Practical vibration energy harvesting tends to be employed where the frequency of the vibration does not vary with time. This is a result of energy harvesting devices resonating at a specific frequency. If the frequency of the vibration changes even slightly from the harvester's natural resonant frequency, then the power generated drops off significantly. Various methods have been proposed to apply kinetic energy harvesting over wider frequency bands, for example by active mechanical tuning of the structure to the vibration, or by adjusting the electrical load. This thesis explores electronic methods of optimally controlling the power electronics that extracts power from the harvester and charges a storage capacitor. The focus is on power levels below lmW, where the power budget for control circuitry is limited to lOs of microwatts or lower. Low-power techniques are investigated that maximise power by optimal resistive loading of the harvester, over a given excitation frequency bandwidth. This includes an experimental investigation of switching-converter control strategies. A low-power power sensing method is presented, which adapts to the available power level to maintain high sensitivity over a wide power range. A discrete component implementation of the method consumes 7.5 /!W and is demonstrated to be effective over a power range of 390 /!W to 750 /!W. A frequency detection method is presented which shows lower power consumption compared to the typically used method. To address the problem of variable frequency excitation, a feed-forward maximum power point tracking control strategy is presented, which demonstrates fast response to changes in the excitation frequency and significant improvement of the performance over the standard maximum power point tracking control. The method relies on an analytical model of the system and the measurement of the excitation frequency. These low-power techniques are implemented in hardware, as a complete self-contained and self-starting energy harvesting system. Depending on the amount of stored energy available, the system moves automatically between passive and active conversion modes, and a mode with optimised feed -forward maximum power point tracking. A minimum operating power of 26 /! W for the active conversion mode has been demonstrated.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:682563 |
| Date | January 2014 |
| Creators | Proynov, Plamen |
| Publisher | University of Bristol |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0021 seconds