Elsevier

ISA Transactions

Volume 69, July 2017, Pages 51-64
ISA Transactions

Research article
Improved finite-control-set model predictive control for active front-end rectifiers with simplified computational approach and on-line parameter identification

https://doi.org/10.1016/j.isatra.2017.04.009Get rights and content

Highlights

  • For the first time, a Finite-Control-Set Model Predictive Control (FCS-MPC) method based on the model reference adaptive system (MRAS) with online parameter identification is proposed for active front-end rectifiers.

  • Compared with the conventional FCS-MPC method, the proposed strategy avoids the exhaustive exploration for testing all feasible voltage vectors, and reduces the eight possible voltage vectors to one. The optimizing efficiency can be improved, and the time-consuming of calculation can be reduced.

  • Compared with existing works, the proposed strategy enables relative accurate identification of the parameters, and further improves the steady-state performance for the whole system.

Abstract

In this paper, an improved finite-control-set model predictive control method is proposed for active front-end rectifiers where the computational effort and parameter mismatch problems are taken into account simultaneously. Specifically, a desired voltage vector which only requires one exploration is directly selected by using a single cost function, and the process of selection of the desired voltage vector is optimized by using a sector distribution method. Meanwhile, a model reference adaptive system-based online parameter identification approach is presented to alleviate the parameter mismatch problem. The advantages of the proposed method summarized as follows: First, the proposed algorithm reduces the eight possible voltage vectors to one. The exhaustive exploration can be avoided while the control performance is not deteriorated. Second, the proposed controller can mitigate performance degradation caused by the model parameter mismatch. Simulation results under various parameters operating conditions are presented to demonstrate the efficacy of the proposed method.

Introduction

Active front-end rectifiers (AFEs) offer several advantages compared to a diode rectifier. The advantages of AFEs include low-harmonic distortion of input alternating current (AC) currents, independent control of active and reactive powers, bi-directional power flow, and high-quality direct current (DC) output voltage. Successful applications can be found in diverse areas, including renewable energy sources [1], energy storage systems [2], [3], uninterrupted power supply (UPS) [4], smart grid technologies [5], and so on.

With the aim of flexible regulation of powers, various control schemes have been proposed for AFEs. Direct power control (DPC) is a well-known efficient control strategy [6], [7]. The DPC technique originates from the direct torque control (DTC) used in adjustable speed drives system. This strategy is based on the evaluation of the active and reactive instantaneous power errors values and the voltage vector position without any internal control loop for the currents. During its implementation process, a predefined lookup table or optimal switching sequence are previously established based on the instantaneous power behaviors of the converter. However, the main problem of this approach is that the resulting switching frequency is variable, and might increase the current harmonic distortions. This complicates the design of the output filter. To overcome this issue, an improved strategy based on space-vector modulation (SVM) is proposed in [8]. It uses the model system to calculate the required voltage in order to simultaneously eliminate active and reactive power tracking errors in one modulation period. Then, the required voltage is applied using a pulse width modulator. Constant switching frequency is thus achieved, and sinusoidal AC currents with low harmonic distortions can be obtained with relatively low sampling frequency. Recently, the model-predictive direct power control (MP-DPC) method based on the concept of a finite control set has been developed as a simple and effective control technique for AFEs owing to its simplicity and flexiblility without using any pulse-width modulation (PWM) blocks [9], [10], [11], [12]. The finite control set model predictive control (FCS-MPC) takes into account the limited number of switching states of the power converter for solving the optimization problem by using a simple iterative algorithm. Because of the simplicity of its implementation, the FCS-MPC technique has attracted great attentions from various research communities [13], [14], [15], [16], [17]. However, the FCS-MPC method selects an optimal voltage vector by evaluating all feasible voltage vectors. Hence, the FCS-MPC requires a large amount of calculation efforts. On a common test bench, the large amount of calculation efforts means long sampling intervals and a low switching frequency, which will degrade the system performance. In order to solve this issue, the simplification methods of the FCS-MPC algorithm are proposed and discussed [18], [19], [20], [21]. In [18], a simplified model predictive current control method is presented in order to reduce the amount of calculations in the practical implementation. The reference voltage space vector is obtained by utilizing the sector information, and the proposed method only needs a subset of all the available voltage vectors for the prediction and optimization. In [19], the main idea is to preselect voltage vectors to decrease switching losses at the next sampling period. The proposed algorithm reduces the eight possible voltage vectors to four. In [20], a control strategy of finite control set model predictive torque control (FCS-MPTC) with a deadbeat (DB) solution is proposed for PMSM drives. By using a deadbeat solution, the process of selections of the best switching vector is optimized. The proposed algorithm reduces the eight possible voltage vectors to three. In [21], the computation effort of FCS-MPC is greatly reduced by equivalent transformation and specialized sector distribution method. The proposed algorithm reduces the eight possible voltage vectors to two. Although aforementioned methods reduce the calculation efforts, there are barely any studies that address the eight possible voltage vectors to one.

