Prediction of the Magnitude of Lightning Impulse Interference in an Electronic Device Using an Artificial Neural Network

Reliable operation of modern electronic devices is the basis for their effective use. Electromagnetic interference, such as lightning, creates the risk of temporary or permanent disruption of the functioning of device elements. Therefore, it is necessary to take into account the possible consequences of such interference on the basis of prediction even at the design stage of electronic devices. The paper a technique for predicting the magnitude of lightning impulse interference in electronic devices using an artificial neural network is proposed. The advantage of this approach over simulation modeling is that rather complex three-dimensional models and long computation times are not needed. A stand was developed and obtained by measuring the amount of lightning interference in an electronic device. The features of the task of training an artificial neural network based on experimental data are considered. The architecture and parameters of the artificial neural network have been selected. An artificial neural network was trained based on experimental data. The root mean square error between the predicted data and the average impulse interference voltage values on the training data does not exceed acceptable values. The results of predicting the magnitude of lightning interference in an electronic device are presented. If we consider further prospects for using an artificial neural network to predict electromagnetic interference, we can note the relevance of the problem of the impact of electrostatic discharge, intentional short pulse and etc.