Particle swarm optimization of a neural network model for predicting the flashover voltage on polluted cap and pin insulator

Abstract

This paper proposes training an artificial neural network (ANN) by a particle swarm optimization (PSO) technique to predict the flashover voltage of outdoor insulators. The analysis follows a series of real-world tests on high-voltage insulators to form a database for implementing artificial intelligence concepts. These tests are performed in various degrees of artificial contamination (distilled brine). Each contamination level shows the amount of contamination in milliliters per area of the isolator. The acquisition database provides values of flashover voltage corresponding to their electrical conductivity in each isolation zone and different degrees of artificial contamination. The results show that ANN trained by PSO can not only provide better prediction results, but also reduce the amount of computation efforts. It is also a more powerful model because: it does not get stuck in a local optimum. In addition, it also has the advantages of simple logic, simple implementation, and underlying intelligence. Compared to the results obtained by practical tests, the results obtained present that the PSO-ANN technique is very effective to predict flashover of high-voltage polluted insulators.

Modeling and Predicting the Mechanical Behavior of Standard Insulating Kraft Paper Used in Power Transformers under Thermal Aging

“…Position, velocity, and fitness are three significant factors that are critical. The PSO is used to solve an optimization problem by performing the following [21]: a.…”

“…The PSO's advantages include simple coding and low processing cost [21]. Since the PSO algorithm solves optimal glob problems with accuracy, it is used in the current work to train the multi-layer perceptron (MLP) network models.…”