Meanwhile, the performance of FCS-MPC system mainly depends on the accurate prediction of the future behaviors of the powers. Inaccurate model parameters can lead to the inaccurate prediction of the future input behaviors. This will result in the selection of an inappropriate switching pattern. For the accurate prediction of FCS-MPC algorithm, the system parameters such as the input coupling inductance and the input resistance must be estimated precisely. Recently, a great deal of efforts have been done in order to relieve the parameter mismatch problem, and several works based on parameter identification techniques have been reported to improve the performance of the control system [22], [23], [24], [25], [26]. In [22], a Luenberger observer is proposed for three-phase voltage source PWM rectifier. The problems of parameter mismatch and model uncertainty are addressed. The system robustness is improved. In [23], an AC-line-current reconstruction algorithm based on the DC-link current measurement is proposed by using a fuzzy logic controller. The parameter variations of line inductance is compensated. In [24], the authors present two simple, fast, and robust algorithms estimating the coupling inductance in voltage-source rectifiers (VSRs) controlled by the DPC method. The dynamics of both methods were proven to be quite fast. In [25], a novel strategy is proposed for the early detection of incipient faults, and enables a fast identification of the defective phase. A significant breakdown can be avoided by using this strategy because the faulty phase is rapidly, simply, and accurately identified. In [26], an adaptive online parameter identification technique is presented to overcome model mismatch and parameter uncertainty by using least-squares estimation for AC-DC AFEs. It should be noted that numerical examples are given to illustrate the effectiveness of the parameter estimation. However, these studies scarcely provide an accurate online estimation of the input coupling inductance that can be used to update the value of the controller by using a model reference adaptive system (MRAS) online identification technique in the system model.

This paper proposes an improved FCS-MPC method, which takes into account the computational effort and parameter mismatch problem simultaneously. To overcome the aforementioned problems, we combine the simplified computational approach and the design framework of MRAS-based online parameter identification technique. In order to reduce the high amount of calculation burdens, a desired voltage vector which only requires one exploration is directly selected by using a single cost function. The proposed algorithm reduces the eight possible voltage vectors to one. The exhaustive exploration is avoided while the control performance is not affected. The execution process of the control algorithm is shortened by reducing the calculation burdens, and thus, the corresponding sampling period will be further shortened. The system performance can be improved. Meanwhile, the theoretical background of the MRAS used for system model parameter estimation is introduced. A MRAS-based online parameter identification technique is presented to alleviate the parameter mismatch problem. The scheme used in this paper has significantly increased the accuracy of the estimation, and it is insensitive to variations of the system parameters. Furthermore, this control scheme can successfully suppress performance deterioration caused by parameter variations and model inaccuracy. In order to demonstrate the feasibility and effectiveness of the proposed scheme, simulation results will be provided and compared with the conventional FCS-MPC method in [10] and the proposed FCS-MPC method in [26].

