A new current control scheme for a 3-level inverter using neural networks

This paper presents the design of a new current controller for a three-level inverter of electric trains using neural networks (NNs). From past experience, the unbalance of DC-link voltage in three-level inverters causes a low power factor and increased harmonics to induction motors of traction vehicles. The proposed current control scheme offers "tight DC-link voltage balancing" for the inverter along with nonlinear mapping of essential switching patterns and fault tolerance, which are inherent characteristics of NNs. A computer simulation shows that a new current control scheme provides a comparable performance to a conventional sinusoidal pulse-width-modulation (SPWM) control in many aspects.