AC power

Electrical engineering

Compensation (psychology)

Mathematics

Space vector modulation

Stochastic neural network

Gene

Support vector machine

Psychoanalysis

Recurrent neural network

Engineering

Voltage

Model predictive control

Artificial intelligence

Psychology

Control (management)

Chemistry

Grid

Geometry

PWM rectifier

Computer science

Control theory (sociology)

Rectifier (neural networks)

Artificial neural network

Biochemistry

Robustness (evolution)

Pulse-width modulation

IEEE Journal of Emerging and Selected Topics in Power Electronics

Among various control strategies for pulsewidth-modulated (PWM) rectifiers, model-based predictive current control (MBPCC) based on space vector modulation (SVM) is a simple and effective method due to its quick dynamic response and good steady-state performance. However, SVM-based MBPCC (SVM-MBPCC) suffers from parameter variations and grid disturbances, which are common in practical applications due to temperature effects, saturation, grid faults, and so on. Recently, model-free predictive current control (MFPCC) based on current differences has been proposed to address the problem of parameter robustness. However, this approach suffers from stagnant updating of the current differences, and the steady-state ripples are still high due to the application of a single vector in one control period. This article proposes an SVM-MFPCC, which combines the merits of both SVM-MBPCC and MFPCC. The proposed method uses an ultralocal model rather than a conventional accurate model of the PWM rectifier, which is updated online based on the voltages and currents in past control periods. By using this ultralocal model, a voltage reference to nullify the current error can be calculated based on the principle of deadbeat control and is subsequently synthesized via SVM. Furthermore, by adding an appropriate compensation power to the original power references, the proposed method can achieve sinusoidal and balanced grid currents even under unbalanced and distorted grid conditions. The proposed method is compared to conventional SVM-MBPCC under various grid conditions, and the results of simulations and experiments confirm its effectiveness.

Model-Free Predictive Current Control of a PWM Rectifier Based on Space Vector Modulation Under Unbalanced and Distorted Grid Conditions