Output Voltage Estimation for Vibration Energy Harvesters Using Relevance Vector Mechines

By converting natural vibration into electricity, vibration energy harvesters (VEHs) provide promising self-power resources to drive wireless electronic systems. The output voltage is a vital parameter to evaluate the performance of a VEH. Considering the fact that relation between the operation conditions and the output voltage is nonlinear, a relevance vector machine (RVM) -based approach is proposed to estimate the output voltage. The vibration frequency, acceleration amplitude, and load resistance are employed as the input vectors so as to model the harvested electricity voltage which is the output vector of the RVM. An experimental set-up is used to collect real data for RVM modeling and precision evaluation. The results show that the proposed method is capable of estimating the output voltage for the VEH with good precision.