An improved algorithm for the estimation of the root mean square value as an optimal solution for commercial measurement equipment

• Published in: Microprocessors and microsystems 2024-04, Vol.106, p.105042, Article 105042
• Main Authors: Bulat, Marina, Mirković, Stefan, Gazivoda, Nemanja, Pejić, Dragan, Urekar, Marjan, Antić, Boris
• Format: Article
• Language: English
• Subjects: Approximation methods; Interpolation; Numerical analysis; Root mean square
• ISSN: 0141-9331; 1872-9436
• DOI: 10.1016/j.micpro.2024.105042

•Numerical methods used for the estimation of the RMS.•Simpson's 1/3 rule and Simpson's 3/8 rule modified for the purpose of general application.•Modified Simpson's rules do not necessitate more complex mathematical calculations than those used in the existing methods.•Modifications provide better measurement results for some a lower ratio of the frequency of sampling and the frequency of the signal.•Modified methods do not require the manufacturers of commercial measurement equipment to additionally invest in it. This paper demonstrates that direct changes in the algorithm for the estimation of the root mean square value of a voltage signal of an arbitrary waveform can lead to improved performances and lower measurement uncertainty of commercially available instruments without requiring any upgrade of their existing hardware. The research conducted and presented here is an original contribution to the development of estimation techniques and mathematical models for measurement oriented purposes regardless of the number of samples in the given period relying on mathematical calculation of the equal complexity as in the methods already in use. The theoretical approach examines the problem of numerical integration focusing on modified Simpson's 1/3 rule and modified Simpson's 3/8 rule used for the purpose of the estimation of the root mean square value when a small number of samples per period is available. It highlights the limitations of Simpson's 1/3 rule and Simpson's 3/8 rule, and shows that the newly proposed algorithm is optimal with respect to measurement accuracy and precision even in cases when the ratio of the sampling frequency and the signal's fundamental frequency is low. All theoretical results have been validated experimentally.