doubt that the cost will be increased. The second is that the supercapacitor is connected to the DC/DC converter and then connected in parallel with the battery. This method greatly reduces the voltage requirement for the supercapacitor, but the battery capacity cannot be fully utilized in this case [15]. The dual DC/DC parallel connection not only can adjust the output power of the energy storage element but also can control the charging and discharging processes of the supercapacitor and the battery separately [16]. In addition, the effect of bus voltage on the voltage of the two types of energy storage elements will also be reduced to a very small degree, more flexible control. In summary, this article decided to use dual DC/DC connection. In order to more easily and directly study the power interaction between the DC bus and the distributed power generation unit, the hybrid energy storage device and the load, and taking into account that the DC microgrid is in the island operation mode. The structure of the DC microgrid HESS shown in Figure 1 can be simplified to the structural model shown in Figure 2. Regardless of the specific working conditions within each module, only the power interaction between these modules and the DC bus is considered, and each part can be equivalent to a module whose power can be changed, thereby simplifying our research and analysis process. In Figure 2, PS is the output power of the distributed power generation unit, PL is the power consumed by the AC-DC load and PES indicates the power released or absorbed by the HESS. Figure 2 Open in new tabDownload slide DC micro grid HESS simplified structure diagram Figure 2 Open in new tabDownload slide DC micro grid HESS simplified structure diagram The power difference between the distributed power supply and the AC/DC load is provided by the HESS, so as to stabilize the DC bus voltage and realize the power balance of the DC micro-grid. Then, the DC bus power balance needs to meet the conditions show as Equation 1. $$\begin{equation} {P}_S-{P}_L={P}_{ES}, \end{equation}$$(1) where PS is the output power of the distributed power generation unit, PL is the power consumed by the AC-DC load and PES = the power released or absorbed by the HESS. 3. THE PRINCIPLE OF DROOP METHOD Centralized control, master–slave control and other power allocation control strategies based on high-speed communications enable rapid power allocation and ensure that all operating parameters of the microgrid are at rated values. However, such control strategies rely too much on communications facilities. When the communications link experiences serious failures and delays, the control system is prone to failure. Droop control method is one of the most commonly used methods in microgrid power control strategies that do not require interconnected communications. Compared with other control strategies that require communications, there is no problem of communication device failure and system crash. And the droop control method has a simple structure and can also reduce the current between the DC/DC converters. The traditional voltage droop method is shown in Equation 2. $$\begin{equation} {u}_{ref}={u}_n-{i}_o\cdot R, \end{equation}$$(2) where uref is the reference output voltage and un is the rated voltage of the converter. When a sudden power change occurs on the DC bus, the supercapacitor can respond quickly, so it can suppress the sharp rise or fall of the DC bus voltage. Figure 3a shows the charging circuit diagram for the series connection of capacitors, resistors and a DC voltage source. Figure 4b shows the discharge circuit diagram for a capacitor and resistor connected in series. Figure 3a and b is equivalent to the RC circuit full-response and zero-input response circuit in circuit theory knowledge, where US and uC represent the DC supply voltage and capacitor voltage, respectively. Figure 3 Open in new tabDownload slide Capacitor charging/discharging circuit diagram Figure 3 Open in new tabDownload slide Capacitor charging/discharging circuit diagram Figure 4 Open in new tabDownload slide Converter control block diagram based on droop method Figure 4 Open in new tabDownload slide Converter control block diagram based on droop method In the charging circuit shown in Figure 3a, the switch S is closed and uR + uC = US can be obtained according to Kirchhoff’s voltage law (KVL). Then, the initial condition and the relation of each element are brought into KVL and Equation 3 can be obtained: $$\begin{equation} RC\frac{d{u}_C}{dt}+{u}_C={U}_S \end{equation}$$(3) By solving the differential Equation 3, the capacitor voltage and current expressions as shown in Equations 4 and 5 can be obtained, where Uo is the initial voltage of the capacitor and Uo is less than the DC supply voltage US, so that the capacitor can be charged after the switch S is closed and its voltage will increase to US at the end of the charge. The capacitance voltage of the charging circuit is shown in Equation 4. The capacitance current of the charging circuit is shown in Equation 5. $$\begin{equation} {u}_C={U}_S+\left({U}_o-{U}_S\right){e}^{-t/ RC} \end{equation}$$(4) $$\begin{align} {i}_C=\frac{U_o-{U}_S}{R}{e}^{-t/ RC}, \end{align}$$(5) where uC is the capacitor voltage, US is the DC supply voltage, Uo is the initial voltage of the capacitor and iC is the charging current. For the discharge circuit shown in Figure 3b, the solution of the capacitor voltage and current expressions is shown in Equations 6 and 7: $$\begin{equation} {u}_C={U}_o{e}^{-t/ RC} \end{equation}$$(6) $$\begin{equation} {i}_C=\frac{U_o}{R}{e}^{-t/ RC}. \end{equation}$$(7) The combination of the voltage and current expressions of the capacitor charging circuit and the discharging circuit is shown in Equation 8. $$\begin{equation} {u}_C={U}_o-m\int{i}_C dt \end{equation}$$(8) 4. HYBRID ENERGY STORAGE POWER DISTRIBUTION CONTROL STRATEGY 4.1. The derivation of equivalent formula of voltage sag based on SOC Since the SOC of the battery is different in the parallel hybrid energy storage unit, the output capability will be different when responding to non-high-frequency fluctuation components. In order to redistribute non-high-frequency fluctuation components, it is necessary to determine the output size according to the battery SOC. A battery with a large SOC absorbs less power and releases more power. A battery with a smaller SOC absorbs more power and releases less power. This helps avoid overcharging and over-discharging of the battery. According to the battery calculation model, the calculation formula for the state of charge of the battery can be obtained as shown in Equation 9. $$\begin{equation} SOC= SO{C}_0-\frac{1}{C_N}{\int}_0^t\eta Id\tau, \end{equation}$$(9) where SOC0 is the initial state of charge of the battery and CN is the capacity of the battery cell. Ignoring the power loss of the converter and expecting the output voltage of each battery group to be equal, the input and output power of the converter can be approximately equal to the product of the input voltage and the output current of the converter. According to the above, the SOC calculation formula of Equation 10 can be obtained. $$\begin{equation} SOC= SO{C}_0-\frac{1}{C_N{U}_t}{\int}_0^t\eta{P}_o d\tau \end{equation}$$(10) The product of the equivalent output resistance R of the converter and the output current io in Equation 2 is replaced with the output power Po to obtain the voltage droop method formula as in Equation 11. Then, considering the relationship between droop coefficient and SOC, it can be changed to the voltage droop formula as shown in Equation 12. n takes a positive integer and the SOC equalization speed changes when the values are different. $$\begin{equation} {u}_{ref}={u}_n-{a}_0{P}_o \end{equation}$$(11) $$\begin{equation} {u}_{ref}^{\hbox{'}}={u}_n-{a}_0/ SO{C}^n\cdot{P}_o={u}_n-a{P}_o, \end{equation}$$(12) where a0 is the initial droop coefficient and a is the new droop coefficient. 4.2. Hybrid energy storage coordination control strategy At present, there are many kinds of mixed energy storage power distribution methods and power distribution methods based on power fluctuation characteristics are most suitable for energy storage element characteristics of HESSs. Among them, the power distribution method based on high–low-pass filtering is the most studied by experts and scholars. From the derivation process of the integral sagging method, it can be seen that this method is suitable for controlling supercapacitors and the voltage sagging method is suitable for the control method of the battery. Because the voltage sagging method and the integral sagging method require coordinated control to suppress bus power fluctuations, Equations 8 and 12 are rewritten as shown in Equations 13 and 14. $$\begin{equation} {u}_1={u}_{n1}-a{P}_D \end{equation}$$(13) $$\begin{equation} {u}_2={u}_{n2}-b\int{P}_G \end{equation}$$(14) The power absorbed or released by the HESS is equal to the high-frequency ripple power plus the non-high-frequency ripple power. Laplacian transformations are performed on Equations 13 and 14 and transformed from time domain to complex frequency domain for analysis. You can get expressions like Equations 15 and 16. $$\begin{equation} {P}_G=\frac{s}{s+b/a}{P}_{ES} \end{equation}$$(15) $$\begin{equation} {P}_D={P}_{ES}-{P}_G=\frac{b/a}{s+b/a}{P}_{ES} \end{equation}$$(16) Equations 15 and 16 show that the coordinated control strategy of the voltage sagging method and the integral sagging method can automatically split the ripple power into high-frequency ripple power and non-high-frequency ripple power without using a high-pass/low-pass filter. Then, Laplacian inverse transformation is performed on the above equation to obtain the respective output of the supercapacitor, as shown in Equation 17, and the battery when the DC bus generates power fluctuations, as shown in Equation 18. $$\begin{equation} {P}_G={P}_{SC}={P}_{ES}{e}^{\frac{b}{a}t} \end{equation}$$(17) $$\begin{equation} {P}_D={P}_{Bat}={P}_{ES}-{P}_G={P}_{ES}-{P}_{ES}{e}^{\frac{b}{a}t} \end{equation}$$(18) The high-frequency fluctuation component is suppressed by using a supercapacitor, and the non-high-frequency fluctuation component is sent to the battery for suppression. The DC/DC converter control block diagram using the voltage sagging method and the integral sagging method is shown in Figure 4. The non-high-frequency