Scenario Probabilistic Load Flow Calculation Method Cosidering Wind Farms Correlation And Its Application In Reactive Power Optimization

As the energy crisis and environment pollution problem increasingly prominent, the development and utilization of new energy has got more and more attention. Among them, the development and utilization technology of wind energy is the most mature, and there are already large-scale wind farms connecting to the grid. However, affected by wind energy resources, wind farm output is highly random, intermittent and cross-correlation. That makes the power flow distribution and voltage stability have been greatly affected. According to the adjacent wind farms output randomness and correlation characteristics, this paper establishes the scenario probability model of wind farms output to analysis the effect after wind farms connect to the grid, and applied to the study of reactive power optimization.First, this paper proposed a scenario probabilistic load flow calculation method considering wind farms correlation based on cluster analysis and Copula function. The cluster analysis can classified data according to the approximate characteristics of itself. Copula function can build the probability model of complex and correlation data. This paper combines their advantage, taking into account the complex and changeable characteristic correlation between the adjacent wind farms output, establishes the scenario probability model of wind farms output, completes the probability assessment of power system in each scene by probability load flow based on Latin hypercube sampling. Two adjacent wind farms in Australia are accessed in IEEE-30 node system to test and verify the rationality and validity of the proposed method. The simulation results show that the proposed model can describe the correlation between wind farm output better, and get more accurate results of probability load flow.Then, this paper applies the proposed method into reactive power optimization. Establishing the probability reactive power optimization model which targets to minimum the weighted value of system network loss expectation, generator reactive power and node voltage deviation expectation. Using particle swarm optimization to solve the model, and get the optimal reactive power control strategy in each wind farm output scene. Compared with control strategy of deterministic scenario reactive power optimization model, the proposed method can improve the adaptability of reactive power control strategy to the random changes of wind farm output, and ensure the power system operating under the optimal condition with maximum probability.