| Summary: | 碩士 === 國立高雄應用科技大學 === 電子工程系碩士班 === 103 === Power Scheduling to balance the peak and off-peak load is a difficult but significant issue. Especially in Taiwan, the hot summer often causes a hugh power consumption more than the power generation, dispatch difficulty and part dilemma. According to the structure of power consumption in Taiwan, the residential usage of electricity is about 20%. If we can lead the behavior of household users by in electricity by taking the advantages of smart grid a and smart meter, the gap between the peak and off-peak load will be reduced to provide a more convenient and secure electricity service. However, in the current solution, the most power generators adjusted electricity support and transmission according to electricity usage. In this way, it can't be really improved the peak demand for electricity consumption without the presentation of real-time and sufficient power consumption information at home. If there are some useful information available for home users to help them adjusting their electricity usage and optimizing peak and off-peak electricity load. We apply genetic algorithms and load balance indicators to design a power schedule method. The proposed method also analyzes the user past history of electricity usage to derive user habits that show the peak load time of user. At the same time, the power supply is another considerations. The proposed method recommends electricity usage time to approximate user habits. The proposed method can reduce high electricity consumption at the peak load time, and improves load balance in smart grid. In addition, it uses the simple structure of Arduino single-chip and Bluetooth 2.1 wireless modules to simulate the Bluetooth 4.0 wireless transmission that can provide one to many communication in this research to fit the future trend of smart home technology. The proposed Bluetooth emulation is applicable to be implemented to integrate the smart home appliances. We develop power manager, power schedule method, and investigate the effect between the electricity data collection and the number of appliances. According to the experimental results, the proposed method can work with at least 10 appliances in a future smart home. If the appliance increase in the home, the transferable model will be a solution to improve the performance. By optimizing peak electricity load, the proposed method can save electricity and improve the load balance problem in comparison with the previous work Therefore, the proposal can effectively improve the unbalance of electricity supply and demand, and have high feasibility for household in future.
|
|---|