ripple power (high-frequency ripple power) is brought into Equation 13a and b to obtain the reference voltage. The reference voltage is compared with the actual output voltage of the battery (supercapacitor) and then through the voltage loop to get the reference current. Then the reference current is compared with the output current of the battery (supercapacitor) and the duty cycle of the converter is calculated after the current loop, so that the PWM wave of the switching tube of the DC/DC converter is obtained. Supercapacitors are used to suppress high-frequency fluctuation components, and non-high-frequency fluctuation components are sent to the battery for inhibition. The DC/DC converter control block diagram using voltage sag method and integral sag method is shown in Figure 4. The non-hf ripple power (HF ripple power) is substituted into Equation 13a and b to obtain the reference voltage, which is compared with the actual output voltage of the battery (supercapacitor), and then the reference current is obtained through the voltage loop. The reference current is then compared with the output current of the battery (supercapacitor), and the duty cycle of the converter is calculated after the current loop, thus obtaining the PWM wave of the DC/DC converter switch tube. 5. SIMULATION AND EXPERIENCE 5.1. Simulation analysis This paper validates the correctness of the proposed power distribution control strategy by building a simulation model in MATLAB/Simulink. Two sets of distributed power generation units are used for simulation. The maximum output power of the first and second groups is 2000 and 5000 W. Two sets of batteries with a rated capacity of 4A•h used, the SOCs of the first and second groups were 60% and 40%, respectively. Two sets of supercapacitors with a rated capacity of 25F and two sets of loads of 4000 W. After the simulation began for a period of time, the system was operating in a stable state. The maximum output power of the distributed power supply was changed from 2000 to 5000 W, and the load was 4000 W. Figure 5 is a simulation result of a DC microgrid when the power of a distributed power generation unit changes. Figure 5 Open in new tabDownload slide Simulation result diagram of DC microgrid Figure 5 Open in new tabDownload slide Simulation result diagram of DC microgrid From Figure 5a and b, it can be seen that the battery does not have time to act when the simulation is started and the output power of the supercapacitor mutates from 0 to 2000 W to supplement the power shortage. Then, the output power of the battery slowly increases to 2000 W, and the output power of the supercapacitor gradually decreases from 2000 to 0 W. At ~0.6 s, the output power of the energy storage element is fixed and the DC microgrid is operating in stable state. At 1 s, the output power of the distributed power generation unit was changed from 2000 to 5000 W and the battery was still in a state of releasing power because of slow motion. Therefore, the supercapacitor needs to absorb 3000 W of power and then decay to 0 W in exponential form and the battery is slowly converted from the original release power to absorbed power and finally tending to be stable. As shown in Figure 5c, the bus voltage increases exponentially to 386 V and the voltage remains unchanged when the power balance is reached. Then, at 1 s, the output power of the distributed generator unit changes and the voltage of the bus is ~384.7 V and then remains unchanged. The DC bus voltage fluctuates little during the entire process and responds quickly. 5.2. Experimental verification A microgrid model was set up in the laboratory for verification. The switching frequency of the DC/DC converter is 10 kHz, the load is a resistive load, the other parameters are the same as those of the simulation model. The experimental results are shown in Figures 6 and 7. Figure 6 Open in new tabDownload slide Experimental results of DC microgrid Figure 6 Open in new tabDownload slide Experimental results of DC microgrid Figure 7 Open in new tabDownload slide Battery power distribution when distributed generation power fluctuation Figure 7 Open in new tabDownload slide Battery power distribution when distributed generation power fluctuation Figure 6a and b is the output power changes of the battery and the supercapacitor. It can be seen from the figure that the change of the battery is relatively slow, while the supercapacitor responds quickly to power fluctuations. Figure 6c shows that the bus voltage can quickly return to stable after power fluctuations occur and the voltage fluctuation is small. Figure 7a shows the battery power distribution obtained by using the original hybrid energy storage control strategy. It can be clearly seen that different battery groups always bear the same non-high-frequency fluctuation components, which is detrimental to the battery SOC balance. Figure 7b shows the battery power allocation when using the hybrid energy storage control strategy proposed in this paper. It can be seen that the first group of batteries with large SOC discharges more power when discharged and absorbs less power when charging. The second group has a small SOC. The discharge power is less when discharged, and more power is absorbed when charging. 