A Model Predictive Control Method for Hybrid Energy Storage Systems


The traditional PI controller for a hybrid energy storage system (HESS) has certain drawbacks, such as difficult tuning of the controller parameters and the additional filters to allocate high- and low- frequency power fluctuations. This paper proposes a model predictive control (MPC) method to control three-level bidirectional DC/DC converters for grid-connections to a HESS in a DC microgrid.


First, the mathematical model of a HESS consisting of a battery and ultra capacitor (UC) is established and the neutral point voltage imbalance of a three level converter is solved by analyzing the operating modes of the converter. Secondly, for the control of the grid-connected converters, an MPC method is proposed for calculating steady state reference values in the outer layer and the dynamic rolling optimization in the inner layer. The outer layer ensures the voltage regulation and establishes the current predictive model, while the inner layer, using the model predictive current control, makes the current follow the predictive value, thus reducing the system current ripple.


This cascaded topology has two independent controllers and is free of filters to realize the high-and low frequency power allocation for a HESS. Therefore, it allows two types of energy storage devices to independently regulate the voltage and realizes the power allocation of the battery and UC. Finally, simulation studies are conducted in PSCAD/EMTDC, and the effectiveness of the proposed HESS control strategy is verified in a case, such as a controller comparison and fault scenario.


  1. Double layer control method
  2.  Hybrid energy storage system (HESS)
  3. Model predictive control (MPC)
  4. Three-level DC/DC converter



Fig. 1. The topology of a HESS.


Fig. 2. Comparison under the proposed MPC method with the PI control method. (a) Bus voltage response under two methods. (b) UC voltage response under two methods.

Fig. 3. The power response of the battery when the controller parameter _i is changed.

Fig. 4. System voltage response in the case of short-circuiting of the UC.

(a) The response of the bus voltage during the short-circuit fault of the UC. (b) The response of the UC voltage during the short-circuit fault of the UC.

Fig. 5. Photovoltaic module output power.

Fig. 6. Comparison between the proposed MPC method and the PI control method. (a) The bus voltage response under two methods. (b) The UC voltage response under two methods.

Fig. 7. The effectiveness under the proposed MPC method. (a) The voltage of capacitors C1 and C2. (b) Inductor L1 reference current iL1ref and actual current iL1.


In this paper, the advantages of a three-level bidirectional DC/DC converter for battery/UC HESS and the effectiveness of the proposed MPC method are discussed from both a theoretical analysis and simulation verification. At the same grid voltage level, the battery can suppress higher voltage level fluctuations after a two-stage boosting structure.


Compared with the PI controller, the MPC controller doesn’t need a tedious step of adjusting parameters and various state variables are considered in each sampling instant. Moreover, the MPC algorithm based on the constant switching frequency achieves fast and accurate regulation of voltage and current with diminished ripples. Finally, the system does not need filters to allocate power fluctuations, and the control structure is optimized while the battery life is prolonged.


[1] D. Liang, C. Y. Qin, S. Y. Wang, and H. M. Guo, “Reliability evaluation of DC distribution power network,” in Proceedings of 2018 China International Conference on Electricity Distribution, Tianjin, 2018, pp. 654–658.

[2] Z. Huang, J. Ma, J. Zeng et al., “Research status and prospect of control and protection technology for DC distribution network,” in Proceedings of 2014 China International Conference on Electricity Distribution, Shenzhen, Sep. 2014, pp. 1488–1493.

[3] K. A. Joshi and N. M. Pindoriya, “Case-specificity and its implications in distribution network analysis with increasing penetration of photovoltaic generation,” CSEE Journal of Power and Energy Systems, vol. 3, no. 1, pp. 101–113, Mar. 2017.

[4] Y. Xu, T. Y. Zhao, S. Q. Zhao, J. H. Zhang, and Y. Wang, “Multiobjective chance-constrained optimal day-ahead scheduling considering BESS degradation,” CSEE Journal of Power and Energy Systems, vol. 4, no. 3, pp. 316–325, Sep. 2018.

[5] Y. Sun, Z. Zhao, M. Yang, D. Jia, W. Pei and B. Xu, “Overview of energy storage in renewable energy power fluctuation mitigation,” CSEE Journal of Power and Energy Systems, vol. 6, no. 1, pp. 160–173, Mar. 2020.

Leave a Reply

Your email address will not be published. Required fields are marked *