Fig.5 shows the general structure of each of the three sliding mode control loops.
The basic sliding mode control system of Fig.1 would suffer from control chatter, i.e., high frequency switching of the control variable, u, between its limits, ±U , and this might interact adversely with the switching of the inverter, so a boundary layer is introduced [1] in which the switching element is replaced by a saturation element with the same control saturation limits having a finite but relatively high gain, K.