It is a great challenge to differentiate Partial Discharge (PD) induced by different types of insulation defects in high voltage cables. Some types of PD signals have very similar characteristics and are specifically difficult to be differentiate, even for the most experienced specialists. To overcome the challenge, a Convolutional Neural Network (CNN) based deep learning methodology for PD pattern recognition is presented in this paper. Firstly, PD testing for five types of artificial defects in Ethylene-Propylene-Rubber (EPR) cables was carried out in the High Voltage (HV) laboratory to generate signals containing PD data. Secondly, 3500 sets of PD transient pulses were extracted and then 33 kinds of PD features were established. The third stage applies a CNN to the data: typical CNN architecture and the key factors which affect the CNN based pattern recognition accuracy, are described. Factors discussed include the number of the network layers, convolutional kernel size, activation function, and pooling method. The paper presents a flowchart of the CNN based PD pattern recognition method and an evaluation with 3500 sets of PD samples. Finally, the CNN based pattern recognition results are shown and the proposed method is compared with two more traditional analysis methods, i.e. Support Vector Machine (SVM) and Back Propagation Neural Network (BPNN). The results show that the proposed CNN method has higher pattern recognition accuracy than SVM and BPNN, and that the novel method is especially effective for PD type recognition in cases of signals of high similarity, which is applicable for industrial applications.

Optimal selection of features of Partial Discharge (PD) signals recorded from defects in High Voltage (HV) cables will contribute not only to the improvement of PD pattern recognition accuracy and efficiency but also to PD parameter visualization in HV cable condition monitoring and diagnostics. This paper presents a novel Random Forest (RF)-based feature selection algorithm for PD pattern recognition of HV cables. Firstly, the experimental data acquisition and feature extraction processes are introduced. PD signals were captured from faults created in a cable to obtain the raw PD data, then a set of 3500 transient PD pulses and a set of 3500 typical interference pulses were extracted, based on which 34 PD features were extracted for further processing. Furthermore, 119 two-dimensional features and 1082 three-dimensional features were generated. The paper then discusses the basic principle of the RF algorithm. Finally, RF-based feature selection was implemented to determine the optimal features for PD pattern recognition. The results were obtained and evaluated with the Back Propagation Neural Network (BPNN) and Support Vector Machine (SVM). Results show that the proposed RF-based method is effective for PD feature selection of HV cables with the potential for application to additional HV power apparatus.

In this study, glass fiber (GF)/phenol formaldehyde resin (PF)/epoxy resin (EP) three-phase electrical insulating composites were fabricated by selective laser sintering (SLS) additive manufacturing technology and subsequent infiltration. In the three-phase composites, glass fibers modified by a silane coupling agent (KH-550) were used as reinforcements, phenol formaldehyde resin acted as the binder and matrix, and infiltrated epoxy resin was the filler. Mechanical and electrical properties such as tensile strength, bending strength, dielectric constant, electrical conductivity, and electric breakdown strength of the GF/PF/EP three-phase composite parts were investigated. The results indicated that after being infiltrated with EP, the bending strength and tensile strength of the GF/PF/EP composites increased by 30% and 42.8%, respectively. Moreover, the flexural strength and tensile strength of the GF/PF/EP composite increased with the increase of the glass fiber content. More importantly, the three-phase composites showed high electrical properties. Significant improvement in the dielectric constant, electric breakdown strength, and resistivity with the increase in the content of glass fiber was observed. This enables the prepared GF/PF/EP composites to form complex structural electrical insulation devices by SLS, which expands the materials and applications of additive manufacturing technology.

Fault diagnosis of power transformers is crucial to the healthy operation of transformers. In order to enhance its accuracy and reliability, a new fault diagnosis method based on Set Pair Analysis (SPA) and association rules was proposed in this paper. Via analyzing the relationship of fault symptoms and fault types, the corresponding association rules could be established. Via computing the support degrees and confidence degrees of the association rules, the constant weight coefficient of each symptom and the variable weight coefficient of each type could be obtained. The diagnosis method could avoid the subjective defects effectively. Via introducing the concept of subordinate degree, the connection degrees of each fault type and the whole running state of transformer could be obtained. This method could also improve the accuracy of uncertainty factors of transformer fault diagnosis. Experimental results of the test substation proved this method had a higher accuracy by comparing with both association rules and SPA.

Condition assessment for power transformer requires not only integrating the known artificial intelligence (AI) technology, but also exploiting the interrelation of the measured data. According to the association rule of information data and the variable weight synthesizing theory of factor spaces, an assessment method of transformer condition was proposed in this paper. Via analyzing the interrelation of the independent status parameters (ISP) and transformer fault types, the set of synthetic status parameters (SSP) can be built up. For avoiding interference from subjective experience, association rule theory was used to calculate the constant weight coefficients (CWC) of the ISPs. Since the true transformer condition may not always be accurately reflected under the condition of CWCs of a few SSPs, the method of variable weight synthesizing was used for computing the variable weight coefficients (VWC) of the SSPs. Then, combining with the existing maintenance procedures, a preferable condition assessing system of power transformer was proposed. Operational example proved the condition assessing system may reflect the real operation condition of power transformer.