Lijing Zhang

Existing interturn fault detecting methods rely on winding impedance, winding current, and dissolved gases. They are effective only when the insulation is severely damaged. This paper proposes a novel detection method based on fusion analysis of electrothermal characteristics including winding currents, temperatures of four areas on the tank wall, top oil and ambient temperatures, which can identify the interturn fault at an early stage. When an incipient interturn fault occurs, the heat generated by the faulty turns is transferred to the oil and tank wall, leading to an increase in top oil and tank wall temperatures. Thus, the incipient fault can be detected by analysing these electrothermal characteristic parameters. Borrowing the idea of digital twin (DT), this method establishes a high‐fidelity simulation model to simulate the transformer electrothermal characteristics under different operating conditions. Afterward, an intelligent neural network is adopted to extract the quantitative relationship between the eight feature attributions and fault conditions. Finally, this neural network is utilized to detect the incipient interturn fault for the transformer entity. Case studies are conducted on a 100 kVA transformer with oil natural air natural (ONAN) cooling mode. The detection accuracy is improved by 68.5% compared to the winding current‐based method.

Accurate assessment of hot‐spot temperature is essential for the safe operation of power transformers. Existing dynamic thermal models cannot estimate hot‐spot temperature accurately since some input parameters are roughly determined by transformer capacity and cooling mode while ignoring the effect of winding structure, tank dimensions, and material physical properties. To improve the accuracy of temperature assessment, empirical parameters of the IEC thermal model including thermal constants, winding and oil exponents are optimised with the help of numerical simulation in this article. Based on energy conservation and heat transfer theory, a computational fluid dynamic (CFD) model of a transformer in oil natural air natural (ONAN) cooling mode is established. This CFD model simulates the entity's structure, sizes, and multi‐stage heat dissipation processes realistically, so it can more precisely calculate the dynamic hot‐spot temperature. According to the simulated temperature curves at different operating conditions, the thermal constants and oil exponent are estimated using non‐linear regression, and the winding exponent is optimised using linear regression. A case study is conducted on an ONAN transformer. It shows the improved IEC model with optimised parameters can more accurately evaluate hot‐spot temperature, and the absolute error is decreased by 2.4 K (38.7%) compared with traditional thermal models.

Temperature is an important factor affecting the dynamic processes of corona discharge, since the ionization, attachment, and mobility of charged particles are directly related to temperature. Most existing research has investigated the temperature effect on corona discharge through experimental works. However, they are hard to acquire the corona discharge microphysical processes, which are related to the formation of the current pulse and electromagnetic (EM) wave signals. This article proposes a numerical model with temperature-related transport parameters to study the temperature influence on microphysical processes and current pulses for negative corona. By comparing microphysical quantities such as the effective ionization coefficient, density, and velocity of space charges, the effect of temperature on the current pulse parameters including the rise and fall times, amplitude, and pulsewidth is systematically explained. It shows that as the temperature rises, the current pulse amplitude increases, while the other three parameters decrease. The rise time is dependent on the effective ionization coefficient. The fall time is affected by the velocities of negative and positive ions. The experimental results indicate the simulated current pulses are consistent with the measured ones, proving the effectiveness of the proposed numerical model with temperature variation.

The tip size of metal protrusions is an important factor affecting corona discharge, since it influences the electric filed, the ionization range and strength around the protrusion defect. Most existing studies on the influence of protrusion tip size on corona discharge are implemented by experimental methods. However, these studies are difficult to obtain the microphysical process of charged particles, which is related to the generation of current pulses and electromagnetic (EM) waves. In this paper, a hydrodynamic simulation model is proposed to study the effect of protrusion tip size on the microphysical process and current pulses of negative corona discharge. A mapping relationship between microscopic behaviors of charged particles and current pulses is established by this simulation method. Based on the comparison of microphysical quantities, the effect of defect tip size on the detailed parameters of current pulses is clearly explained. It shows that with the increase of tip size under a constant voltage, all the five parameters of current pluses increase. The rising time is related to the effective ionization coefficient, while the falling time is dependent on the velocities of positive and negative ions. The experiment and simulation results coincide in current waveforms and all five parameters.

Dissolved gas analysis is an important technique for the insulation condition assessment and incipient fault diagnosis of power transformers. However, the performance of the traditional ratio methods can be hardly improved due to the overreliance on absolute ratio threshold. In this paper, a novel method combining oversampling and cost‐sensitive learning is proposed to improve the diagnosis accuracy of all fault types of transformers. The radial‐based oversampling (RBO) method is adopted to synthesise samples for the complex fault classes. With the newly generated samples, the deep belief network (DBN) can effectively learn the features of complex fault classes and distinguish them from the other fault classes. Moreover, by integrating a cost matrix into the loss function, the parameters of DBN are adaptively updated so as to ensure the correct classification of the fault class with less samples. Based on the oversampling and cost‐sensitive learning, the proposed method can form suitable classification boundaries amongst thermal, discharge and complex fault classes. The effectiveness and generalisation capability of the proposed method are verified by case studies in a real‐world fault dataset of power transformers with multi‐source samples. The results demonstrate that the proposed method improves the classification accuracies in all fault classes, especially in the complex fault classes. The overall accuracy can be reach over 90% by applying both RBO and cost‐sensitive learning.

The function of the fluxgate sensor in the aircraft navigation system is to calibrate the heading of gyro. Its magnetic heading signal is weak and anomalous synchro signal. Experimental results show that this signal cannot be measured directly and exactly by using SDC. This paper designs a negative feedback amplifying circuit which uses the servo amplifying setup and synchronizer equipment to convert the magnetic heading signal to the standard synchro signal. Meanwhile, we study and manufacture the SDC and DSC modules which integrate the multifunction based-on PC-104 bus synchro/digital converting card. In the practical operation, we adopt the above-mentioned scheme and calibrate the compass error by the software. The result demonstrates that the acquisition accuracy of magnetic heading signal reaches 0.1deg and meets the demands of the navigation accuracy.