Distortion suppression based on adaline for single-phase SPWM inverter

Power switching components´ dead time, voltage drop and the change in load introduce the distortion in the output of SPWM inverter. This kind of outputs will trigger a series of serious problems, like lowering the stability of running equipment, increasing the loss, etc. In this paper, based on the causes of different kinds of distortion, aiming at the dead-zone effect, a algorithm to minimize dead time and to reduce the output zero-crossing distortion is given; aiming at the change of model parameters and load, based on adaptive linear neural network, a algorithm about distortion detection and compensation is proposed. Finally,using STM32F103ZE6 (Cortex-M3 32-bit ARM microprocessor), single-phase SPWM inverter results show that the distortion of output is only 0.9%, lower than about 85% before compensation.