| Summary: | 碩士 === 國立清華大學 === 電子工程研究所 === 101 === The price of gas and electricity has increased rapidly in recent years, so do environmental protection attract more and more attention. Many countries are encouraging the development of alternative and renewable energy sources. Under the concern of space limitation and noise reduction in a city, the vertical, small-scale wind power generation system turns out to be a good option for distributed clean energy sources. In order to increase the system’s overall efficiency, as well as to decrease the cost, many maximum-powerpoint-tracking (MPPT) algorithms have been proposed in literatures. This thesis aims to identify MPPT algorithms suitable for wind-power generation and to implement the algorithms as a microsystem on a chip. In a windpower generation system, the operating voltage and operating current are nonlinearly dependent on the wind speed and the temperature. By controlling the operating voltage and current optimally, the energy-generation efficiency can be boosted greatly. The optimal-torque-control method is able to track MPPT much faster than other algorithms when the wind speed varies continuously. However, the optimal torque control could track the wrong maximum power point as the air density (environment) changes. Therefore, this thesis investigates the feasibility of improving the optimal-torque-control method and realizing the algorithms as a microsystem on a chip.
The major contributions of this thesis are described as follows. “An optimal torque control method with self-adaptability to environmental changes” is developed and verified by both simulation and implementation in a microprocessor. Taking into account the fact that the wind speed changes dynamically and continuously, this thesis further adopts the “dynamic optimal torque control (DOTC) method”to track maximum power point fast and reliably as wind speed changes. After verifying the capability of the DOTC method in simulation, this DOTC algorithm is further realized as a microsystem with analog integrated circuits. The microsystem is designed and fabricated with the TSMC CMOS 0.18m process. The power supply voltage is 1.8V. Total chip layout area is 1.963*1.538mm2 and it consumes less than 1.5mW. The experimental results reveal that the microsystem achieves a power-conversion coefficient of 27.7%, compatible with the ideal powerconversion coefficient of 28.1%, and the error is merely 1.42%.
|
|---|
