Indoor environment PV applications: Estimation of the maximum harvestable power

Abstract

The use of PV modules for powering sensors in an indoor environment requires that, during the design process, the harvestable power be evaluated and compared with the power requirements of the load device to validate their compatibility for a given type of light source and level of illumination. The models reported recently to relate power to illumination during power estimation are seen to result in large estimation errors in the case of an indoor environment. This paper seeks to propose an appropriate model that can be exploited during maximum power estimation in an indoor environment to relate power to illumination level. Proceeding from fitting measured data to the three-diode model of the PV module and based on the errors and execution time, the Grey Wolf Optimisation Algorithm, among others, proved to be accurate for estimating maximum power. The relationship between maximum power and illumination was seen to take the form of a second-degree polynomial. This was validated by comparing the values of power estimated by this model to the measured values as well as values estimated with the linear model, the exponential model, and the model proposed by Joseph Amajama. The proposed model proved to be more accurate than the models reported, recording errors of not more than 5 %, which is acceptable for system design. This will go a long way towards facilitating the design of indoor energy harvesting systems that can be used for powering smart sensors and other IoT devices.