Study of Discharge in Point-Plane Air Interval Using Fuzzy Logic

Abstract

-The objective of this paper is to study the discharge phenomenon for a point-plane air interval using an original fuzzy logic system. Firstly, a physical model based on streamer theory with consideration of the space charge fields due to electrons and positive ions is proposed. To test this model we have calculated the breakdown threshold voltage for a point-plane air interval. The same model is used to determine the discharge steps for different configurations as an inference data base. Secondly, using results obtained by the numerical simulation of the previous model, we have introduced the fuzzy logic technique to predict the breakdown threshold voltage of the same configurations used in the numerical model and make estimation on the insulating state of the air interval. From the comparison of obtained results, we can conclude that they are in accordance with the experimental ones obtained for breakdown discharges in different point-plane air gaps collected from the literature. The proposed study using fuzzy logic technique shows a good performance in the analysis of different discharge steps of the air interval.

A Feature Set for Structural Characterization of Sphere Gaps and the Breakdown Voltage Prediction by PSO-Optimized Support Vector Classifier

“…On the other hand, some artificial intelligence algorithms are useful to describe the relations between the influencing factors and the breakdown voltage. For example, the artificial neural network (ANN) [11], [12], the support vector machine (SVM) [13]- [15] and the fuzzy logic [16], [17] have already been used by some researchers to predict the external insulation strength. These achievements preliminarily verify the feasibility of air insulation prediction by some mathematical algorithms, but many interesting topics are still worthy to be studied in detail.…”

Discharge voltage prediction of UHV AC transmission line–tower air gaps by a machine learning model

“…For the past few years, some artificial intelligence algorithms, e.g. artificial neural network (ANN), fuzzy logic, and support vector machine (SVM), have been introduced to predict the strength of external insulation, including the flashover voltage of insulators [16–18] and the breakdown voltage of air gaps [19–24]. These researches provide new ideas to predict the dielectric strength of air insulation by modelling the relationships between the flashover voltage and the related influencing factors directly, instead of considering the complex and random discharge process.…”

“…Based on the fact that the electric field distribution has a decisive influence on the value of the gap factor and the air gap discharge voltage, if their relationships can be established by a mathematical model, it is possible to predict the strength of air insulation. Previous studies in [16–20] extracted some features from the static electric field distribution calculated by finite element method (FEM) to characterise the gap structure, and establish their multi‐dimensional nonlinear relationships with the breakdown voltage by a machine learning algorithm, namely, the SVM. This method has satisfactory performance for breakdown voltage prediction of typical air gaps, such as the sphere gap, rod‐plane, rod‐rod gaps etc.…”

Electrostatic field features on the shortest interelectrode path and a SVR model for breakdown voltage prediction of rod–plane air gaps

“…The above idea considers the discharge process as a grey box; therefore, some artificial intelligence algorithms may be suitable to describe this problem. Intelligent algorithms such as the artificial neural network [14,15], the fuzzy logic system [16,17], the support vector machine (SVM) [18][19][20] etc. have been applied to analyse air-gap breakdown process or predict the breakdown voltage in previous studies.…”