An Empirical Investigation in Applying Reliable Industry 4.0 Based Machine Learning (ML) Approaches in Analysing and Monitoring Smart Meters using Multivariate Analysis of Variance (Manova)
Md. Sohel Rana, Anil Kumar Dixit, M. Sundar Rajan, Sumit Malhotra, S. Radhika, Bhasker Pant
Abstract
Conventional power systems have been evolving towards prospective smart grids in recent decades as market needs have changed and independent innovations, such as the “Internet of Things” (IoT), have developed. The further automation of traditional development of industrial operations with a variety of current advanced techniques is termed as Industry 4.0. The most efficient strategy to boost the productivity of the enterprise environment, such as smart grids, is to use contemporary control technology that regulates and supervises interaction between smart devices. Furthermore, the advancement of loT -enabled smart meters is aided by a separate data connection, such as “Bluetooth” or the “Global System for Mobiles” (GSM), for remotely inspecting and operating smart meters. In this context, SPSS software is used for analysing the role of Industry 4.0 based Machine Learning or ML in analysing Smart Meters and MANOV A (and correlation) approach is considered to get proper result. This study proposes a machine learning-based architecture for analysing and monitoring smart meter data. The advantages of loT integration into Industry 4.0 include the ability to exchange actual information across a wide range of smart meters for various applications, such as factories, hospitals and residences, which is critical for smart grids. Results suggested that Smart Grid is beneficial to the extent of reducing electricity cost; however, maintenance cost and security issues are the major challenging factors in the new technology.