Fault detection in copper-rotor SEIG system using artificial neural network for distributed wind power generation

Too much dependence on large, polluting and expensive generation is no longer an option that Canadians would endorse in this era of distributed generation through renewable energy systems. Understanding the significance and prospects of self-excited induction generators (SEIGs) in distributed wind power generation, this paper presents an exclusive study of fault and a artificial neural network (ANN) based technique for its detection across the stator terminals of the SEIG. Firstly, two-axis model of a 7.5 hp industrial copper-rotor SEIG is developed to perform numerical investigations under static loading conditions, faulty conditions and hence derive data for designing the ANN based detection scheme. Fault tolerant capability of the machine is experimentally elicited by applying a short-circuit fault across the terminals of the machine and the need for fault detection in the SEIG system is discussed. Lastly, a novel ANN based scheme is developed for fault detection and numerical investigations are performed to illustrate the performance of the developed scheme. This paper aims to provide a good study to understand and develop a ANN based device for fault detection in a SEIG system.