Computationally-efficient direct model predictive control of three- level neutral-point clamped back-to-back converters for wind turbine systems with permanent-magnet synchronous generators
Electronic Semiconductors, Components and Circuits, Integrated Systems, Sensor Technology, Theoretical Electrical Engineering
Final Report Abstract
In this project, we proposed two computationally efficient CE-DMPC schemes for 3L-NPC back-to-back power converters in wind turbine systems with direct-drive PMSG. More detailed conclusions are as follows: The derivation of the overall discrete-time mathematical model of the wind turbine system with PMSG and 3L-NPC back-to-back power converter required for the implementation of the DMPC schemes; the discussion and implementation of the classical DMPC scheme for torque and power control of the considered wind turbine system; the introduction, discussion, and analysis of two CE-DMPC schemes with either HCR or TCR. Both CE-DMPC schemes were applied to torque and power control of the considered wind turbine system with PMSG and 3L-NPC back-to-back power converter. Based on an intelligent but applicationdependent selection of the CRs, the possible reductions of the computational load for CE-DMPC with HCR and for CE-DMPC with TCR were analyzed theoretically: by CE-DMPC with HCR up to 55,6% and by CE-DMPC with TCR up to 83,0% of the computational load can be saved; the implementation of all three DMPC schemes in Mat lab/Simulink and their application to control of the derived wind turbine system model. The presented simulation results illustrated that the overall control performance of CE-DMPC with HCR and TCR is still acceptable and comparable to classical DMPC, although only a reduced number of switching states is evaluated during prediction and optimization; the implementation of all three DMPC schemes on an FPGA-based real-time system. It was shown that real-time implementation is feasible. Moreover, the theoretically derived bounds on the potential reduction of the computational load for the CE-DMPC schemes were validated on the FPGA-based real-time system: the computation times can be reduced by 55.1% with CE-DMPC with HCR and by 79.6% with CE-DMPC. Finally, all three DMPC schemes were experimentally tested at a labconstructed 3L-NPC back-to-back power converter PMSG wind turbine system prototype. The presented measurement results clearly illustrated that the control performances of CE-DMPC with HCR and TCR are still very satisfactory and comparable to the control performance of the classical DMPC scheme while significantly lower computation times were achieved. Concluding, in this project, we investigated the computationally efficient direct model predictive control methods for grid-tied three-level neutral point clamped back-to-back power converter PMSG wind turbine systems. Two computationally efficient solutions were successfully proposed and comprehensively evaluated at a lab constructed test-bench of the aforementioned system. With the proposed solutions, the computational load and time was significantly reduced. The significantly reduced computation time of the proposed CE-DMPC schemes eases real-time implementation on common industrial real-time platforms without the need of newer/more powerful processors or large FPGAs. Moreover, it might give room to combine the CE-DMPC schemes with online parameter estimation.
Publications
- ``Fully FPGA Based Performance Enhanced DMPC for Grid-Tied AFE with Multiple Predictions and Reduced Computational Efforts'', IECON 2015, Japan, 2015
Z. Zhang, Z. Chen, F. Wang, R. Kennel
(See online at https://doi.org/10.1109/IECON.2015.7392277) - ``Computational Efficient Direct Model Predictive Control for 3L-NPC Back-to-Back Converter PMSG Wind Turbine System'', SPEEDAM-2016, Capri, Italy, 2016
Z. Zhang, C. Hackl, R. Kennel
- “A Computationally- Efficient Quasi-Centralized DMPC for Back-to-Back Converter PMSG Wind Turbine Systems Without DC-link Tracking Errors”, IEEE Transactions on Industrial Electronics, 2016
Z. Zhang, T. Sun, F. Wang, J. Rodriguez, R. Kennel
(See online at https://doi.org/10.1109/TIE.2016.2573768) - ``Multiple Vector Model Predictive Power Control for Grid-Tied Wind Turbine System with Enhanced Steady State Control Performances”, IEEE Transactions on Industrial Electronics, 2017
Z. Zhang, H. Fang, F. Gao, J. Rodriguez, R. Kennel
(See online at https://doi.org/10.1109/TIE.2017.2682000) - “Computationally Efficient DMPC for Threelevel NPC Back-to-Back Converters in Wind Turbine Systems with PMSG”, IEEE Transactions on Power Electronics, 2017
Z. Zhang, C. Hackl, R. Kennel
(See online at https://doi.org/10.1109/TPEL.2016.2637081) - “Nonlinear Control Strategies for 3L NPC Back-to-Back Converter PMSG Wind Turbine Systems: Experimental Assessment Using FPGA'', IEEE Transactions on Industrial Informatics, 2017
Z. Zhang, F. Wang, J. Wang, J. Rodriguez, R. Kennel
(See online at https://doi.org/10.1109/TII.2017.2678500)
