Model Predictive Current Control Algorithm Research Articles

Model Predictive Current Control of an Induction Motor Considering Iron Core Losses and Saturation

The paper considers the model predictive current control (MPCC) of an induction motor (IM) drive and evaluates five IM models of different complexities—from conventional to magnetic saturation, iron losses, and stray-load losses—for the MPCC design. The validity of each considered IM model and the corresponding MPCC algorithm is evaluated by comparison of the following performance metrics: the total harmonic distortion of the stator current, the average switching frequency, the rotor flux magnitude error, the rotor flux angle error, and the product of the first two metrics. The metrics’ values are determined in wide ranges of the rotor speed (0.1–1 p.u.) and load torque (0–1 p.u.) through simulations performed in the MATLAB Simulink environment. The obtained results allow us to identify the IM model that offers the best tradeoff between the practicability and accuracy. Furthermore, a control effort penalization (CEP) is suggested to reduce the average switching frequency and, hence, the power converter losses. This involves constraining the simultaneous switching to a maximum of two branches of the three-phase power converter, as well as inclusion of the weighted switching penalization term in the cost function. Finally, the performance—both steady-state and dynamic—of the proposed MPCC system with CEP is compared with that of the analogous field-oriented controlled (FOC) IM drive. The inverter switching frequency is reduced more than twice by including the frequency-dependent iron-loss resistance in the MPCC. It is additionally reduced by implementing the proposed CEP strategy without sacrificing many other performance metrics, thus achieving a performance comparable to the FOC IM drive.

Model predictive and SoC balancing control of a CHB inverter in battery energy storage application using reward functions

SummaryThis article presents an improved model predictive current control algorithm combined with a novel state of charge (SoC) balancing approach for a three‐phase cascaded H‐bridge inverter in battery energy storage applications. Based on the model predictive control with elimination of redundant voltage vectors, neighboring vectors of the last voltage vectors are used to find the optimal ones in only one control cycle, which solves the problems of high computational burden and slow dynamic response. To overcome the unsuitability of the conventional cost function in SoC balancing, the reward functions that effectively balance the inter‐phase SoCs, inner‐phase SoCs, and inner‐phase dc voltages are presented respectively. Simulation and hardware‐in‐the‐loop‐based experimental results validate the effectiveness of the proposed strategy.

Low complexity duty cycle modulation model predictive current control of permanent magnet synchronous motor

This study proposes a model predictive current control algorithm for the synchronous permanent magnet motor with low complexity duty cycle modulation. Three single-phase voltage vectors are simulated to synthesize the target voltage vector that enables the current to achieve deadbeat control. The algorithm can change the amplitude and direction of the synthesized vector and extend the selection range of the synthesized vector to the whole hexagonal region. That is to say, this method can realize the deadbeat control of the d-axis and q-axis currents, and reduce the current pulsation. Moreover, this method only needs one action time operation, which can directly convert the action time of the voltage vector into a three-phase duty cycle after equivalent processing and over-modulation of the action time of the voltage vector. It is not necessary to traverse multiple sectors for optimization, which simplifies the modulation method and reduces the computational complexity. The effectiveness and reliability of the proposed method are verified by theoretical derivation and experimental results.

Improved Model Predictive Current Control of NPC Three-Level Converter Fed PMSM System for Neutral Point Potential Imbalance Suppression

For the neutral point clamped (NPC) three-level converter fed permanent magnet synchronous motor (PMSM) system, the performance of the conventional model predictive current control (MPCC) algorithm will be deteriorated if the amplitude of the neutral point potential (NPP) is large. Additionally, the adjustment process of the weighted coefficients of the conventional MPCC algorithm is complex because of numerous control terms in the cost function. To solve the above issues, an improved MPCC algorithm is proposed in this paper. Firstly, Newtonian iteration is used to transfer the stator current into stator voltage in the cost function. Then, the NPP term in the conventional cost function can be eliminated by introducing the partition control of the NP potential, which also eliminates the whole adjustment process of weighting coefficients. Finally, based on the amplitude of the NPP, the amplitude and phase angle of medium and small vectors are modified to improve the control performance of the torque and flux. Experimental results show that the fluctuation of the neutral point potential can be suppressed rapidly. Meanwhile, the performance of the torque, flux and current are also improved compared with the conventional MPCC.

Deadbeat two-vector model predictive current control for open-winding primary permanent-magnet linear motor

The structure of open-winding motor fed by the dual-inverter can increase the winding terminal voltage of the primary permanent-magnet linear motor (PPMLM). So, it is an effective method to improve the thrust performance of the PPMLM. Based on the detailed derivation of the mathematical model of the open-winding PPMLM and the in-depth analysis of the space voltage vector distribution characteristics of the dual-inverter, a deadbeat two-vector model predictive current control (MPCC) model is proposed in this paper. First, the optimal vector of the inverter 1 is determined by judging the position of the large sector. Second, two optimal vectors of the inverter 2 is determined by judging the position of the small sector. Finally, the duty cycle of the two optimal vectors of the inverter 2 is calculated to realize the proposed deadbeat two-vector MPCC algorithm. Compared with the traditional model predictive torque control (MPTC) and MPCC strategies, the proposed algorithm has the advantages of lower current harmonics (lower 34 % and 44 % under rated conditions, respectively) and better thrust performance (higher 77 % and 61 % under rated conditions, respectively) as well as lower computational complexity. Simulation and experimental results show that the proposed two-vector MPCC has good steady-state and dynamic performances, which can reduce the current harmonics and thrust ripple, and thus improve the motor drive performance compared with the traditional method.

