Experimental studies with simple neuro-controller based excitation controller

The implementation and experimental verification are described of a simple neurocontroller (SNC) as an excitation controller for a physical model of a single-machine infinite-bus power system. The SNC with one weight, one bias and a linear hard-limit activation function is trained online. A modified error function with additional variables, used for training, ensures good performance and allows additional control functions to be performed. The SNC is used to control the synchronous generator terminal voltage and to function as a power system stabiliser at the same time. The SNC has been implemented using a DSP board mounted on a PC. Its performance has been investigated by experimental tests for various operating conditions and disturbances.