Robust predictive current control of PWM rectifiers with LCL filters under unbalanced and distorted network conditions

Abstract

Pulse-width modulation (PWM) rectifiers with LCL filters can achieve better filtering and have lower total inductance than its counterpart of L filters. However, the control complexity is also increased due to the higher system order. Furthermore, the performance may deteriorate significantly when the inductance used in the controller deviates from its actual value due to saturation, temperature, and so on. This paper proposes a robust predictive current control (RPCC) scheme based on the ultra-local model for LCL-filtered PWM rectifiers. Excluding the converter voltage related term, the other part of the system F including the disturbance caused by parameter mismatches and other dynamic process, can be estimated using differential algebra (DA) or an extended state observer (ESO). After obtaining F, the reference voltage vector can be calculated based on the principle of deadbeat current control. The proposed RPCC is extended to an unbalanced and distorted grid by modifying the grid current reference; thus, the three-phase grid currents are balanced and sinusoidal. The proposed RPCC based on DA and ESO is compared to the conventional model-based predictive current control (MPCC), and the experimental results confirm its robustness and effectiveness under mismatched parameters and network disturbances.