A new control scheme for variable speed wind turbines using neural networks

In this paper a new neurocontrol scheme for a variable speed wind turbine (VSWT) is presented. A voltage source converter cascade is used employing neurocontrol for the generator side converter and three independent hysteresis controllers for the grid side, respectively. A new method for estimating the optimal rotating speed is also proposed. The repercussions from the connection of a wind park (WP), composed of wind turbines (WTs) of this type, on the operation of weak electric distribution systems are also analyzed. The effects on the voltage profile caused by variable speed (VS) wind turbines are compared to the effects caused by fixed speed (FS) WTs. Both types of WTs are assumed to be equipped with asynchronous machines. The advantages of the studied VS operation scheme are confirmed and the possible increase in the installed capacity of the WP over the FS mode, maintaining the same power quality standards, is estimated.