Research On Optimal Dispatching Of Power System Containing Wind Energies Based On Transfer Capacity Constraints

A remarkable feature of wind power in China is that it is geographically far away from the load center and needs to rely on the external transmission of multi-layer transmission sections for absorption.The strong uncertainty of wind power generation makes the dispatchers need to set different security constraints on different transmission sections.The traditional dispatching methods are often limited by these constraints,resulting in a large number of wind abandonment,which can not meet the dispatching needs of large-scale wind power connected to the power grid.Therefore,it is necessary to establish a new optimal dispatching model,consider the uncertainty of wind power in the model,and formulate a dispatching scheme that can make full use of cross-section external transmission for wind power consumption.Based on this,this paper takes opportunity constrained programming as the theoretical basis and section constrained optimal dispatching of power system with wind farm as the research content.The main work and achievements are as follows:The probability distribution models of wind power prediction error under normal distribution,Laplace distribution,logistic distribution and Weibull distribution are established respectively.The randomness of wind power prediction power is characterized by the randomness of prediction error.Based on the theory of chance constrained programming,an optimal dispatching model of power system with wind farm based on section constraints is established.For the non-convex chance constraints in the model,they are transformed into second-order cone constraints and linear constraints under the condition of normal hypothesis,which is convenient to be solved by commercial solvers such as GUROBI.Combining the traditional Benders Decomposition Method with the polyhedron approximation of the second-order cone,an improved Benders Decomposition algorithm is obtained.The more complex second-order cone constraints are relaxed in the model,and the optimal solution obtained by solving the relaxed model is substituted into the second-order cone constraints for verification.Only the violated constraints are added back to the original problem through polyhedron approximation,which avoids the difficulty caused by adding a large number of second-order cone constraints to the model in the early stage of solution.Taking the three area IEEE rts-96 system as an example,the algorithm is verified.The results show that the improved Benders Decomposition method can significantly improve the solution speed of the model,and the scheduling plan formulated by the model can make full use of the section for wind power consumption on the premise of ensuring the safety of the system,and can ensure a certain economy.The relationship between wind power prediction error and prediction lead time is analyzed,and the rolling mechanism of opportunity constraint confidence level is determined according to the modified wind power prediction information echelon.The rolling optimal dispatching model of power system with wind farm considering the change of wind power prediction accuracy is established and solved by the improved Benders Decomposition method.The example analysis shows that the proposed rolling optimal scheduling model can properly focus on different scheduling requirements in different periods of the scheduling cycle,and effectively reduce the operation cost of the system without increasing the risk of wind abandonment and load loss.