6. CONCLUSION Usually, the battery stabilizes the fluctuating power at the non-high-frequency part and the supercapacitor supplants the fluctuating power at the high-frequency part. However, this strategy does not take into account the overcharge and over-discharge of the battery. In this paper, based on the original voltage droop method formula, the voltage droop coefficient is divided by the nth power of SOC to obtain a new voltage droop coefficient. This can redistribute non-high-frequency ripple power based on battery SOC size, the SOC with large battery has less output when charging and has more output when discharging. A battery with a small SOC has less output when it is charged and has more output when discharged. By building a simulation model in MATLAB/Simulink and setting up a microgrid model in the laboratory, the results show that the proposed control strategy can suppress the bus power fluctuations while also achieving the battery SOC balance in the energy storage unit, avoiding overcharge and over-discharge of the battery. References [1] Zhi N , Zhang H, Xiao Y. Research on improved droop control strategy to improve dynamic characteristics of DC microgrid . J Electr Eng Technol 2016 ; 31 : 31 – 9 . Google Scholar OpenURL Placeholder Text WorldCat [2] Yang J , Jin X, Wu X et al. Non-interconnection communication network power allocation strategy of hybrid energy storage system in DC microgrid . J China Electrotech Soc 2017 ; 32 : 135 – 44 . Google Scholar OpenURL Placeholder Text WorldCat [3] Lu X , Guerrero JM, Sun K et al. An improved droop control method for DC microgrids based on low bandwidth communication with DC bus voltage restoration and enhanced current sharing accuracy . IEEE Trans Power Electron 2013 ; 29 : 1800 – 12 . Google Scholar Crossref Search ADS WorldCat [4] Oliveira TR , Donoso-Garcia PF. Distributed secondary level control for energy storage management in DC microgrids . IEEE Trans Smart Grid 2016 ; 99 : 1 – 11 . Google Scholar OpenURL Placeholder Text WorldCat [5] Li M , Fan J, Qiao L. Adaptive droop control of a multibus DC microgrid based on consensus algorithm . Math Probl Eng 2021 ; 13 : 22 – 30 . Google Scholar OpenURL Placeholder Text WorldCat [6] Zhao X , Peng K, Zhang X et al. Droop control method for flexible DC distribution system based on feedforward compensation . Autom Electr Power Syst 2018 ; 61 : 635 – 47 . Google Scholar OpenURL Placeholder Text WorldCat [7] Mi Y , Wu Y, Zhu Y et al. Coordinated control of independent DC microgrids with automatic distribution of load power . Power Syst Technol 2017 ; 41 : 440 – 7 . Google Scholar OpenURL Placeholder Text WorldCat [8] Yang J , Jin X, Wu X et al. An improved current load distribution control strategy for DC microgrid . Proc CSEE 2016 ; 36 : 59 – 67 . Google Scholar OpenURL Placeholder Text WorldCat [9] Zhang G , Wang X, Wang S. Research on DCS microgrid energy coordination control strategy based on DBS . High Volt Appar 2018 ; 19 : 77 – 86 . Google Scholar OpenURL Placeholder Text WorldCat [10] Li P , Zhang C, Yuan R et al. Load current distribution method of distributed energy storage system based on improved SOC droop control . Proc CSEE 2017 ; 37 : 3746 – 54 . Google Scholar OpenURL Placeholder Text WorldCat [11] Lin P , Wang P, Xiao J et al. An integral droop for transient power allocation and output impedance shaping of hybrid energy storage system in DC microgrid . IEEE Trans Power Electron 2017 ; 18 : 129 – 38 . Google Scholar OpenURL Placeholder Text WorldCat [12] Salman S , Xin A. Droop control based approach for frequency and voltage in hybrid AC/DC microgrid . J Electr Eng Technol 2020 ; 29 : 177 – 87 . Google Scholar OpenURL Placeholder Text WorldCat [13] Fu Y , Zhang Z, Mi Y et al. Droop control for DC multi-microgrids based on local adaptive fuzzy approach and global power allocation correction . IEEE Trans Smart Grid 2018 ; 13 : 146 – 55 . Google Scholar OpenURL Placeholder Text WorldCat [14] Zhu Y , Fan Q, Xiong L et al. Coordination control strategy for battery—ultracapacitor hybrid energy storage system in microgrids with unbalanced and nonlinear loads . IEEE Access 2019 ; 20 : 1772 – 81 . Google Scholar OpenURL Placeholder Text WorldCat [15] Shi M , Xia C, Zhou J et al. Advanced secondary voltage recovery control for multiple HESSs in a droop-controlled DC microgrid . IEEE Trans Smart Grid 2019 ; 10 : 3828 – 39 . Google Scholar Crossref Search ADS WorldCat [16] Wen B , Qin W, Han X et al. DC micro-grid hybrid energy storage system control strategy based on voltage sag method . Power System Technology 2015 ; 39 : 892 – 8 . Google Scholar OpenURL Placeholder Text WorldCat © The Author(s) 2021. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com
TI - Research on control strategy of battery-supercapacitor hybrid energy storage system based on droop control
JF - International Journal of Low-Carbon Technologies
DO - 10.1093/ijlct/ctab062
DA - 2021-09-11
UR - https://www.deepdyve.com/lp/oxford-university-press/research-on-control-strategy-of-battery-supercapacitor-hybrid-energy-15DmaGbr1V
SP - 1
EP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -