Research On The Optimization Of Active Power Command Distribution In Wind Farm Based On The Condition Evaluation Of Wind Turbines

Compared with traditional thermal power generation,solar power generation,hydropower generation and nuclear power generation,wind power generation has the advantages of no pollution,low technical difficulty and little impact on the environment.Therefore,wind power generation is highly valued by all countries in the world,and the installed capacity of wind power increases year by year.With the rapid development of wind power industry,how to reduce the operation and maintenance costs of wind farms has become a rising concern in the wind power industry.On the other hand,large-scale wind power fluctuations limit the ability of the grid to absorb wind power,so wind farms are mostly in the state of power limitation.When the wind farm is in the state of power limitation,considering the health status difference between wind turbines,reasonably distributing the active power command of the wind farm and reducing the generation task of the wind turbine with poor health status are of great significance to reduce the operation and maintenance cost of the wind farm and improve the operation reliability of the wind farm.In order to solve the problem of active power command assignment in wind farms based on the health assessment of wind turbines,the main contents of this paper are as follows:(1)After using k-means algorithm to classify the operation conditions of wind turbine by wind speed,ambient temperature and output power,according to the degree that the values of temperature parameters of the main components of the wind turbine such as main shaft,gearbox and generator deviate from the values that the wind turbine is in good operation condition,the operation condition of the wind turbine is evaluated by evaluation method of wind turbine operation state based on SOM-MQE.The feasibility of this method is verified by the actual operation data of wind turbine.(2)After selecting the operation parameters of the wind turbine used to predict the output power by the ReliefF algorithm,using the power generation performance evaluation method of wind turbine based on LSTM network to predict the output power of wind turbine and evaluate the power generation performance of wind turbine by the deviation between the predicted output power and the measured output power of wind turbine.The feasibility of this method is verified by the actual operation data of wind turbine.(3)The health status of wind turbine is divided into three categories:good,general and poor,and the health status of wind turbine is evaluated by the operation status and generation performance of wind turbine.In each scheduling cycle,according to the health status of each wind turbine in the wind farm,the maximum power generation capacity and the generation instructions issued by the grid dispatching department to the wind farm,the objective function and constraints formula are established,and the genetic algorithm is used to get the active power command allocated to each wind turbine.The simulation results show that the proposed method in this paper can not only make the output power of the wind farm meet the requirements of power grid dispatching instructions,but also reduce the output power of the wind turbine in poor health status.