The winding is the core component of a transformer, and the technology used to diagnose its current state directly affects the operation and maintenance of the transformer. The mechanical vibration characteristics of a dry-type transformer winding are studied in this paper. A short-circuit test was performed on an SCB10-1000/10 dry-type transformer, and the vibration signal at the surface was measured. Based on actual experimental conditions, a vibration-simulation model of the transformer was established using COMSOL Multiphysics software. A multiphysics coupling simulation of the circuit, magnetic field, and solid mechanics of the transformer was performed on this model. The simulation results were compared with measured data to verify the validity of the simulation model. The simulation model for a transformer operating under normal conditions was then used to develop simulation models of transformer-winding looseness, winding deformation, and winding-insulation failure, and the winding fault vibration characteristics were analyzed. The results provide a basis for detecting and analyzing the mechanical state of transformer windings.

In order to improve the intelligent diagnosis level of an on-load tap-changer’s (OLTC) mechanical condition, a feature extraction method based on variational mode decomposition (VMD) and weight divergence was proposed. The harmony search (HS) algorithm was used to optimize the parameter selection of the relevance vector machine (RVM). Firstly, the OLTC vibration signal was decomposed into a series of finite-bandwidth intrinsic mode function (IMF) by VMD under different working conditions. The weight divergence was extracted to characterize the complexity of the vibration signal. Then, weight divergence was used as training and test samples of the harmony search optimization-relevance vector machine (HS-RVM). The experimental results suggested that the proposed integrated model has high fault diagnosis accuracy. This model can accurately extract the characteristics of the mechanical condition, and provide a reference for the practical OLTC intelligent fault diagnosis.

Winding hotspot temperature is the key factor affecting the load capacity and service life of transformers. For the early detection of transformer winding hotspot temperature anomalies, a new prediction model for the hotspot temperature fluctuation range based on fuzzy information granulation (FIG) and the chaotic particle swarm optimized wavelet neural network (CPSO-WNN) is proposed in this paper. The raw data are firstly processed by FIG to extract useful information from each time window. The extracted information is then used to construct a wavelet neural network (WNN) prediction model. Furthermore, the structural parameters of WNN are optimized by chaotic particle swarm optimization (CPSO) before it is used to predict the fluctuation range of the hotspot temperature. By analyzing the experimental data with four different prediction models, we find that the proposed method is more effective and is of guiding significance for the operation and maintenance of transformers.