ADAPTIVE MATHEMATICAL MODEL OF THERMAL PROCESSES IN A COSINE POWER CAPACITOR

Abstract

The existing protection and diagnostics systems are unable to detect the power capacitor abnormal heating caused by the development of its malfunctions. A diagnosis technique is presented that provides discovering such a heating at its early manifestation. The technique includes algorithms and an apparatus component in a form of a digital device, and it is based on continuous measuring of a capacitor body surface temperature, ambient temperature, voltages and currents from the power source. On the basis of measured values the active power losses in the capacitor and the temperature of the hottest point of its dielectric were calculated. Thereafter, the calculated average daily values of the temperature was analyzed, and, should the tendency of permanent increase of these values is detected, the diagnosis alarms of danger levels of abnormal heating are formed, viz. a low level, an average level, a high level and a very high level. The presented algorithms have been developed heuristically. Their final formation is possible only after years of operation of the proposed diagnosis system applied to the real objects. Because of application of the diagnosis system the probability of capacitor units’ failure will be lower and thus the dependability of the power supply may be higher. The implementation of the developed system will reduce the probability of sudden failure of capacitor units, and, correspondingly, will increase the reliability of the electricity supply system of an enterprise.