Multiobjective Approach to Privacy Preservation and Energy Scheduling in Smart Grid

• Main Authors: Hsuan-Hao Chang, 張軒豪
• Other Authors: Wei-Yu Chiu
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/39150453536070608824

碩士 === 元智大學 === 電機工程學系 === 105 === A novel load forecasting mechanism that uses fuzzy logic and big data, termed multipoint fuzzy prediction (MPFP), is proposed. The MPFP can be combined with green buildings and renewable energy sources to reduce peak loads and energy consumption. On the basis of a prediction of load curves, the energy management system (EMS) can discharge energy storage devices when electricity prices are high and charge them when electricity prices are low, reducing costs. Real power demand data were employed to illustrate the validity of the proposed MPFP scheme. After having the prediction of power demand, a power scheduling in a smart home equipped with an energy storage device is investigated. Two objectives are considered: minimizing the energy costs and maximizing the privacy protection. A multiobjective approach is developed to achieve these objectives of a residential user. A multiobjective optimization problem (MOP) is first formulated, and a hybrid evolutionary algorithm is proposed. By solving the MOP, a Pareto optimal solution can be obtained and is further used by the central control unit of the smart home to adjust the loads and storage level over time. Simulation results show that as compared to existing energy management methods, the proposed multiobjective approach can maintain a reasonable energy cost while preserving the user’s privacy at a satisfying level; and it is scalable to a group of smart homes so that a superior peak-to-average ratio (PAR) can be achieved.