Plenary session

Plenary lecture „Advanced control methods of DC/AC and AC/DC power converters – look-up table and predictive algorithms” will be given by prof. Andrzej Sikorski

AC/DC/AC converters (back-to-back converters), as a whole, are used to connect induction motors of adjustable drives and also wind or water turbines renewable energy sources with the grid. In recent years a growing interest has been observed in AC/DC converters connecting DC voltage systems to the grid. This is due to their advantages such as availability of a bidirectional power flow and the unity power factor. AC/DC converters are widely used especially to connect different types of renewable energy sources, such as photovoltaic cells or fuel cells to the grid. The dependences describing the direction and speed of current vector change (the grid or IM stator current vector), caused by each voltage vector of the converter, enables the analysis and synthesis of look-up table direct torque and flux control of an induction motor fed by a DC/AC inverter as well as the direct power control of an AC/DC converter cooperating with the grid. The alternative nonlinear control approach is based on Model Predictive Control methods for AC/DC and DC/AC converters. Proposed control algorithms DPC-3V and DTC-3V belong to ICS-MPC methods. They are based on current error vector minimization criteria. The converter output set voltage vector is reproduced by SVM that provides the advantages typical for linear methods, i.e. constant switching frequency and low grid current or torque ripples. In transient states, they provide maximum dynamics comparable to the fastest nonlinear control methods. New FCS-MPC method for the voltage source rectifier with input LCL filter, similar to ICS-MPC will be presented. Differential equations describing the converter and the load are used to predict the direction and the rate of changes of the converter current and the filtering capacitor voltage vectors. Predicted values are used to define errors for every output voltage vector, which allows us to find the minimal value of the cost function. It results in choosing an optimal voltage vector and controlled quantities error mini-mization.