Publications

Adaptive Control Design for Power Electronics Converters Using Kolmogorov-Arnold Networks

The design of control algorithm parameters has a significant impact on the reference tracking accuracy and efficiency of power electronic converters. Therefore, it is important to relate the control parameter design to the converter's performance metrics and operating conditions in order to achieve optimized performance. In conventional implementations, these parameters are precomputed and remain nearly constant throughout the converter's operating range. This can lead to sub-optimal performance when the metrics are highly dependent on the converter's loading conditions. Moreover, the relationship between control algorithm parameters, operating conditions, and performance metrics can be complex and difficult to model. In this paper, the Kolmogorov-Arnold Network (KAN) is proposed to derive an analytical expression that maps control parameters to converter performance under varying load conditions. The approach is demonstrated through the design of weighting factors in a finite-set model predictive control (FS-MPC) algorithm. Experimental validation confirms that the resulting analytical expression effectively controls the average switching frequency under different loading conditions. The proposed method offers high accuracy with low computational complexity.

Generic and Flexible Digital Model of DAB Converters With Various Topologies and Control for Efficient Optimal Design

To evaluate the performance of dual-active-bridge (DAB) converters across various topologies and control schemes for optimal design, this paper proposes a generic explicit digital model applicable two-level (2L)/three-level (3L) half-bridge (HB), full-bridge (FB), and three-phase (3P) topologies with a unified control structure. Voltage and current expressions for a single bridge leg considering both 2L and 3L topologies are developed in a generalized manner, and thus the model of DAB topologies with different number of legs can be constructed by superimposing the corresponding number of individual bridge leg models. Subsequently, the digital model, i.e., discrete-time model of DAB converters with different topologies and control configurations, is derived based on the circuit differential equations and a unified closed-loop control system by applying backward Euler method. Therefore, both steady-state and dynamic performance of DAB converters can be predicted in the proposed model. Owing to the analytical formulation of operating parameters and generic expressions for different topologies and control, the calculation speed of the proposed digital model can be accelerated compared to commercial simulation platforms such as MATLAB/Simulink and PLECS, and the design efficiency can be significantly enhanced, especially when it eliminates the need for frequent manual adjustments of topologies, modulation, and loss-thermal models of power devices. Simulation and experimental tests validate that the digital DAB model can achieve high accuracy under various topologies with fast execution speed.

Screening Small-Signal Stability Risks of Converter-Based Power Systems Using Balanced Truncation and Perturbation Theory

Screening small-signal stability risks in converter-dominated power systems comprising multiple black-box converter-based resources is challenging, since a high-dimensional operating point (OP) space is involved to identify the small-signal instability boundary (SSSB). This paper introduces a SSSB generation method that combines balanced truncation and eigenvalue perturbation theory within an eigenvalue-based analysis framework. The method systematically identifies the critical mode(s) that can become unstable under OP variations and efficiently computes the corresponding SSSB. Its effectiveness is demonstrated on a converter-based energy island type system, where the identified SSSB is validated against electromagnetic transient simulations. The computational efficiency of the proposed method and its expandability to large-scale power systems is quantified by comparison with an existing SSSB generation method, showing a substantial reduction of up to 99.6% in computational time.

Large-Signal Modeling of Virtual Admittance-Based Grid-Forming Inverter

As wind power capacities increase, the decreasing inertia of modern power systems challenges its frequency stability. Unlike traditional grid-following inverters, grid-forming (GFM) inverters can create and stabilize a grid independent of the utility. Hence, GFM inverters are seen as a potential solution, but their transient stability is a growing concern. While large-signal models are effective tools for analyzing transient stability, most existing research focuses on typical GFM control schemes without using virtual impedance. To address such a research gap, this paper proposes several simplified large-signal models for GFM inverters with virtual admittance, showing that the second-order model closely matches the accuracy of the full-order EMT model. So, the second-order large-signal model is a competing candidate for transient stability analysis due to its simplicity and accuracy. Finally, the correctness of the proposed models has been verified by time-domain simulations.

