What question did this study set out to answer?

To enhance energy harvesting in residential photovoltaic systems using AI-assisted control and battery storage.

March 26, 2026Results in Engineering0 citationsOpen Access

AI-Augmented Residential PV System with Battery Storage for Enhanced Energy Harvesting

DJD. R. Binu Ben JoseNational Institute of Technology Tiruchirappalli MIM A InayathullahVellore Institute of Technology University VRViza Nagarjuna Reddy

Key Points

To enhance energy harvesting in residential photovoltaic systems using AI-assisted control and battery storage.
Developed a grid-connected and stand-alone PV-BESS residential system
Implemented adaptive maximum power point tracking (MPPT) using a PIC microcontroller
Designed reconfigurable switching logic for power flow management
Utilized hysteresis current control and AI algorithms for efficient operation
Modelled and simulated the system using MATLAB/SIMULINK
Achieved system efficiency of 94% when feeding grid and load
Power tracking efficiency reached 99.4%
Loss of power supply probability at 0.13 indicated high reliability
Experimental results closely matched simulation data

Abstract

• PV-BESS residential system with grid connected and stand-alone operation. • Reconfigurable switching logic for seamless power flow to battery bank, load, and grid. • Adaptive MPPT control and AI-assisted HCC maximizes PV power, ensures grid injection at unity power factor. • System efficiency is 94 % when feeding grid and load, while the power tracking efficiency is 99.4 %. • Loss of power supply probability of 0.13 ensures reliable power. This paper presents a highly reliable residential photovoltaic (PV) fed grid-connected power electronic interface with battery energy storage system (BESS) for enhanced energy harvesting. Being a residential system, a switching logic has been designed which reconfigures the proposed system by connecting the dc–dc converter either with the PV array or with the BESS to operate in grid-connected or stand-alone modes. This logic enables the system to match the grid voltage of 110 V even when the number of batteries in the BESS is less. A PIC16F876 microcontroller (PICuC) has been programmed to implement a maximum power point tracking (MPPT) algorithm, which adjusts the duty ratio of the dc–dc converter using the adaptive step perturb and observe method. To enhance the system-level tracking efficiency, a hysteresis current controller (HCC) assisted by one-class support vector machine (OCSVM) plus XGBoost artificial intelligence (AI) algorithm has been implemented using another PICuC of the same series programmed to control the VSI as well as the power at the dc link to achieve a coordinated control for MPPT. The HCC enables the inverter to feed the actual maximum power to the grid at unity power factor, considering the system efficiency at that level of power. The complete system has been modelled and simulated in the MATLAB/SIMULINK platform. A 500 W laboratory prototype has been fabricated and tested to show the viability of the proposed scheme. Results presented in the experimental section show a close correlation between the experimental and simulation results.

KI fragen

Bookmark

View Full Paper