Search

Due the development of the renewable energy installations, the equipment for these systems has known substantial progress. Such equipment must meet a number of conditions both on power safety as well on the quality of energy supplied. As is recognized in literature, the power quality is given by frequency, voltage, harmonic pollution, source-consumer, interdependence, interdependence-source system (smart grid). Power converters DC/DC, DC/AC are the reference equipment in a renewable energy system and because of quickly evolution of the power electronic, allowing these to be increasingly complex. The paper present a monitoring/diagnosis system for an inverter, using an artificial neural network. The training data for artificial neural network are obtained through EMTP simulation of different cases of failure in switching system.

The renewable energy system contains the equipment which must work in properly conditions for different operating states. These devices must have a high reliability in order to not disturb the functionality of the entire system, when some faults or defects occur during the normal running. The equipment might be monitored through traditional methods (from main parameters: current, voltage, frequency, operating times, etc) also through modern methods like electrical and magnetic fields. This paper presents the monitoring and diagnosis of the DC/AC inverter from one 5 kW wind turbine, off grid connection, in case of supplying of different loads and in case of several faults. The monitoring is performed according with the electromagnetic field radiation generated by inverter for these operating regimes.