Cloud-integrated battery management system using IoT and machine learning for solar photovoltaic applications

Batteries are essential energy storage devices that enhance the reliability and efficiency of renewable energy systems. The Battery Management System (BMS) is necessary for ensuring the battery’s consistency, safety, performance improvement, and efficiency. It is able to discriminate the discharging and charging current, to provide the warning information very early and manage the batteries connected economically. In case of large scale BMS; it needs complex wiring setup, expensive hardware, air-conditioners and regular maintenance with man-power. To address these problems, the Cloud Integrated Battery Management System (CIBMS) has been proposed to monitor battery characteristics continually and that the proposed controller use sophisticated computational techniques to predict the battery’s State of Charge (SOC) and State of Health (SOH). It stores the bulk amount of measured data to the Amazon Web Services (AWS) 1 GB RAM and 40 GB Space cloud. The proposed system regulates the battery charging from solar PV and the maximum discharge rate. The hardware setup has been implemented in the institutional laboratory and tested for solar powered lead acid batteries. The sensors connected to Internet of Things (IoT) devices continuously collect and store real-time battery statistics in cloud databases. The robust machine learning method of the Support Vector Regression (SVR) technique has been used to examine these data to estimate the SOH with the values of 99.14% to 79.87% corresponding to 100% − 80% of actual SOH and SOC with the values of 99.61% to 11.29% corresponding to 100% − 10% of actual SOC. Additionally, it anticipates the battery’s age, fault, and maintenance needs and best manages the battery’s charging and discharging limits. The proposed CIBMS has been executed using various algorithms and the results are validated with the actual measurement.