The aim of this work is to estimate the flashover voltage of a high-voltage insulator that has been deliberately polluted using fuzzy logic (FL). Initially, experimental experiments on a high-voltage insulator were used to collect a data set that was then used to implement the idea of artificial intelligence. These studies were conducted using varying degrees of fake pollution, namely saline distilled water. Each pollution level indicated the quantity of artificial pollution, measured in millilitres, in each section of the insulator. The collecting database provides flashover voltage measurements associated with varying levels of artificial pollution in each insulator zone and its conductivity. Furthermore, we have used fuzzy logic (FL) to forecast the flashover voltage of the high-voltage insulator and assess the insulating condition of simulated pollution. The suggested prediction model, which is based on Federated Learning (FL), is implemented using MATLAB's graphical user interface. Ultimately, a comparison was conducted between the outcomes achieved by FL and real-world ones. The database used in this comparison differs from that used in concepts based on programming language implementation taken from previous literature. The findings demonstrate the superior effectiveness of the FL approach in predicting the flashover voltage of high-voltage insulators when compared to data acquired from practical testing.
PL
Celem pracy jest oszacowanie napięcia przeskoku izolatora wysokiego napięcia, który został celowo zanieczyszczony przy użyciu logiki rozmytej (FL). Początkowo eksperymenty eksperymentalne na izolatorze wysokiego napięcia służyły zebraniu zbioru danych, który następnie wykorzystano do wdrożenia idei sztucznej inteligencji. Badania te przeprowadzono przy użyciu różnego stopnia fałszywych zanieczyszczeń, a mianowicie destylowanej wody solankowej. Każdy poziom zanieczyszczenia wskazywał ilość sztucznych zanieczyszczeń, mierzoną w mililitrach, w każdej sekcji izolatora. Zbierana baza danych zapewnia pomiary napięcia przeskoku związane z różnymi poziomami sztucznych zanieczyszczeń w każdej strefie izolatora i jego przewodnością. Ponadto wykorzystaliśmy logikę rozmytą (FL) do prognozowania napięcia przeskoku izolatora wysokiego napięcia i oceny stanu izolacji symulowanego zanieczyszczenia. Sugerowany model predykcyjny, oparty na Federated Learning (FL), jest implementowany przy użyciu graficznego interfejsu użytkownika MATLAB-a. Ostatecznie przeprowadzono porównanie wyników uzyskanych na platformie FL z wynikami uzyskanymi w świecie rzeczywistym. Baza danych wykorzystana w tym porównaniu różni się od bazy danych stosowanej w koncepcjach opartych na implementacji języka programowania zaczerpniętych z wcześniejszej literatury. Odkrycia wskazują na wyższą skuteczność metody FL w przewidywaniu napięcia przeskoku izolatorów wysokiego napięcia w porównaniu z danymi uzyskanymi z testów praktycznych.
The main insulation layer is the most important layer of the high-voltage cable, and the quality of this material directly affects the life of the cable. It is also known that contamination, porosity and associated partial discharges in the insulation can affect the service life of cables. In this paper, we use the COMSOL Multiphysics software, which is based on the finite element method in AC/DC, 2D electrostatic. Our study shows the effect of heterogeneous cavities on the functioning of electrical cables. This work contains the study of electric field distribution and potential of a model of high voltage cable; we took into account the absence and the presence of heterogeneous cavities. The study was conducted using numerical results with mathematical validation. The obtained results are considered satisfactory, favorable and very promising.
Investigations of the flow in a building system are crucial for understanding the fundamental basis of the aerodynamic structure characteristics. The CFD simulations were conducted using ANSYS Fluent 17.0 software, which solves the Navier-Stokes equations in conjunction with different turbulence models and by a finite volume discretization method. Particularly, a comparison between the experimental and standard k-ω, BSL k-ω, SST k-ω, standard k-ɛ, RNG k-ɛ and Realizable k-ɛ turbulence model has been developed. The comparison between the founded results affirms that the standard k-ω turbulence model is the most efficient to model the air flow in the present application. Indeed, the numerical results compared using the experimental data developed in the LASEM laboratory confirms the validity of the numerical method. The good agreements validate the considered computational method.
The flashover of pollution, observed on the insulators used in high voltage transmission, is one of the most important power transmission stakes. It is a very complex problem due to several factors including the modelling difficulties of complex shapes of insulators, different pollution densities at different regions, non-homogeneous pollution distribution on the insulator surface and unknown effect of humidity on the pollution. In the literature, some static and dynamic models have been developed by making some assumptions and omissions to predict the flashover voltages of polluted insulators. This paper aims to experimentally analyse the flashover process and simulation of the distributions of the potential and the electric field under 50 Hz applied voltage on a real model simulating the 175CTV outdoor insulators largely used by the Algerian Company of Electricity and Gas (SONELGAZ). This real model is studied under non-polluted (distilled water), and polluted (distilled water and sand) environments. The simulations were carried out by using the COMSOL multiphysics software. This program uses the finite element method to solve the partial differential equations that describe the field. Experimental results made in the laboratory and simulation results are original and found to be congruent.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.