Research On Reliability And Dynamic Economic Dispatch Of Power System Considering Wind Power

With energy and environment issues becoming more prominent, renewable energy especially wind power has been developing rapidly and has caused extensive attention around the world. However, wind power has characteristics of randomness and intermittence. Large-scale wind generation connected to power grid has adverse effect on stable operation and generator schedule of the power system. In planning and dispatching power system, it is important to do in-depth study on effect of large-scale integrated wind power to reliability and generator schedule of the power system. Main research work and innovative points of this dissertation are as follows.1) In order to evaluate the effect on reliability of power system with wind farms, reliability model of composite generation and transmission system with wind farm is established considering factors of conventional generating units, transmission lines and transformers etc. A hybrid model of autoregressive moving average (ARMA) and generalized autoregressive conditional heteroscedasticity (GARCH) is proposed to forecast wind speed. The proposed model is proved based on real wind speeds of Donghai wind farm. Meanwhile, an optimal load shedding model is introduced to decrease shedding load. IEEE-RTS is used to verify the proposed method. Analysis and comparison of results show that reliability of wind power can be improved after wind farm joined in. It also indicates that generators have more contribution to reliability of power system than the same rated wind power.2) A new analytical reliability model of wind farms based on frequency and duration approach is proposed. Stochastic characteristic of wind farm output is considered thoroughly in this model and wind farm can be modeled as a conventional unit with derated power states. Roy Billinton Test System (RBTS) is used in this paper. A wind farm model from reference is used to compare with the proposed model. Feasibility and validity of the proposed model is verified by simulation. The proposed analytical model can be utilized by power system planners to conduct reliability calculation with large-scale wind farms.3) An improved particle swarm optimization (IPSO) is developed in this paper. The IPSO combines solution-sharing and Elitist learning strategy (ELS) to enhance computation ability and jump out of local optimal. Dynamic economic dispatch (DED) model with wind power is formulated. In this optimization model, constraints of up spinning reserve and down spinning reserve are introduced to deal with influence of wind power on DED, and valve point effect is taken into accounted in objective function. Simulation results show that the proposed method have good convergence and great economic effect. The proposed method and model has much practical value for DED problem with wind farms.4) A stochastic model is formulated to investigate representation of wind power forecasting uncertainty in dynamic economic dispatch problem. Stochastic model provides a rational and adaptive way to provide adequate spinning reserves at every hour. Cost of shedding load is taken part in objective function to get dispatching schedule considering both economy and reliability. Wind power forecasting inaccuracies are modeled as scenario trees using Monte Carlo simulation method. Scenario reduction method is introduced for enhancing a tradeoff between calculation speed and accuracy of DED solution. IPSO and interior point method are used to solve the proposed model. IEEE 30 and IEEE RTS systems are introduced individually as example system to verify effectiveness of the proposed formulation.5) A new DED formulation with wind farm considering reliability of power system operation is proposed. Conventional constraints and reliability of system are incorporated into formulation simultaneously. Interruption load costs are taken into account in the objective function. Q-function approximation is introduced to obtain analytical LOLP and EENS expression. Different factors, such as LOLP, value of lost load (VOLL), forecast error of wind power and load, are studied. Simulation results show that the proposed model is very effective and feasible. It has important theoretical significance and practical cost value for dispatching operation with wind farms.