Harvesting-aware energy management for environmental monitoring WSN

Abstract

An intelligent energy controller is proposed to manage operation of wireless sensor nodes equipped with energy harvesting devices. The energy controller uses Takagi-Sugeno fuzzy logic and has inputs for the state of the energy buffer and forecasts of solar energy available for harvest. Two different forecasting horizons were investigated, current and next-day, using ideal and pressure-based forecasts. Differential evolution is used to optimize the controller. To validate the evolved controller, a wireless sensor network is simulated using real field-collected environmental data. The optimization goal is to best utilize the solar energy available for harvest while preserving a backup energy reserve. Performing the highest number of operations possible while leaving the energy reserve intact increases deployment time and reliability. The controller using current and next-day energy forecasts made better use of the available energy, indicated by a lower fitness function. However, while it took more measurements when compared to the controller only using the current-day forecast, it also used more reserve energy while still remaining at only a small fraction of the total available reserve. Reserve energy usage using the pressure-based forecast was higher for both forecasting horizons compared to the ideal energy forecast, pointing to further performance improvements possible for a more accurate forecast.