A Biased-SECE Interface for Piezoelectric Energy Harvesting with Geometric-Mean-Computational MPPT Achieving 99.9% MPPT Efficiency, $8.75\text{Cycles}/\Delta \mathrm{V}_{\text{OC}}$ Tracking, and 9.3x Energy Extraction

As the demand for sustainable energy solutions grows, piezoelectric energy harvesting (PEH)-which converts ambient vibrations into electrical energy-has emerged [1]–[6]. The interface circuit in a PEH system plays a crucial role not only in energy extraction and conversion but also in enabling the piezoelectric (PZT) to operate near its maximum power point (MPP). Enhancing the efficiency of PEH systems requires careful attention to both static performance (MPP accuracy) and dynamic performance (MPP tracking responsiveness). Rapid online tracking of the MPP is particularly critical, as it allows PEH to adapt seamlessly to the time-varying open-circuit voltage$(V_{\text{OC}})$of the PZT, as well as to different types of input vibrations, whether periodic or shock-induced [7]. Figure 31.2.1 (top-left) shows previous approaches. The MPPT scheme based on fractional open-circuit voltage (FOCV) [1] directly samples$V_{\text{OC}}$from the input, leading to interruptions in energy extraction. Furthermore, to maintain tight MPPT under rapidly fluctuating$V_{\text{OC}}$conditions, more frequent$V_{\text{OC}}$-sampling is necessary, reducing the amount of energy harvested. In contrast, the perturb-and-observe (P&O) [2] and duty-cycle-based (DCB) [3] MPPT methods sense the harvested output to avoid discontinuities in energy extraction. However, the intermediate$V_{\text{RECT}}$and$C_{\text{RECT}}$affect both MPPT speed and accuracy, resulting in a trade-off between these two aspects.