基于多传感器融合的电力变压器内部放电定位与辨识技术

针对电力设备状态监测中传感器信息融合的难题，应用油中溶解气体、特高频、光纤超声等监测传感器构建了适用于变压器的融合传感阵列，提出了变压器内部放电的辨识策略。采用超声信号到达时间差估算实现了放电源定位，建立了多传感器监测特征融合算法，并结合K邻近（K nearest neighbors，KNN）算法实现了放电模式辨识。实验结果表明：放电源定位平均误差为45.1 mm，放电类型辨识准确率达到90.6%；所构建的多传感器融合感知策略与系统，可实现变压器内部放电源的准确定位与辨识，具有一定的应用推广价值。

Internal Discharge Location and Identification for Power Transformers Based on Multi-Sensor Fusion

To solve the problem of senor information fusion in power equipment condition monitoring， a sensor fusion array is constructed with several monitoring sensors for power transformers， i.e.， dissolved gas in oil， UHF and fiber optic ultrasound， and the identification strategy of transformer internal discharge is proposed in this paper. Firstly， the arrival time estimation is used to realize the discharge location by acoustic signal. Secondly， a novel feature fusion algorithm based on the multi-sensor is established， and the KNN （K nearest neighbor） algorithm is used to identify internal discharge in transformers. The experimental results show that the average error of discharge location is 45.1 mm， and the identification accuracy of discharge type reaches 90.6%. The multi-sensor fusion sensing strategy and system constructed in this paper could realize the accurate discharge location and pattern recognition in transformers with good application and promotion value.