Accurate fault location is crucial for finding and clearing faults in distribution networks. It can help to reduce the loss of power failure and ensure safe and stable operation. Without efficient observation points and the lack of phase angle information, traditional fault location methods of distribution networks usually have large errors. As the application of μPMUs in the distribution network becomes more and more common, voltage and current waveforms, as well as phase angle can be recorded and provide more information for accurate fault location. In this paper, an intelligent location method is proposed to pinpoint the fault location based on the information of μPMUs which are properly allocated. The fault section is firstly determined by comparing the zero-sequence current waveforms on both sides of the fault section. Then, a Stack Auto-Encoder (SAE) is modeled to provide an end-to-end solution to pinpoint fault point with the voltage and current phasors. Finally, the performance of the proposed method is tested by a simulated distribution network on the platform of PSCAD. The results show that the proposed method is effective in locating faults and can withstand the effects of transition resistance, fault type, and noise. Compared to another popular method, the proposed method shows better location accuracy.

This paper presents an end-to-end approach for locating faults on high-voltage dc (HVDC) transmission lines. Different from traditional methods which rely on communications between different measuring units or feature extraction of post fault transients, the proposed algorithm takes the raw data of locally detected traveling current surges as the only-input and outputs the fault locations directly. Especially, the stacked auto-encoder (SAE) is utilized to model the relationship between fault currents and fault locations. The SAE-based method is performed in time domain and tested with a simulated HVDC transmission line modeled in PSCAD/EMTDC. The simulation results show that this method is effective in locating faulted points and robust against attenuation, overlapping of traveling surges, and various ground resistances.

Partial discharge (PD) is caused by the localised electrical field intensification in insulating materials. Early detection and accurate measurement of PD are very important for preventing premature failure of the insulating material. Detection of PDs in metal‐clad apparatus through the transient Earth voltage method is a promising approach in non‐intrusive on‐line tests. However, the electrical interference from background environment remains the major barrier to improving its measurement accuracy. In this study, a wavelet‐entropy‐based PD de‐noising method has been proposed. The unique features of PD are characterised by combining wavelet analysis that reveals the local features and entropy that measures the disorder. With such features, a feed‐forward back‐propagation artificial neural network is adopted to recognise the actual PDs from noisy background. Comparing with other methods such as the energy‐based method and the similarity‐comparing method, the proposed wavelet‐entropy‐based method is more effective in PD signal de‐noising.

Distribution networks have complex topologies and limited number of monitoring devices. Because quick and accurate fault location is critical for the operation quality and reliability of distribution systems, a lot of methods have been proposed, including installation of additional monitoring equipment. Although accurate fault location is possible by dividing the complex topologies into simple ones with fewer branches, the cost of installing additional devices is uneconomic, which makes it hard to be widely applied in practical systems. Based on the theoretical analysis of the relationship between electrical measurements and fault points, an intelligent location method with limited measurement information for multi-branch distribution networks is proposed. Logical reasoning and machine learning are combined to locate fault points with limited measurement information. The relationship between the fault point and unknown distribution of uncontrollable factors in fault equivalents are fitted by machine learning model which can preserve time-related features of electrical waveforms. An adaptive migration strategy by transfer learning is proposed to deal with the problem of small samples in practical application. The proposed method is tested in different topologies. Simulation results prove the effectiveness and generalization capability of the proposed method under complex operating conditions.

Compared with AC distribution networks, DC ones have a number of advantages. Intensive connections of distributed renewable energy can lead to large amount of power electronic converters and complex models. Underground cable is widely used in DC distribution networks. Accurate location of faults can help engineers find the fault points and shorten the time of maintenance. In DC distribution networks, where only a few measuring units are equipped and low sampling rates are adopted, there is limited data used for fault location. Also, for monopole grounding fault, the fault features are sometimes unobvious for recognition. Deep learning which provides feature hierarchy can learn experiences automatically and recognise raw data as human brain does. It reveals a high potential to solve location problems in DC distribution systems. This paper proposes a depth learning based fault location for DC distribution networks. First, a DC distribution network with radiant topology is modelled, and faults are added with different parameters to simulate various scenarios in practical projects. Then, a deep neural network is generated and trained with normalised fault currents. The parameters of network are discussed according to particular application. Finally, the location performance of deep neural network is tested.