Study On UFLS Optimization Based On The Sharing Information In The Smart Grid Environment

Exsiting under frequency load shedding method uses local information inside a substation to perform load shedding, and does not take advantages of smart grid development when wider information, more interaction and networking are readily available. Therefore, this thesis seeks to study the optimal under frequency load shedding strategy based on sharing information in the smart grid environment. The study looked into the available information inside and outside a substation, different technical architectures and decision-making methods. From this study, an optimal networked under frequency load shedding scheme is proposed. The scheme consists of a technical architecture which achieves wide-area coordination and user interactive coordination, and a function which decides optimally the shedding load and the shedding capacity.This thesis first summarizes the key under frequency load shedding information according to the shedding flow, then presents in detail the information available in the smart grid environment from the system level, the substation level and the load level. Among them, it focuses on the load characteristic information and friendly user (electric vehicles) interactive features. The acquisition and transmission of all kinds of information is also analysed.Under frequency load shedding technology based on communication network is studied after analysis of characteristics and problems of traditional under frequency load shedding techniques, which accords with the current smart grid developing trend. The optimized substation under frequency load shedding architecture and its control strategy is then presented, based on the intelligent substation and networked information. The wide-area under frequency load shedding architecture using station control system as the subunit is discussed in order to realize the double coordination optimization.In order to realize coordinated optimization on the user’s side, a load sorting algorithm using grey correlation analysis is proposed to calculate the load importance, load frequency effect coefficient, load factor and other load characteristics. A precise load shedding strategy based on the real-time load detection is proposed, which meets the needs of the total shedding capacity. The strategy is further improved with coordinated optimization considering the connection of electric vehicles to make full use of demand side response and to reduce shedding cost. The proposed algorithm is verified using the simulation examples.In order to realize coordinated optimization on the wide-area side, a wide-area real-time intelligent load capacity distribution model is built. A unit load factor is introduced as a weighting coefficient of the shedding load to calculate the comprehensive cost. This factor is calculated using an analytic hierarchy process method. The steady state frequency constraint is obtained using direct frequency stability analysis method, with information obtained from the wide-area measurement data. This multiple constraint problem is resolved using particle swarm optimization algorithm. The effectiveness of the proposed comprehensive cost optimization model is verified using simulation example.This thesis studies the problems to decide optimally the shedding load and the distribution shedding capacity, and exploites thinking for further research on the improvement of under frequency load shedding adaptability in the smart grid environment.