The Study On Optimal Generation Dispatch Considering Large Scale Wind Power Integrating

Wind power as a clean energy source has applied to modern power systemswidely. With wind power integrating into grid increasing, fluctuation and randomnessof wind should have an impact on dispatch. Therefore research in traditionalgenerators economic dispatch under certain reliability is meaningful to cope with thedifficulties. In our country, the distribution of large-scale wind resources isinconsistent with the critical load centers, so the fact is large capacity wind powertransmitted by long distance. Therefore this paper takes the static voltage stability intoeconomic dispatch model. This paper studies the economic dispatch considering largescale wind from the following aspects:(1) In order to meet the static voltage stability of the system, the system staticvoltage stability constraints is added into the economic dispatch. In this paper, twoindicators have introduced to reflect the system static voltage stability, and they arethe static reserve factor and voltage stability indicator. Each of indicators has its owncharacteristic, and is suitable in different application scenarios.(2) In order to deal with the fluctuation of wind power, chance constrainedprogramming and latin hypercube sampling are proposed. Considering the fuel costsof traditional power plants and environmental factors as the objective function,building uncertainty model through using the principle of chance constrainedprogramming into constraints to former the probability constraints, and sampling bylatin hypercube sampling method to get all possible wind power sampling valueswhich are applied into the constraints in fitting that can simulate the randomness ofwind power accurately.(3) In the optimization algorithm, the adaptive differential evolution algorithmand augmented Lagrangian algorithm are combined to solve the problem. On one side,through the adaptive improve of differential evolution algorithm, adjustingevolutionary parameters which include the variation factor and cross constant, andachieving the optimization of variables, on the other hand, by the augmentedLagrangian algorithm, transforming the constrained condition into function which isassociated the penalty function with the Lagrange multiplier, added into the objectivefunction.