High-level modelling and performance optimisation of mixed-technology energy harvester systems Tom J Kazmierski, Leran Wang, Bashir M Al-Hashimi University of Southampton, UK MOS-AK, Edinburgh 19 September 2008
nergy harvesting in electronics State-of-the-art Energy harvesting from the environment Pervasive computing 2
Vibration-based energy harvester system AC Voltage Voltage booster System consisting various components from different physical domains Reported equivalent circuit models are inadequate VHDL-AMS as modelling language Performance loss due to close mechanical-electrical interaction Systematic EH optimisation based on HDL model 3
EH modelling approaches macro-generator models - Ideal voltage source - Equivalent circuit 4
Proposed accurate HDL model - Use VHDL-AMS as modelling language - Describe micro-generator as a series of analytical equations - Mixed-technology (mechanical, magnetic, electrical etc.) 5
Case study design State-of-the-art EH (R. Torah, et.al, Development of a cantilever beam generator employing vibration energy harvesting. In Proceedings of The 6th Int. Workshop on Micro and Nanotechnology for Power Generation and Energy Conversion Applications, 2006.) Vibration-based based Electromagnetic micro-generator Voltage multiplier as booster Super capacitor as storage 6
Experimental micro-generator Electromagnetic Cantilever based Fixed coil, moving magnets Small volume: 150 mm 3 7
Micro-generator model B B R r -B z(t) R r -B z(t) (b) Small displacement (c) Large displacement (a) Φ = ( R + r)*2* B* N (b) Φ = ( 2 2 2 2 R z ( t) + r z ( t))*2* B * N (c) Φ = ( 2 2 2 2 R ( H z( t) ) + r ( H z( t) ) )* B* N 8
Voltage booster Voltage multiplier as the voltage booster 6-stage Villard configuration Schottky diodes (BAT760) C1 C3 C5 Microgenerator D1 D2 D3 D4 D5 D6 0.22F C2 C4 C6 9
Experimental setup Micro-generator sitting on vibration generator Frequency: 50Hz, acceleration level: 0.6 m/s 2 Data collected by LabView 10
Simulation and verification Different energy harvester models compared with experimental measurements Ideal voltage source fails: voltage booster can greatly affect the behaviour of micro-generator Voltage (V) 3 2.5 2 1.5 1 0.5 Ideal voltage source Equivalent circuit model Experimental measurement Proposed HDL model 0 0 30 60 90 120 150 Time (min) 11
Simulation and verification Equivalent circuit model is inaccurate: L=m, C=1/k, R=b is over simplification 0.8 Voltage (V) 0.0-0.8 1.0 0.0-1.0 1.0 0.0 Equivalent circuit model Proposed HDL model Experimental measurement -1.0 1200 1220 1240 1260 Time (ms) 12
EH performance optimisation Performance loss due to mechanical-electrical interaction η Loss =(E Harvested -E Delivered ) / E Harvested Integrated performance optimisation in VHDL-AMS testbench. (L. Wang and T.J. Kazmierski, VHDL-AMS based genetic optimization of a fuzzy logic controller for automotive active suspension systems, BMAS 2005) 13
Genetic optimisation in VHDL-AMS testbench GA parameters: 7 genes: 3 from micro-generator, 4 from transformer Fitness: super capacitor charging rate v dot Tournament selection Elitism Arithmetic crossover Gene mutation Parallel GA: genes in one generation are evaluated simultaneously VHDL-AMS finite state machine 14
H performance optimisation Use a voltage transformer as booster Optimisation algorithm found that the transformer exhibits better performance than voltage multiplier 15
Optimisation results Super capacitor charging simulation waveforms Voltage (V) 2.0 1.6 1.2 Optimised Un-optimised 0.8 0.4 0.0 0 30 60 90 120 150 Time (min) 16
Optimisation results Micro-generator Outer radius of coil (R) 1.2mm Coil turns (N) 2300 Internal resistance (Rc) 1600Ω Voltage transformer Resistance (Ω) No. of turns Primary winding 400 2000 Secondary winding 1000 5000 Un-optimised Micro-generator Outer radius of coil (R) 1.1mm Coil turns (N) 2100 Internal resistance (Rc) 1400Ω Voltage transformer Resistance (Ω) No. of turns Primary winding 340 1900 Secondary winding 690 3800 Optimised 17
Further work: automated design flow Architectural synthesis GA-based performance optimisation 18
Conclusion Integrated approach to EH modelling and performance optimisation has been developed Existing electrical equivalent circuit models of microgenerator cannot predict the voltage booster s performance accurately HDL model based on analytical equations can describe the actual shape and size of an EH Through performance optimisation it was possible to increase the energy harvesting rate by 30% 19