Скачать презентацию Energy Harvesting for Pervasive Sensing Paul D Mitcheson Скачать презентацию Energy Harvesting for Pervasive Sensing Paul D Mitcheson

6aa97d41f299b5dba7307469830be0c3.ppt

  • Количество слайдов: 38

Energy Harvesting for Pervasive Sensing Paul D. Mitcheson, Eric M. Yeatman Department of Electronic Energy Harvesting for Pervasive Sensing Paul D. Mitcheson, Eric M. Yeatman Department of Electronic & Electrical Engineering Imperial College London P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Energy Harvesting: what is it? • Taking useful advantage of power sources already present Energy Harvesting: what is it? • Taking useful advantage of power sources already present in the local environment • This energy would otherwise be unused or wasted as e. g. heat • “local” being local to the powered device or system • Extracted power levels generally not limited by source, but by size and effectiveness of generator (“harvester”) P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Energy Harvesting: what is it for? • Normally not as a primary source of Energy Harvesting: what is it for? • Normally not as a primary source of power, but for applications where mains power is not suitable, because of: • Installation cost • Mobility • Remote/inaccessible/hostile location • Usual alternative is batteries: • Avoid replacement/recharging • Avoid waste from used batteries P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

How Much Power? World electrical generation capacity 4 terawatts Power station 1 gigawatt House How Much Power? World electrical generation capacity 4 terawatts Power station 1 gigawatt House 10 kilowatts Person, lightbulb 100 watts Laptop, heart 10 watts Cellphone power usage 1 watt Wristwatch, sensor node 1 microwatt Transmitted Cellphone signal 1 nanowatt P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Cost example: • Mains electricity: consumer price 15¢ / k. Whr • Alkaline AA Cost example: • Mains electricity: consumer price 15¢ / k. Whr • Alkaline AA battery: 1 € / 3 Whr • Factor of 2, 000 P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Energy Harvesting Applications • Key application is wireless sensor networks • Sensors can be Energy Harvesting Applications • Key application is wireless sensor networks • Sensors can be very low power • Small size often important • Minimal maintenance crucial if many nodes • Implementation of WSNs could lead to higher energy efficiency of buildings etc P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

1 cc wireless sensor node, IMEC P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, 1 cc wireless sensor node, IMEC P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Sensor Node Power Requirements – How much power does our harvester need to supply? Sensor Node Power Requirements – How much power does our harvester need to supply? • Sensing Element • Signal Conditioning Electronics • Data Transmission P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Sensing Element Simple signals - temperature, pressure, motion – require electrical power above thermal Sensing Element Simple signals - temperature, pressure, motion – require electrical power above thermal noise limit. NT 10 -20 W/Hz For most applications, this is negligible P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Signal Conditioning Likely principal function: A/D Converter Recent results: Sauerbrey et al. , Infineon Signal Conditioning Likely principal function: A/D Converter Recent results: Sauerbrey et al. , Infineon (’ 03) Power < 1 m. W possible for low sample rates! P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Data Transmission: Required Power 50 Transmit Power (d. Bm) 40 Conclusions: Typical indoor Loss Data Transmission: Required Power 50 Transmit Power (d. Bm) 40 Conclusions: Typical indoor Loss exponent (3. 5) 30 20 10 0 Power independent of bitrate for low bit-rate -30 d. Bm (1 m. W) feasible for room-scale transmission range Ideal free-space propagation -10 -20 -30 -40 -50 1 10 1000 Range (m) Figure: F. Martin, Motorola P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Estimated Total Power Needs • Peak power 1 – 100 u. W • Average Estimated Total Power Needs • Peak power 1 – 100 u. W • Average power can be below 1 u. W Batteries: Present Capability • 10 m. W yr for 1 cm 3 battery feasible • Not easy to beat! • Useful energy reservoir for energy harvesting P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Fuel-Based Power Sources • Energy density much higher than for batteries, 10 k. J/ Fuel-Based Power Sources • Energy density much higher than for batteries, 10 k. J/ cm 3 • Technology immature, fuel cells most promising Micro fuel cell, Yen et al. Fraunhofer Inst. P. Mitcheson, Nov. 2008

Energy Scavenging : Sources Energy Source Light Ambient light, such as sunlight Thermal Temperature Energy Scavenging : Sources Energy Source Light Ambient light, such as sunlight Thermal Temperature gradients Magnetic and Electro-magnetic waves Kinetic Volume flow (liquids or gases) Movement and vibration Conversion Mechanism Solar Cells Thermoelectric or Heat Engine Magnetic induction (induction loop) Antennas Magnetic (induction) Piezoelectric Electrostatic P. Mitcheson, Nov. 2008

Solar Cells • highly developed • suited to integration • high power density possible: Solar Cells • highly developed • suited to integration • high power density possible: Ø 100 m. W/cm 2 (strong sunlight) • but not common: Ø 100 m. W/cm 2 (office) • Need to be exposed, and oriented correctly Solar cell for Berkeley Pico-Radio P. Mitcheson, Nov. 2008

Solar Cells in Energy Harvesting Applications: • Cost not the main issue • Availability Solar Cells in Energy Harvesting Applications: • Cost not the main issue • Availability of light is key P. Mitcheson, Nov. 2008

