Ambient energy is omnipresent in our environment in the form of solar and radio frequency radiation, thermal energy, energy from chemical and biological sources—such as salinity gradient and blood sugar levels, respectively—and mechanical energy from natural phenomena such as waves or from vibrations generated by man-made activity. Ambient vibrations, in particular, were until recently considered of little interest as an energy source due to their very low energy level. Past attempts to harvest them have focused on physical energy conversion methods, such as microelectromechanical systems, which do not require fossil fuels.
Today, the main challenge is to harvest enough energy to power electronic devices. Since these devices are becoming smaller in size and consume less power, they may be well suited for ambient vibration conversion systems for charging batteries or supplying power directly.
The integration of multidisciplinary research produced mechatronic harvesting systems: fruits of the synergistic integration of mechanics, electronics, control theory, and computer science within product design and manufacturing to improve functionality. These can be used as autonomous source of electrical energy for remote or wireless applications powered by ambient mechanical vibrations from machines, aircraft, ships, bridges, buildings, and so forth.1 Until now, these systems were limited by their design, which did not allow optimization of the amount of electric energy harvested.
Here, we report our work on a generic vibration energy harvester2 optimized using artificial intelligence methods. It works by amplifying ambient vibrations before converting them into electrical energy. This device consists of three integrated components, including a precise mechanical part, an electromagnetic energy converter, and an electrical load (see Figure 1). Due to the integration of the harvester's components, its parameters are interdependent. As a result, it only operates correctly in a narrow bandwidth once tuned up at the resonance frequency, where the amplified movement is at a maximum and generates the largest amount of electricity.
Figure 1. Schematic diagram of the principle of the vibration energy harvester, in which a precise mechanical part—consisting of a mass m, a spring k, and a mechanical damping element bm—amplifies the ambient vibration (Av) movement through a patented springless resonance mechanism. The electromagnetic energy converter (magnet movement against coils L and Rc) then converts the amplified movement x into electrical energy, according to Faraday's law. The electrical energy is consumed by the electrical load RL (e.g., wireless application), and the power consumption causes a damping effect be.
A major part of our work involved optimizing our device. Classical optimization methods3 rely on analytical techniques based on differential calculus. However, these have limited scope because they cannot solve practical problems described by functions that are not continuous or differentiable. Instead, our method employs artificial intelligence consisting of a stochastic algorithm—SOMA (for self-organizing migrating algorithm)—which was inspired by the pattern of behavior of animal groups in the wild.4
The aim of our process5 was to maximize the harvested power by identifying optimal mechanical, electrical, and electromagnetic parameters with the goal of operating it with traditionally sized batteries. Given that harvested power is proportional to the weight and movement of the mass within the device, we strove to minimize its volume and weight. We also performed several studies, including maximization of the harvested output power obtained with a fixed-volume harvester, minimization of the volume or weight of the harvester at constant output power, and complex optimization of all of the parameters combined.
We then used the computing language MATLAB to create a dynamic model of the vibration energy harvester and a simulator called Simulink to model the response of our device to ambient vibrations.6 In these simulations, the harvester showed it could operate as an autonomous source of energy in a vibratory environment. By varying the 22 independent parameters connected to its design and geometry, we produced simulations of their outputs, which were then used as inputs for the SOMA application to find the best combination of independent parameters to optimize the device. These results show the power of the SOMA method for analyzing complex systems with dozens of parameters.
By accounting for the nonlinear behavior of the harvester, our simulation model closely reflects the actual device. Optimization can therefore help in generating the maximum amount of electrical power. As next steps, we plan to improve the quality factor of the model and to use it for a new harvester design tailored to real-life conditions and to meet the requirements of wireless applications.
Zdenek Hadas, Vladislav Singule
Faculty of Mechanical Engineering
Brno University of Technology (BUT)
Brno, Czech Republic
Zdenek Hadas was awarded a PhD (2007) from BUT for his work on energy harvesting from mechanical vibrations. He is currently a researcher at the BUT Institute of Solid Mechanics, Mechatronics, and Biomechanics. His research focuses on energy harvesting from mechanical vibrations and modeling mechatronic systems.
Vladislav Singule is head of the Department of Electrical Enneering at the Institute of Production Machines, Systems, and Robotics. He became an associate professor at the Military Academy in Brno in 1984 while focusing on line measurement techniques. His current research interests include electrical drives, power electronics, control motion of electromechanical systems, and mechatronics.
Jiri Kurfust, Cestmir Ondrusek
Faculty of Electrical Engineering and Communication
Brno, Czech Republic
Jiri Kurfurst is a PhD student in the Department of Power Electrical and Electronic Engineering at BUT. His thesis focuses on optimizing electrical machines.
Cestmir Ondrusek is head of the Department of Power Electrical and Electronic Engineering at BUT. He became an associate professor at the Faculty of Electrical Engineering and Communication in 1986. His main interests are electrical machines, electric drives, diagnostics, chaos in electromechanical systems, the use of artificial intelligence in the design of electromechanical systems, and simulating the dynamic behavior of electromechanical systems.
1. S. Priya, D. J. Inman, Energy Harvesting Technologies, Springer, 2009.
2. S. P. Beeby, M. J. Tudor, N. M. White, Energy harvesting vibration sources for microsystems applications, Meas. Sci. Technol. 17, no. 12, pp. 175-195, 2006.
3. S. S. Rao, Engineering Optimization: Theory and Practice, Wiley, 2009.
4. I. Zelinka, SOMA—Self-Organizing Migrating Algorithm, ch. 7, New optimization techniques in engineering, Springer, 2004.
5. Z. Hadas, C. Ondrusek, V. Singule, Power sensitivity of vibration energy harvester, Microsyst. Technol. 16, no. 5, pp. 691-702, 2010.
6. Z. Hadas, V. Singule, C. Ondrusek, M. Kluge, Simulation of vibration power generator, Recent Advances in Mechatronics, pp. 350-354, Springer, Berlin, 2007.