Efficient design for micro-scale energy harvesting systems

Chao Lu, Purdue University

Abstract

Rapid advances in nanoscale integration and low power design techniques have resulted in a variety of highly miniaturized electronic systems for use in emerging application domains such as wireless sensor networks. Essential requirements for these systems include ultra-low cost, stringent size or volume constraints, long operation time and maintenance-free operation. A key challenge during the design and operation of these systems is the issue of conveniently and cost-effectively provide energy to these systems in order to sustain long-term operation. Conventional solutions based only on energy storage buffers, such as rechargeable batteries, cannot meet all of the above requirements. Micro-scale energy harvesting from the environment has the potential to satisfy all of these application requirements and, hence, is the focus of this research. Even though there have been several micro-scale energy harvesting prototypes presented in literature, the design of efficient micro-scale energy harvesting systems has not yet been comprehensively studied. An efficient design implies rapid development time, low design effort, a cost-effective solution and optimal system performance. The design of efficient micro-scale energy harvesting systems requires an in-depth understanding of numerous design tradeoffs and considerations at all layers of design abstraction. This dissertation makes several contributions towards this. The first contribution is energy-efficient power converter circuit design using several new techniques in terms of circuit topology, transistor sizing and the number of stages. These optimization techniques reduce internal power loss, improve the power transfer capability, and hence boost the energy harvesting performance substantially. The second contribution is a new automatic maximum output power tracking method. The proposed technique consumes very little power, have a very low hardware cost, and has fast tracking speed. Therefore, it is well suited for micro-scale energy harvesting systems. The third contribution is based on the fact that currently, design of micro-scale energy harvesting systems depends heavily on designer's intuition and experience. Due to the lack of efficient design and simulation tools, current designs are mainly point solutions and no design methodologies exist for finding optimal design solutions. In order to overcome this challenge, a design framework that enables systematic design space exploration and quick performance evaluation of micro-scale energy harvesting systems at the early design phase is proposed. Unlike current design practices, which optimize each system building block in isolation, the proposed design framework supports cross-layer optimization of heterogeneous system buildings blocks. The proposed framework takes various design parameters into account, can accurately mimic system operation behavior and can evaluate the overall impact of various design parameters on the system. In addition to significantly reducing the design effort and enabling easy optimization of design parameters, the proposed framework also leads to a three orders of magnitude reduction in simulation time compared to low-level simulation tools. The accuracy and efficacy of the proposed framework are validated and demonstrated through multiple case studies.

Degree

Ph.D.

Advisors

Raghunathan, Purdue University.

Subject Area

Computer Engineering|Electrical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS