Research On Reactive Power Optimization In Complex Power Grid Containing Wind Farm Groups

Wind power, as the most widely used of all renewable energy, has got greatlydeveloped in recent years. Large-scale wind farms and wind farm groups integration willhave a lot of influence on security, stability and economy of power system. So it isnecessary to take reactive power compensation measures to optimize the power flowdistribution, schedule reactive power distribution reasonably, reduce the power losses andimprove the voltage stability of wind farms.A new modeling method for wind farm groups based on hierarchical clusteringmethod is proposed in this paper. According to different impacting extent of the windfarm with DFIG during system faults, wind turbines with similar dynamic behaviors areidentified. Then define a cluster classification index, select the appropriate clusteringthreshold to aggregate the same wind turbines, so the dynamic multi-machine equivalentmodel of wind farms with DFIG is obtained. The proposed model more accuratelydescribes the dynamic characteristics of the grid connection point and has importantengineering value.From the point of view of the probability, the random output of doubly-fed windfarm has great impact on the grid voltage. The stochastic response surface method isapplied to wind power systems with uncertainty analysis, you can quickly estimate theprobability of voltage response. The sum of power loss is selected as the objectivefunction and IA-PSO algorithm is used to solve the objective function. Under thecircumstances that the installed capacity and position of reactive power compensationdevices are unknown, the optimal solution of reactive power compensation is obtained.Considering the cost of double-fed wind farm reactive power regulation andinvestment of reactive power compensation equipment, a strategy of comprehensivereactive power optimization is proposed. Two-stage optimizations, i.e. the planoptimization and the real-time optimization are presented. The reactive power of windfarm is taken as the control variables in the real-time reactive power optimization stage.The IA-PSO algorithm is adopted in the two-stage optimization. So the comprehensiveoptimization result of active power loss, reactive power compensation equipment investment and cost of wind farms providing reactive power is acquired. The proposedstrategy is applied to the IEEE30bus test system, and the simulation results are verifiedto meet the demand of dispatching and have a high practicability.