欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
以永磁同步风电机组的实测数据样本为例,研究基于遗传算法的支持向量回归（GA-SVR）的噪声预测。开展了永磁同步风电机组的空载、负载及风速变化实验;对不同运行工况下的振动及噪声实测数据进行分析处理,建立了数据样本;以发电机主轴的径向与轴向、齿轮箱高速轴和低速轴径向与轴向的振动值作为输入变量,机组的噪声值作为输出变量,建立了GA-SVR预测模型;通过数据样本训练,验证了该预测模型。研究结果表明,应用GA-SVR预测模型对机组噪声进行预测,能够较精确地获得噪声波动趋势及预测值。将该模型用于风电机组的噪声预测是可行的。
英文摘要：
Taking the measured data of permanent magnet synchronous wind turbines as an example, a noise prediction is researched based on the genetic algorithm support vector regression （GA-SVR）. The permanent magnet synchronous wind turbines are respectively tested without a load, with load under variable wind speed conditions. The vibration data and noise data under different operation conditions are processed, the data samples are established. A GA-SVR prediction model is created, in which the vibration of generator shaft in radial and axial direction, the vibration of gearbox high-speed shaft and low-speed shaft in radial and axial direction are chosen as input variables, and the noise as the output variable. The data sample training validates the prediction model. The results showed that the noise prediction with GA-SVR can get more accurate noise fluctuation trend and predictive value. The application of the prediction model on wind turbines has practical feasibility.
同期刊论文项目
同项目期刊论文
期刊信息
位置：