A Memory-Efficient True-RMS Estimator in a Limited-Resources Hardware

Abstract

More and more human activities are increasingly dependent on the power quality energy they are supplied. In particular, those loads that have an electronic switching mode power supply (SMPS) and microcontrollers or microprocessors are very susceptible to de-energization of the AC line as voltage monitoring takes place on the DC end and may not have enough time to cope with a voltage event involving energy loss. There have been many proposals that analyze power quality or current consumption, even those using limited resources hardware or the classic formulas of discrete root mean square (RMS). In this proposal, an approximation to the problem by means of an RMS value estimator that uses as a base a microcontroller of basic range and low economic cost and algorithms of calculation of very low computational cost that elude complex arithmetic operation to controllers, such as powers or roots, is presented. The results of the experiments to which the proposal was subjected show its ability to provide an estimation of the RMS value of voltage with sufficient precision or an event alarm in less time than the options analyzed. The validation tests and functional comparison of the prototype which were carried out support its viability.