Automation of electric power source changeover switches deploying artificial intelligence.

Abstract

Unstable, unreliable, and expensive electric power supply systems have created a need for multiple alternative power sources to manage power failures and cost. Manually controlled changeover switches are limited by human cognitive ability and availability of personnel to operate the device, making them inefficient and expensive to maintain. This paper proposes an automated electric power changeover switch that deploys artificial intelligence to process real-time generated data such as solar availability, grid stability, and fuel level in fossil fuel power generators, using forecasting and optimization techniques to accurately select preferred shared active power sources during operation. After each power changeover, the device sends a short message service (SMS) to enrolled users on the electric power status and also allows the remote control of the device. The obtained trial test results validated its operational efficiency. The developed device is targeted at low-income countries of Sub-Saharan Africa where electric supply is highly unstable.