Advanced Grid Synchronization System for Power Converters under Unbalanced and Distorted Operating Conditions

This paper proposes a new technique for grid synchronization under unbalanced and distorted conditions, i.e., the dual second order generalised integrator - frequency-locked loop (DSOGI-FLL). This grid synchronization system results from the application of the instantaneous symmetrical components method on the stationary and orthogonal alphabeta reference frame. The second order generalized integrator concept (SOGI) is exploited to generate in-quadrature signals used on the alphabeta reference frame. The frequency-adaptive characteristic is achieved by a simple control loop, without using either phase-angles or trigonometric functions. In this paper, the development of the DSOGI-FLL is plainly exposed and hypothesis and conclusions are verified by simulation and experimental results

Årsrapport:Danmarks Forsknings- og Innovationspolitiske Råd 2020

A novel physics-inspired method for efficient and robust ultra-short-term photovoltaic power forecasting

Loss of Generation Ratio Analysis for Offshore Wind Farms

With the development of wind energy techniques, the capacity of offshore wind farms (OWF) is getting larger and the distance to shore is getting longer. This fact results in many different kinds of configurations of the electrical system design within the wind farm. Besides the costs of the wind farm system, which is concerned all the time, the reliability of the electrical system of the wind farm is also very important in the design phase. The index - loss of generation ratio probability (LOGRP) has been proposed to evaluate the reliability of the electrical system within an offshore wind farm. This paper provides further discussion of this index and applies it to a complex system. Comprehensive studies are conducted to investigate the effects of the component ratings, the network topology and other factors on LOGRP with a sample OWF as a reference. The reliability sensitivity is also discussed. The analysis presented in this paper is useful both for future wind farm planning and existing OWF evaluation

Experimental verification of an optimized control strategy for a medium-voltage DVR

Voltage sags are a major problem in present distribution systems. Therefore, different solutions have been examined to compensate these sags to avoid production losses at sensitive loads. In particular, Dynamic Voltage Restorers (DVRs) can realize this goal. Presently, a system wide integration of DVRs is hampered because of their high cost, in particular, due to the expensive DC-link energy storage devices. The cost of these DC-link capacitors remains high because the DVR requires a minimum DC-link voltage to be able to compensate a sag. As a result, only a small fraction of the energy stored in the DC-link capacitor can be used, which makes it almost impossible for the DVR to compensate relatively long voltage sags. Present control strategies are only able to minimize the distortions at the load or to allow a better utilization of the storage system by minimizing the needed voltage amplitude. To avoid this drawback an optimized control strategy is presented in this paper, which is able to reduce the needed injection voltage of the DVR and concurrently to mitigate the transient distortions at the load side. In the following paper a brief introduction of the basic DVR principle will be given. After this, an optimized control strategy will be briefly described. Finally, experimental results using a medium-voltage 10 kV DVR setup will be shown to verify and prove the functionality of the presented control strategy in both symmetrical and asymmetrical voltage sag conditions.

Physics-based Modeling of Degradation in Alkaline Water Electrolysis Cells due to Reverse Currents

The operation of water electrolyzers with renewable energies means that long-term operation also involves several start-up/shut-down cycles of the system. In particular, for alkaline water electrolyzers (AWE) it is known that shut-down leads to reverse currents, which causes electrochemical degradation and thereby reducing the efficiency. In this work, a modeling framework is proposed and implemented for the electrochemical degradation of nickel electrodes in AWE stacks under start-up/shut-down (SU/SD) operation, and it is achieving good agreement with experiments. Such a model can be used for digital twins of electrolysis plants as well as optimizing its operation.

Modeling of wind farm controllers

This paper describes modeling of wind farm controllers. The paper contributes to the IEA Annex XXI “Dynamic models of wind farms for power system studies” and was presented at the Annex XXI workshop in EWEC 2006. The wind farm controllers are developed with features, which make the wind farms able to participate in the control of the power system. The modeled wind farms provide active and reactive power control in the point of common coupling of the wind farm, which can be combined with automatic frequency and voltage control with droop and dead band.

