Equivalent laser radar wind speed calculation method based on long short-term memory neural network

Abstract

The invention discloses an equivalent laser radar wind speed calculation method based on a long short-term memory neural network, and the method comprises two parts: the first part is to organize and count related operation data in a wind turbine generator equipped with laser radar equipment, and through the training of a long short-term memory neural network model, a final model structure and specific parameters are obtained; and the second part is that the trained model structure and specific parameters are adapted to a control system of a wind turbine generator which is not provided with laser radar equipment, the generator set inputs the real-time running state of the generator set into the model, and equivalent laser radar wind speed is obtained and used for the control system of the generator set, so that a specific control function is realized. The method uses the long-short-term memory neural network, generates equivalent wind speed signals which are the same as the wind speed measured by the laser radar through training, is used for control technologies such as feedforward and correction, reduces the number of laser radars used in a project, saves the cost of sensor equipment, effectively reduces the load of a unit, and improves the power generation amount.

Description

technical field [0001] The present invention relates to the technical field of wind turbines, in particular to an equivalent laser radar wind speed calculation method based on long-short-term memory neural network. Background technique [0002] It is well known in the industry that the nacelle-type LiDAR anemometer uses the principle of Doppler frequency shift to measure the wind speed and related environmental parameters in the upwind direction of the wind turbine. In recent years, it has been rapidly applied and promoted as an advanced sensing device in the wind power industry. At present, relatively mature control technologies based on lidar data mainly include feedforward control technology using lidar to collect wind speed, yaw correction based on collected wind direction data, and cabin wind speed transfer function correction. [0003] Through the laser radar to measure the front wind speed signal in the upwind direction, the wind turbine can reduce the fluctuation ran...

AI Technical Summary

Problems solved by technology

[0003] Through the laser radar to measure the front wind speed signal in the upwind direction, the wind turbine can reduce the fluctuation range of the generator and other components through control, effectively reduce the unit load, and increase the power generation. However, due to the cost limitation of the laser radar hardware equipment, the existing domestic wind farms In the project, laser radar equipment is rarely installed for each aircraft position, which limits the application of relevant control strategies and cannot be widely promoted in a large area.

Images