Direct drive wind power plant cluster division method considering low voltage ride through characteristic

Abstract

The invention discloses a direct drive wind power plant cluster division method considering a low voltage ride through characteristic. The method includes the following steps: 1, establishing a detailed module of a direct drive wind power plant; 2, collecting input wind energy of each wind generating set; 3, collecting a terminal voltage value of each wind generating set at a fault moment; 4, setting a grid-side converter active current reference value when network voltage drops; 5, judging the conduction condition of an unloading circuit of each wind generating set; and 6, applying an immunerandom k value and an improved K mean clustering algorithm of a sensitive clustering center to complete direct drive wind power plant cluster division. The direct drive wind power plant cluster division method considering the low voltage ride through characteristic can accurately represent the conduction condition of the unloading circuit of each wind generating set in a direct drive wind power plant, and improve direct drive wind power plant cluster division quality, and thus effectively improve fitting precision of a direct drive wind power plant equivalent model to wind power plant output external characteristics.

Description

technical field [0001] The invention relates to the technical field of equivalent modeling of direct-drive wind farms, in particular to a direct-drive wind farm fleet division method considering low-voltage ride-through characteristics. Background technique [0002] The direct-drive permanent magnet wind turbine saves the gearbox with a high failure rate, and has the advantages of small mechanical loss, high operating efficiency, low maintenance cost, and wide speed range. It is connected to the grid through a full-power converter, and the generator does not directly Coupled with the grid, it has stronger low-voltage ride-through capability than doubly-fed wind turbines, and has increasingly become the mainstream wind turbine type in current wind farms. With the rapid development of wind power technology, the scale of grid-connected wind farms is gradually expanding. If each wind turbine uses a detailed model, many problems will arise, such as too long simulation calculation...

AI Technical Summary

Problems solved by technology

[0004] However, most of the existing multi-machine characterization methods fail to fully consider the difference in the terminal voltage drop of wind turbines when dividing the fleet, and ignore the influence of the conduction of the unloading circuit on the difference in the transient characteristics of the direct-drive permanent magnet wind turbine. , cannot accurately fit the output external characteristics of the wind farm

In addition, the traditional kmeans clustering algorithm has the disadvantages that the number of clusters k needs to be given in advance; it is heavily dependent on the selection of the initial cluster center, making the clustering result easily fall into a local optimal solution; it is easily affected by outlier data and other defects

The above defects of the traditional kmeans clustering algorithm greatly reduce the clustering quality, resulting in the inaccuracy and instability of the cluster division results

Images