+86 18988945661 
contact@iflowpower.com 
+86 18988945661
Language
+86 18988945661 contact@iflowpower.com +86 18988945661
Author ：Iflowpower – Portable Power Station Supplier
In this paper, the modular multi-level transformer (MMC) is used as a grid-connected transformer of the battery energy storage system, which can simultaneously control the flexibility of the high voltage grid. For battery energy storage modular multi-level transform (B-MMC) system, a hybrid model prediction control (H-MPC) method that can effectively reduce calculation amount is proposed. The H-MPC method consists of PI control and MPC.
Among them, the PI control section is used to obtain a sub-module access number of full-stream current output and loop-flow control; the MPC is responsible for common mode voltage (CMV) suppression, and the number of sub-module access is appropriately adjusted. A specific switch signal can occur in the case where the sub-module access is the sorting result of the battery pack charge state (SOC). For different applications, the PI control section and the control target of the PI control section and the MPC are more flexible.
Taking circulation control as an example, a brief analysis is made to the case of the MPC portion. Finally, the correctness and effectiveness of this method is verified by matlab / simulink simulation and experiment. The wide range of applications can effectively reduce people's demand for oil energy, and they have advantages such as strong renewable ability, emissions and pollution, and is important for alleviating energy crises and environmental deterioration [1].
Since new energy power generation has intermittent and uncertainty, it is usually combined with the energy storage device during the grid network. The energy storage device can perform fast power absorption, release, effectively reduce the new energy output fluctuations to the impact of the grid, realize the friendly access and coordination control of new energy [2], wherein the battery energy storage is occupied in a large-scale energy storage system. Important position.
Traditional energy storage and network system must string battery packs, in parallel, through the pre-stage DC-DC converter boost, passed the parallel inverter circuit. For battery packs, if you want to charge it, the discharge state monitoring should add additional battery energy management system, and the production cost will also improve accordingly; for power electronic converters, the voltage level required to be affected by the switching device is more High, systematic security will be seriously affected. When the battery energy storage system (BATTERYENERGYSTORAGESYSTEM, BESS) is combined with the modular multi-level converter (MMC), the distributed access of the energy storage unit can be realized, and the high voltage grid is enhanced to improve the operating efficiency of the system.
And reliability [3]. Current research on the BatteryIntegratedModularmultilevelContegratedModuLarmultilevelConvertilevertevertevertileVelconvertiler, B-MMC) system is still relatively small. Document [4] The balance between the chargeOfcharge state (SOC) is achieved by adjusting the modulation depth of each sub-module.
Document [5] is theoretically analyzed the impact of different cyclic components on the SOC balance of battery packs. In the literature [6], this structure is applied to the field of electric vehicles, and the three different working conditions of payment, DC charging and normal driving are respectively. In response to the control of the B-MMC structure, it is still mainly based on the classic PI controller.
Model PredictiveControl, MPC) method is gradually being promoted to power electronic control field due to its excellent advantage in processing nonlinear system complex constraints. [7]. A finitecontrol set model predictive control (FCS-MPC) method suitable for MMC structure is introduced in [8].
The method is simple, and the value function can cover multiple control targets, but when the number of submodules is more, the amount of data calculation is relatively serious. Document [9] decomposes the solving process of the overall value function into a plurality of sub-target functions, and the control method does not involve the selection of weight coefficients, which reduces the design difficulty of the control system. Document [10] combines the sorting actions and packet thoughts, derived an optimized MPC strategy for MMC engineering applications, which does not increase the burden of the processor computing with the increase of the number of sub-modules.
Based on the above research status, this paper proposes a hybrid model prediction control (HybridPredictiveControl, H-MPC) method suitable for B-MMC structure. H-MPC mode can be divided into two parts: PI control and MPC. The PI control part is important to implement a simple logic such as AC output current tracking, and the MPC part is used to process complex logic components such as common mode voltage suppression.
Compared to traditional PI control methods, H-MPC effectively reduces the number of PI controllers, reduces the design complexity of the control system; and the H-MPC is reduced in each sampling cycle. The number of switches, which thereby reducing the operational demand. This paper analyzes the operation characteristics of the B-MMC system, the traditional MPC method of MMC, and a H-MPC method is proposed, and an AC output current tracking, circulation control, battery pack SOC equalization and common mode voltage Inhibition control goal.
This article finally verifies the correctness and effectiveness of the control strategy through MATLAB simulation and experiment. Figure 7 Three-level B-MMC experiment platform author is the researcher of Shandong University Power Grid Intelligent Dispatch and Control Education Key Laboratory Researcher Li Nan, peak.
Copyright © 2024 iFlowpower - Guangong iFlowpower Technology Co., Ltd.
