Disturbance detection for optimal database storage in electrical distribution systems using artificial immune systems with negative selection

• The main disturbances in distribution systems were analyzed by the intelligent system.
• The intelligent system presented high-computing performance.
• The artificial immune systems have natural characteristics of pattern recognition.
• Results indicate that the proposed method is efficient (precise, fast, and robust).
• The abnormality database allows making decision, as well as contribute to the development of smart grid systems.

This paper presents the development of an intelligent system named “normal pass filter” to generate a disturbance database in electrical distribution systems. This is a system that aims to extract examples (and proper registration) of real disturbances from voltage and current measurements that are available by SCADA system. This filter is developed based on negative-selection artificial immune systems. The negative selection algorithm of an immune system is used to determine the presence of abnormalities. If an abnormality is detected, the system records the abnormal signal in a database. This database is a set of disturbance examples (e.g., harmonic, sag, high-impedance fault) for use in many purposes, for example, for training artificial neural networks for intelligent fault diagnosis and prognosis of electrical distribution systems. Recently, these diagnosis systems have been emphasized, particularly in smart grid environments. To exemplify the efficiency of the method, two electrical distribution systems with 33, and 134 busses were examined.