The main advantages of the proposed methods are summarized as follows: (i) Compared with the conventional FCS-MPC method in [10], the proposed strategy avoids the exhaustive exploration for testing all feasible voltage vectors, and reduces the eight possible voltage vectors to one. The optimizing efficiency can be improved, and the time-consuming of calculation can be reduced. (ii) Compared with the proposed FCS-MPC method in [26], the proposed strategy not only identifies the parameter accurately, but also mitigates performance degradation resulting from the model parameter mismatch. Besides, with the MRAS-based online parameter identification technique, robustness performances against model inaccuracy and parameter variations can be further obtained.

This paper is organized as follows: Section 2 introduces FCS-MPC method of the AFEs. In Section 3, the proposed online parameter identification method for AFEs is derived. In Section 4, the simplified computational method for AFEs is described. Section 5 provides the simulation results to illustrate the proposed control scheme. Section 6 concludes this article.

Section snippets

FCS-MPC method of AFEs

The common topology of the AFEs is shown in Fig. 1. It is composed of a three-phase full bridge converter with six power transistors which are connected to the grid voltage eg through the coupling inductances Lg and the input resistances Rg.

Based on Fig. 1, the mathematical model of AFEs in the αβ stationary reference frame is characterized by the following equationseg=23(ega+αegb+α2egc),ig=23(iga+αigb+α2igc),where α=ej(2π/3). The rectifier voltage vector is calculated from the switching state

Online parameter identification method for AFEs

In this section, a MRAS-based online parameter identification technique is presented to alleviate the parameter mismatch problem. In order to resolve this problem, different control strategies are developed for the coupling inductance estimate to handel the mismatch that inherently exist in their feature. In addition, due to the proposed model based adaptive law that has been included in the controller design, the behavior of the system is robust and less sensitive to system parameters values

Simplified computational method for AFEs

In the FCS-MPC strategy, all the possible rectifier voltage vectors are used to predict the future input real and reactive powers. However, the conventional FCS-MPC method has a significant number of switching states, making the implementation of the FCS-MPC strategy challenging as the computational burden becomes a major problem. To resolve this problem, the proposed FCS-MPC algorithm excludes seven voltage vectors during the next sampling period, and a desired voltage vector which only

Simulation results

To verify the feasibility of the proposed adaptive scheme for the coupling inductance estimations, simulation results are presented with a sampling period of Ts=50μs in this section. Since the conventional FCS-MPC method in [10] is commonly used, it is used here as a benchmark for the proposed strategy to compare to. In order to achieve a fair comparison between the three methods, the external PI controllers are configured with the same parameters. The system and control parameters are

Conclusions

In this paper, the FCS-MPC scheme is proposed by combining the simplified computational approach and MRAS online parameter identification technique for AFEs. Firstly, the FCS-MPC method and DB-PDPC technique with SVM are brought together to reduce the high amount of calculation efforts. Compared with the conventional FCS-MPC method, the proposed strategy avoids the exhaustive exploration for testing all feasible voltage vectors, and reduces the eight possible voltage vectors to one. The

Acknowledgement

This work was supported in part by the National Natural Science Foundation of China under Grants 51579023, 61673081, and in part by the China Postdoctoral Science Foundation under Grant 2015M570247, and in part by the Fundamental Research Funds for the Central Universities under Grants 3132016201, 3132016313, and in part by the National Key Research and Development Program of China under Grant 2016YFC0301500, and in part by High Level Talent Innovation and Entrepreneurship Program of Dalian

References (35)

  • k. Heussen et al.

    Unified system-level modeling of intermittent renewable energy sources and energy storage for power system operation

    IEEE Syst J

    (2012)
  • B.M. Ge et al.

    An energy-stored quasi-Z-source inverter for application to photovoltaic power system

    IEEE Trans Ind Electron

    (2013)
  • B.M. Ge et al.

    Novel energy stored single-stage photovoltaic power system with constant DC-link peak voltage

    IEEE Trans Sustain Energy

    (2014)
  • P. Cortés et al.

    Model predictive control of an inverter with output LC filter for UPS applications patricio

    IEEE Trans Ind Electron

    (2009)
  • A.P.S. Meliopoulos et al.