Thermal • need reasonable temperature difference (5 – 10 C) in short distance • Thermal • need reasonable temperature difference (5 – 10 C) in short distance • ADS device 10 m. W for 5 C • even small DT hard to achieve Heat engine, Whalen et al, Applied Digital Solutions P. Mitcheson, Nov. 2008

Seiko Thermic (no longer in production) P. Mitcheson, Nov. 2008 Seiko Thermic (no longer in production) P. Mitcheson, Nov. 2008

Ambient Electromagnetic Radiation Graph: Mantiply et al. 10 V/m needed for reasonable power: not Ambient Electromagnetic Radiation Graph: Mantiply et al. 10 V/m needed for reasonable power: not generally available P. Mitcheson, Nov. 2008

Motion Energy Scavenging • Direct force devices • Inertial devices P. Mitcheson, Nov. 2008 Motion Energy Scavenging • Direct force devices • Inertial devices P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Direct Force: Heel Strike Heel strike generator: Paradiso et al, MIT P. Mitcheson, Nov. Direct Force: Heel Strike Heel strike generator: Paradiso et al, MIT P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Direct Force: larger scale East Japan Railway Co. • Energy harvesting ticket gates P. Direct Force: larger scale East Japan Railway Co. • Energy harvesting ticket gates P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Inertial Harvesters • Mass mounted on a spring within a frame • Frame attached Inertial Harvesters • Mass mounted on a spring within a frame • Frame attached to moving “host” (person, machine…) • Host motion vibrates internal mass • Internal transducer extracts power P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Available Power from Inertial Harvesters • Peak force on proof mass • Damper force Available Power from Inertial Harvesters • Peak force on proof mass • Damper force < F or no movement • Maximum work per transit W = Fzo = mw 2 Yozo • Maximum power P = 2 W/T F = ma = mw 2 Yo = mw 3 Yozo/p P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

How much power is this? 10 x 2 mm 3 x 0. 6 mm How much power is this? 10 x 2 mm 3 x 0. 6 mm Plot assumes: • Si proof mass (higher densities possible) • max source acceleration 1 g (determines Yo for any f) P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Achievable Power Relative to Applications Plot assumes: • proof mass 10 g/cc • source Achievable Power Relative to Applications Plot assumes: • proof mass 10 g/cc • source acceleration 1 g Sensor node watch cellphone laptop P. Mitcheson, Nov. 2008

Implementation Issues: Transduction Mechanism Piezoelectric? • Difficult integration of piezo material • Reasonable voltage Implementation Issues: Transduction Mechanism Piezoelectric? • Difficult integration of piezo material • Reasonable voltage levels easy to achieve • Suitable for miniaturisation P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Typical Inertial Generators Piezoelectric Wright et al, Berkeley Ferro solutions P. Mitcheson, Nov. 2008 Typical Inertial Generators Piezoelectric Wright et al, Berkeley Ferro solutions P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Implementation Issues: Transduction Mechanism Electromagnetic? • Dominant method for large scale conversion • Needs Implementation Issues: Transduction Mechanism Electromagnetic? • Dominant method for large scale conversion • Needs high df/dt to get damper force (f = flux) • df/dt = (df/dz )(dz/dt ) • Low frequency (low dz/dt) needs very high flux gradient • Hard to get enough voltage in small device (coil turns) • Efficiency issues (coil current) Variant: magnetostrictive P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Typical Inertial Generators Magnetic Southampton U. CUHK P. Mitcheson, Nov. 2008 P. D. Mitcheson, Typical Inertial Generators Magnetic Southampton U. CUHK P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Implementation Issues: Mechanism Electrostatic? • Simplementation, no field gradient problem • Suitable for small Implementation Issues: Mechanism Electrostatic? • Simplementation, no field gradient problem • Suitable for small size scale • Damping force can be varied via applied voltage • But needs priming voltage (or electret) P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Typical Approach: Constant Charge Input phase Output phase P. Mitcheson, Nov. 2008 Typical Approach: Constant Charge Input phase Output phase P. Mitcheson, Nov. 2008

Prototype MEMS Device Assembled generator Detail of deep-etched moving plate P. Mitcheson, Nov. 2008 Prototype MEMS Device Assembled generator Detail of deep-etched moving plate P. Mitcheson, Nov. 2008

Device Operation Output > 2 m. W P. Mitcheson, Nov. 2008 P. D. Mitcheson, Device Operation Output > 2 m. W P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Other Options: Rotating Mass Example : Seiko Kinetic P. Mitcheson, Nov. 2008 P. D. Other Options: Rotating Mass Example : Seiko Kinetic P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Large Inertial Generators Backpack: U Penn • 7 watts! P. Mitcheson, Nov. 2008 P. Large Inertial Generators Backpack: U Penn • 7 watts! P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Pervasive Sensing for Energy Generation P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March Pervasive Sensing for Energy Generation P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009

Conclusions • Power levels in the microwatt range are enough for many wireless sensor Conclusions • Power levels in the microwatt range are enough for many wireless sensor nodes • Small energy harvesters can achieve these levels • Help enable pervasive sensing by eliminating maintenance burden Contact: paul. mitcheson@imperial. ac. uk Review Paper: Mitcheson, Yeatman et al. , “Energy Harvesting From Human and Machine Motion for Wireless Electronic Devices”, Proceedings of the IEEE 96(9), 1457 -1486 (1998). P. Mitcheson, Nov. 2008 P. D. Mitcheson, IOM, March 2009