Research On Partial Discharge Location Of Transformer Based On Ultrasonic Array Sensor

As the core of energy conversion in the process of power transmission and distribution,transformers are the most important and critical power equipment in grid operation.Transformer fault diagnosis and online monitoring are of vital importance to the safety of the entire grid operation.Insulation fault is the biggest factor causing transformer fault.Partial discharge is the most fundamental cause of insulation failure.Therefore,it is very important to accurately determine and identify the partial discharge position of the transformer.In this paper,a planar cross array monitoring system for partial discharge fiber FP sensors for transformer high and low voltage windings is proposed.Directional algorithm:Firstly,the traditional positioning and orientation algorithm:Steered Response Power(SRP)sound source localization theory and Multiple Signal Classification(MUSIC)algorithm are used to be studied for the partial discharge localization and direction finding of the transformer.Then the MUSIC-SRP positioning theory is proposed to locate the local discharge of the high and low voltage windings of the transformer.After that,through the simulation of partial discharge in multiple voltage grade transformers’ windings.It is found that music-srp has higher positioning accuracy and better applicability.Finally,the internal partial discharge experiment of the transformer in the oil proves that the MUSIC-SRP positioning algorithm can be used in the inter-turn partial discharge positioning of transformer high and low voltage windings.It has better effect.The internal structure of the transformer is complicated.The partial discharge ultrasonic signal received by the sensor is subjected to multiple reflections.In order to reduce the influence of the reverberation signal for the partial discharge location,a minimum phase decomposition de-reverberation algorithm based on complex cepstrum is proposed.Based on the simulation of partial discharge localization,the comparison results of MUSIC-SRP positioning error before and after de-reverberation.It is proved that the minimum phase decomposition complex cepstrum de-reverberation can improve the positioning accuracy.The experimental results also confirm the effectiveness of the algorithm.