Santiago A. Verne

Predictive control of a back to back motor drive based on Diode Clamped Multilevel converters

In this work, a Back to Back converter based on a Diode Clamped Multilevel topology is presented. Both converters are controlled using model predictive approach. The control objective of the front-end converter is the DC bus voltage regulation with sinusoidal current and unity power factor, while the load converter drives an induction motor through predictive direct torque control. Both algorithms search for the best switching combination in order to optimize their corresponding control targets and also the voltage balance issue. Both switching sequence controllers are carried out autonomously with no need of information exchange about converter states. The performance of the conversion system is evaluated by means of computer simulations.

Active power filter with predictive current control of a Diode clamped Multilevel inverter

In this work, a predictive current control algorithm of an active filter built with multilevel diode clamped inverter is developed. Cycle by cycle current optimization is achieved through current prediction based on converter switching states and system parameters. The presented current control algorithm is naturally applicable to converters with an arbitrary number of levels with constant computational effort. The performance of the proposed algorithm is evaluated by means of computer simulations.

Predictive control of a back to back Diode Clamped Multilevel converter

In this work, a back-to-back Diode Clamped Multilevel converter system using Model Predictive Control is evaluated. DC link voltage balance and currents as main parameters are optimized by means of a proper cost function design. This strategy allows fast system response and capacitors voltage balance under various operating conditions including active power flow, different modulation algorithms, and different modulation indexes for the front-end and the load converters. The obtained results allows to conclude that proper independent selection of redundant states of both converters leads to good performance of the controlled variables including the balance of DC bus voltages.

Predictive control of a cascade asymmetric multilevel converter with split redundant states

This work presents a control strategy for the cascade asymmetric multilevel converter (CAMC). The analysis of the switching states of the topology is perfomed. A new voltage ratio for the phase capacitor, which splits redundant states and expands the number of levels from five to seven, is proposed. This improves the harmonic profile of the output voltage without additional switching devices. A Finite-states model predictive controller (FS-MPC) that uses simple dynamic models, jointly with properly defined cost functions performs the regulation of the internal voltages of the converter (phase capacitors and DC bus capacitors) and the line currents of a three phase load.

A hybrid predictive controller for a Cascade Asymmetric Multilevel Converter

This work presents a simplified predictive controller for the Cascade Asymmetric Multilevel Converter (CAMC). Instead of evaluating all target variables simultaneously (phase currents, voltages on the flying capacitors and midpoint voltage on the DC bus), the method uses the line voltage redundancy and the phase redundancy in a separate way and exploits their effects over each group of target variables. This scheme reduces the number of calculations when compared to the FCS-MPC with a multi-termed cost function. The method does not need for weighting factors and thus removes the need to tune the controller.

Modulación Vectorial de Inversores Multinivel con Diodos de Enclavamiento con Balance del Bus de Continua

In this work, a modulation algortihm for the Diode Clamped Multilevel Inverter with DC bus voltage balancing capability is presented. Space Vector modulation technique for the multilevel inverter is described, jointly with a DC bus voltage balancing algorithm that takes advantage from converters redundancy. The algorithm can be generalized to an arbitrary number of levels and takes into account two important cases: presence or absence of a DC voltage source on the DC bus. The performance of the presented algorithm is evaluated by means of computer simulations.

Predictive control of a single phase current source-fed multilevel converter

In this work, a single-phase multilevel current source DC/AC converter with sinusoidal output is presented. The voltage waveform is achieved by using the stepped current waveform of the multilevel current inverter and a single capacitor at the output. The proposed controller is based on the Finite-Control Set Predictive Control approach and regulates the output voltage simultaneously with the DC current on the converters internal inductor. The synthesis of a multilevel current at the output of the converter jointly with a single capacitor allows one to obtain a high quality output voltage at steady state with fast transient response capability without resonances.

Induction Motor Driven by a CAMC Using Predictive Control

This work presents a control strategy for an induction motor fed by a Cascade Asymmetric Multilevel Converter (CAMC). The approach is based on the Finite-States Model Predictive Control technique and uses the dynamic motor model and the converter model to pre-calculate the evolution of the variables of interest. These target variables are jointly compared with their reference values by means of an optimization function which allows to determine the best state among all switching combinations. This selection is carried out in a direct way, in each sampling period without modulator.