Wake identification method of wind driven generator

Abstract

The invention provides a wake identification method of a wind driven generator. The method comprises the steps: collecting all flow field data of a tested wind driven generator, carrying out derivation calculation on the flow field data through a differential method, obtaining a speed gradient tensor, and then, calculating flow invariant data of each point of a flow field; analyzing the known data, and dividing the collected data into strong turbulence and weak turbulence according to a preset standard; taking the flow invariant data as an input quantity, taking strong turbulence and weak turbulence data as learning objects, and generating a recognizer through machine learning algorithm software; and inputting the invariant data of the flow field required to be identified into the recognizer, and then, drawing a data area conforming to the strong turbulence according to a predetermined standard to obtain a wake area of the wind driven generator which generates the current flow field required to be identified. According to the method, a big data analysis method in modern computer science is used, only sufficient data needs to be provided and other human interferences do not need tobe added, so that objectivity of calculation results can be guaranteed.

Description

technical field [0001] The invention relates to the field of wind power generation, in particular to an identification method for establishing a learning model through known data to identify the wake of any wind power generator. Background technique [0002] Traditional wind turbine wake identification usually adopts the mode of "one physical quantity + one threshold". For example, for the time-averaged wake field, the center position can be determined by the maximum value point of the speed loss, and the wake range can be determined by the speed loss greater than a certain threshold. , for the instantaneous velocity field, the wind turbine wake can be identified by the vorticity of the flow field and a certain threshold, but this method has the following two disadvantages: [0003] 1. The robustness of the univariate recognizer is poor. For wake recognition with different flow parameters, the threshold needs to be adjusted artificially. However, the selection of the thresho...

AI Technical Summary

Problems solved by technology

[0003] 1. The robustness of the univariate recognizer is poor. For the wake recognition of different flow parameters, the threshold needs to be adjusted artificially. However, the selection of the threshold depends heavily on human subjective cognition, so it cannot objectively reflect the real situation.

[0004] 2. Turbulence is a multi-scale physical phenomenon, and a variable can only reflect the flow characteristics at a certain characteristic scale, so the flow state cannot be fully identified, and the wake of the wind turbine cannot be accurately identified

Images