Efficient processing of alarms avalanche using artificial neural networks and classification techniques

In modern control centers, system operators usually have to handle a large number of alarms in real time, and to take decisions on power system operation. In these conditions, the location and the failed element must to be determined in very short time; else the system can continue to deteriorate and, in some extreme cases, can achieve the collapse. This work presents an efficient processing of alarms avalanche using artificial neural networks and classification techniques based on a multilayer perceptron formed by an input vector consisting of the 58 predefined protection signals, which are most representative transmission line alarms. The proposed methodology has been tested on events in the Peruvian Interconnected System and the results show its excellent capacity to identify the failed element.