Model Predictive Current Control Algorithm Based on Joint Modulation Strategy for Low-Inductance PMSM

The disc coreless permanent magnet synchronous machine (PMSM) has many advantages. The inductance is far lower than that of the machines with the core structure. However, the general converter is designed for general machines. The current fluctuations are large when the low-inductance PMSM (LIPMSM) is driven by the general converter. To solve the above problem, the integrated design concept is adopted to design the converter, including the modulation algorithm, control algorithm, and the corresponding topology. The joint modulation strategy is proposed to increase the modulation frequency, which adjusts the output voltage by using the bus circuit and inverter circuit. The model predictive current control (MPCC) algorithm based on the joint modulation strategy is proposed, which has a small computational burden and high control accuracy. The most important feature of the proposed MPCC is that it has a modulation algorithm. To further reduce the current fluctuations, the bus voltage is adjusted in real time according to the bus voltage optimization algorithm. A new topology of the converter is designed to implement the algorithm function. Finally, the correctness and feasibility of the proposed method are verified. Compared with the general converter, the new one has a great improvement on the LIPMSM.

Model Predictive Control and Position Sensorless Control Algorithm for Induction Motor

In view of the traditional finite set model predictive control of induction motor, the enumeration method is used to select only one optimal voltage vector in one control cycle, which causes the deviation of stator current in the tracking control process, and a duty cycle model predictive current control algorithm is proposed. In this algorithm, duty cycle control is introduced on the basis of traditional model prediction algorithm. In a control period, an optimal non-zero vector and a zero vector act together, and the action time of the two is calculated. Based on the duty cycle model predictive control algorithm, the model reference adaptive speed estimation algorithm is introduced to improve the robustness of the model. Finally, the dynamic and static characteristics of vector control and two model predictive control algorithms are analyzed through simulation comparison, and the feasibility of the proposed speed estimation algorithm is verified on the basis of the duty ratio model predictive algorithm, and the robustness of the two estimation methods is compared. The superiority of the prediction algorithm of duty cycle model and the improved speed estimation algorithm is proved.

Model predictive current control for two-phase switches clamping of three-phase inverter

Traditional finite-control-set model predictive control has been widely concerned for heavy power voltage source inverters owing to its simplicity, as well as control flexibility. However, the drawbacks of widespread current harmonic, high switching loss and heavy computational burden are fully exposed. In this article, three-phase voltage source inverter is taken as the research object and a model predictive current control algorithm for two-phase switches clamping is proposed. Firstly, deadbeat control and delay compensation strategy are combined to calculate reference voltage vector and determine two legs that have maximum and minimum reference voltage vectors. Then, the optimal voltage vector set can be directly selected by clamping switches on two legs that have maximum and minimum reference voltage vectors, which decreases the number of commutations on three legs with a large current. The proposed algorithm not only reduces the switching loss and computational burden, but also increases the efficiency of operation. The simulation experiments are implemented to estimate the control effect of the proposed algorithm. Comparison results indicate the proposed algorithm can be more effective than existing algorithms.

Improved Model Predictive Current Control of Single-Phase Five-Level PWM Rectifier

In this paper, a single-phase five-level rectifier with coupled inductors is studied. First, a discrete mathematical model of a single-phase five-level rectifier is created in two-phase(αβ). A traditional single vector finite control set model predictive control (FCS-MPC) algorithm is improved to overcome the problems of a varying switching frequency, the large amount of time needed for calculation, and the inaccurate setpoint of current loop tracking. Then, the objective function of the system is established, and a simple objective function is used instead of an iterative optimization of the traditional FCS-MPC algorithm. At the same time, to eliminate the delay error and to reduce harmonic distortion, deadbeat control technology is introduced. Finally, the simulation and experimental results show that the improved model predictive current control algorithm not only retains the fast response of traditional model predictive control, but also has the advantages of fixed switching frequency, small calculation time, and small current steady-state error.

Model Predictive Current Control of Two-Level Four-Leg Inverters—Part II: Experimental Implementation and Validation

This paper describes the experimental implementation of the model predictive current control algorithm for a two-level four-leg inverter operating under balanced, unbalanced, and nonlinear loading conditions. The proposed scheme is designed to predict the future behavior of the load currents for each of the 16 possible switching states of the converter. The control method chooses a switching state that minimizes the error between the output currents and their references. The algorithm is embedded using a MATLAB/Simulink software environment, and experimental results based on the dSPACE DS1103 controller are provided. These results verify the advantages of the proposed control strategy in terms of robustness under filter and load parameter variations, average neutral-leg switching frequency reduction, reference tracking error, and percentage total harmonic distortion.

Model Predictive Control of Multilevel Cascaded H-Bridge Inverters

This paper presents a model predictive current control algorithm that is suitable for multilevel converters and its application to a three-phase cascaded H-bridge inverter. This control method uses a discrete-time model of the system to predict the future value of the current for all voltage vectors, and selects the vector which minimizes a cost function. Due to the large number of voltage vectors available in a multilevel inverter, a large number of calculations are needed, making difficult the implementation of this control in a standard control platform. A modified control strategy that considerably reduces the amount of calculations without affecting the system's performance is proposed. Experimental results for five- and nine-level inverters validate the proposed control algorithm.