    Smart grid technologies for autonomous operation and control

    IEEE Trans Smart Grid

    (2011)
  • J.B. Hu et al.

    Investigation on switching patterns of direct power control strategies for grid-connected DC-AC converters based on power variation rates

    IEEE Trans Power Electron

    (2011)
  • J.A. Restrepo et al.

    Optimum space vector computation technique for direct power control

    IEEE Trans Power Electron

    (2009)
  • A. Bouafia et al.

    Predictive direct power control of three-phase Pulse Width Modulation (PWM) rectifier using Space-Vector Modulation (SVM)

    IEEE Trans Power Electron

    (2010)
  • J.F. Hu et al.

    Model predictive control of grid-connected inverters for PV systems with flexible power regulation and switching frequency reduction

    IEEE Trans Ind Appl

    (2015)
  • P. Cortés et al.

    Direct power control of an AFE using predictive control

    IEEE Trans Power Electron

    (2008)
  • S. Vazquez et al.

    Direct power control for three-phase power converters

    IEEE Trans Ind Electron

    (2008)
  • D.E. Quevedo et al.

    Model predictive control of an AFE rectifier with dynamic references

    IEEE Trans Power Electron

    (2012)
  • S. Vazquez et al.

    Model predictive control: a review of its applications in power electronics

    IEEE Trans Ind Electron Mag

    (2014)
  • S. Kwak et al.

    Predictive control method with future zero-sequence voltage to reduce switching losses in three-phase voltage source inverters

    IEEE Trans Power Electron

    (2015)
  • D.K. Choi et al.

    Dynamic performance improvement of AC/DC converter using model predictive direct power control with finite control set

    IEEE Trans Ind Electron

    (2015)
  • S.A. Davari et al.

    An improved FCS-MPC algorithm for an induction motor with an imposed optimized weighting factor

    IEEE Trans Power Electron

    (2012)
  • T. Geyer et al.

    Model predictive direct torque control-part I: concept, algorithm, and analysis

    IEEE Trans Power Electron

    (2009)
  • Cited by (32)

    • Lyapunov-based model predictive control of dual-induction motors fed by a nine-switch inverter to improve the closed-loop stability

      2023, International Journal of Electrical Power and Energy Systems
      Citation Excerpt :

      Regarding the control of the numerous control goals, the MPC provides a good solution due to its simple feedback structure [6,7]. The control goals can be easily added to the MPC loop, and the control constraints related to these control goals can be introduced as system restrictions [8–10]. The ease of including the desired control goals enable to use of MPC in induction motor control applications.

    • Finite control set model predictive control approach of nine switch inverter-based drive systems: Design, analysis, and validation

      2021, ISA Transactions
      Citation Excerpt :

      This strategy is simply called receding horizon control. Because of its capability to handle constraints and add nonlinearities, MPC has been used in many power electronics applications [15–17]. To verify the all theoretical analysis, several simulation works are conducted.

    • Model predictive control of microgrids – An overview

      2021, Renewable and Sustainable Energy Reviews
      Citation Excerpt :

      This is different from undermentioned grid-level MPC which usually employs a specific toolbox to help solve a more complex algorithm. It is worth mentioning that various converter topologies can be applicable in Fig. 2, including but not limited to dc-dc [37–39], dc-ac [40–42], ac-dc [43–45], ac-ac [46,47], multi-level converters [48–50], or the converters with complex topology [51–53]. During these years, converter-level MPC has been actively developed.

    • Multi-level model predictive controller with satisfactory optimization for multi-level converters

      2019, Simulation Modelling Practice and Theory
      Citation Excerpt :

      Its discrete nature allows its direct implementation and eliminates the necessity of PWMs. It is commonly used in controlling DC-DC converters and AC-DC rectifiers [18,19]. FS-MPC is also applied in multilevel inverters such as the ANPC [20] and the Packed-U-Cells (PUC) single-phase multilevel inverters[8,21], as it exploits the limited number of voltage states to conduct predictive computations.

    View all citing articles on Scopus

    Fully documented templates are available in the elsarticle package on CTAN.

    View full text