Co-design of adaptive-hysteresis solar MPPT and energy-neutral duty scheduling for ultracapacitor-powered wireless sensor nodes

Solar-powered wireless sensor nodes are constrained by both harvesting-stage losses and temporal mismatch between harvested energy and node service demand. This paper develops a co-designed harvesting and energy-management framework that addresses these limitations jointly. A closed-form expression for the polarization efficiency of a hysteretic maximum power point tracking stage is derived for an ideal single-diode photovoltaic model, accounting simultaneously for ripple loss and set-point error. The formulation converges to the conventional second-order Taylor approximation in the narrow-band limit and remains within 0.09% points of a full single-diode model, including series and shunt resistances over 50–1000 W/m2. A loss-aware hysteresis band is obtained by minimizing ripple, switching, and resonant-conduction losses, while the tracking set point is determined analytically using a Lambert-W-based fractional open-circuit-voltage relation. A predictive scheduler allocates only stored energy exceeding a forecast-based reserve, for which conditional energy neutrality is established under a stated forecast-conservatism assumption. For a 25 mW PV module, 30 F ultracapacitor, and 3.3 V mote, seven-day mixed-weather simulations yield 0.933 mean efficiency vs 0.898 for a fixed-band reference, double the service events, and eliminate service interruptions. Ablation analysis further distinguishes harvesting-stage improvements from system-level recovery of otherwise curtailed energy.