基于数字孪生的风机实时载荷预估研究

为推动风电产业技术的持续进步,改善风机控制系统性能,利用数字孪生方式,实现风机实时载荷的预估.利用风机原有的高可靠采集信号作为输入,搭建风电机组回归模型,建立系统权重矩阵,并进行线性插值使模型适应整个发电工况,完成了风电机组载荷实时预估.载荷预估值可替代目前的应变片载荷测试系统.载荷预估经评估计算精确度在 ９６.００％ 以上.预估载荷在风机塔架推力消减实时控制系统中的应用表明,减载效果显著.

Real Time Load Prediction of Wind Turbine Based on Digital Twin

In order to promote the continuous progress of wind power industry technology and improve the performance of wind turbine contro lsystem , the real-time load prediction of wind turbine is realized by using digital twin mode．Using the original high reliable acquisition signal of wind turbine as input , the regression model of wind turbine is built , the system weight matrix is established , and the linear interpolation is carried out to make the model adapt to the whole power generation condition , and the real-time load prediction of wind turbine is completed．The load prediction value can replace the current strain gauge load test system．The accuracy of load prediction is more than ９６.００％．The application of load prediction in real-time control system of wind turbine tower thrust reduction shows that the effect of load reduction is remarkable．