Abstract
This paper describes a new developed automated partial discharge (PD) measuring system, which is not only for the small scale PD measurement of sample test in laboratory, but also for the large scale after-laying test of power cable line on site. The system comprises an 8-channel fast A/D transformer, a central signal processing unit and four computers, providing real time measuring for monitoring eight signals at the same time automatically. The judgment of PD or discrimination of PD and noise can be divided into three phases. In the first phase, PD judgment is executed automatically based on multiple signal gates: f-gate for frequency, n-gate for pulse count rate, q-gate for pulse magnitude, φ-gate for PD phase position, t-gate for continuous time, set logically in serial or in parallel. The alarm signal in this phase has three levels determined by the gate output. When an abnormal signal is detected the abnormal signal channel can be automatically fixed and connected to other sub-units, and the signal will be judged in the next phase. In the second phase, the abnormal signal will be recognized by three pattern discrimination units: a neural network based on the φ-q-n pattern, a multiple frequency pattern logic gate based on f-q-t pattern, and a statistical x pattern logic gate (x represented source position of the signal) based on x-q-t pattern. In the third phase, the operator can give a final judgment according to the summarized information including the original waveform displayed by oscilloscope, spectrum analyzer, and the signal processing results shown with more than six kinds of 3 D graphics by computer display. Furthermore, a helpful offline processing can give several kinds of serial graphics based on time history or voltage history to help observations and comparisons. The system has been put into practical uses for several PD measurements and all results have shown the system can perform with high sensitivity of PD detection and high accuracy of discrimination of PD and noise. In this paper, some examples of practical PD measurement on site are introduced.
