Use of artificial neural networks to assess the impact of electromagnetic interference

Abstract

Statement of a problem. In the process of testing electronic devices for the effects of electromagnetic interference, it is necessary to process large amounts of data. Such data includes information about the operation of the device under the influence of electromagnetic interference (EMI), configuration (EMI), polarization and many other parameters that affect the assessment of the quality of electromagnetic compatibility (EMC). Based on the fact that most modern devices operate at high and ultra-high frequencies, the amount of data coming from them can reach tens of gigabytes. Analytical processing of such data arrays can be difficult, and in some cases impossible at all. An alternative option for processing such arrays is the use of systems based on artificial intelligence and neural networks of various configurations. Purpose. Conducting an effective assessment of the behavior of the device when it is exposed to electromagnetic interference. Results. A method of using an artificial intelligence system for processing experimental data when testing electronic devices for EMF exposure has been developed. Practical importance. The developed method is aimed at improving the efficiency of the hardware and software complex for assessing the effect of electromagnetic interference on electronic means. Improving the quality of data processing will make it possible to more accurately predict the behavior of the device when working in a real operating environment, as well as to point out to the developer potentially problematic places in the design of the electronic means.