A Single-Stage Bridgeless Buck-Boost PFC Converter with Two-Terminal Outputs

This paper proposes a single-stage bridgeless buck-boost power factor correction (SSBLBB-PFC) converter with two-terminal outputs. By adopting a minimized loop inductance design and using only one high-frequency switching device in both the boost and buck modes, the circuit loss has been significantly reduced. Additionally, the twoterminal output structure is more versatile than the conventional three-terminal output structure, and it can also reduce voltage stress. Additionally, the single inductor is reused to achieve power conversion in the positive and negative input voltage, achieving 100% utilization of the DC inductor. The dead beat control (DBC) is used to control the inductor current to achieve PFC. A detailed analysis was conducted on the working modes and control strategies of the converter. The simulation results under the 220 V power grid prove that the proposed inverter has a better dynamic performance.

Critical review of battery health state estimation with deep learning methods

Modeling of DC/DC Converter for DC Load Flow Calculation

With the development of power electronic techniques, more and more attention has been paid to DC power systems. This paper analyzes the load flow issues containing DC/DC converters in a DC power system. The load flow model of DC/DC converters, which considers the power losses and control strategy of the converter, has been presented and integrated into the traditional load flow algorithm by modifying the Jacobian matrix. Two typical DC/DC converter examples have been presented: the boost converter and full bridge converter

New State Observers and Sensorless Control of Wound Rotor Induction Generator (WRIG) at Power Grid with Experimental Characterization

The paper deals with a wound rotor induction generator (WRIG), also known as the doubly-fed induction generator. A complete experimental set-up is presented and analyzed. It is composed of: WRIG, two power electronics converters connected in the rotor side of WRIG, a line filter, and the data acquisition and control system (dSpace DS 1103). Both converters are commercial units and are vector controlled using appropriate interfaces. They are back-to-back connected, sharing the same DC bus, one supplied through a line filter from the power grid, and the other one with the output on the rotor of the generator. The stator of the generator is directly connected to the power grid. Two stator flux observer topologies were investigated and compared, one with the voltage model in parallel with the current model and the other one with both models connected in series. A speed estimation strategy, which works also during the synchronization procedure, was implemented and tested. It is based on model reference adaptive system (MRAS) principles. All control strategies, the flux observers and the MRAS, were developed in Matlab-Simulinkreg and implemented using a dSpacereg DS1103 single-board control and acquisition interface. Different tests were performed, and sample results are presented and discussed in the paper. The schemes used are illustrated in the paper, and the experimental results are shown and analyzed

A Novel Bipolar Rectifier with Output Short-Circuit Current Limitation for Microgrids

In microgrids, DC-side short circuits can produce high fault currents that threaten converter safety. Conventional converters often struggle to limit these currents effectively. This paper presents a bipolar single-phase AC/DC converter designed for microgrids, featuring inherent output short-circuit current limiting. To reduce fault current, a small DC bus capacitance is used, which poses challenges in suppressing voltage ripple. To address this, a ripple suppression control strategy is proposed, leveraging the converter's operating modes to maintain stable output under small capacitance conditions. The proposed solution improves system protection and enhances overall reliability.

Wind Energy Conversion

Performance Improvement of Sensorless Vector Control for Matrix Converter Drives Using PQR Transformation

This paper presents a new method to improve sensorless performance of matrix converter drives using PQR power transformation. The non-linearity of matrix converter drives such as commutation delay, turn-on and turn-off time of switching device, and on-state switching device voltage drop is modelled using PQR transformation and compensated using a reference current control scheme. To eliminate the input current distortion due to the input voltage unbalance, a simple method using PQR transformation is also proposed. The proposed compensation method is applied for high performance induction motor drives using a 3 kW matrix converter system without a speed sensor. Experimental results are shown to illustrate the feasibility of the proposed strategy