Quantitative Modeling of Effects of Extreme Low Temperature on Wind Turbine Blade Performance Based on Nonlinear Random-Coefficient Regression Model

In order to quantitatively analyze the influence of extreme low temperature on wind turbine blade performance, considering the uncertainty of its operation process, this paper proposed a quantitative modeling and analysis method based on nonlinear random-coefficient regression model for the influence of extreme low temperature on wind turbine blade performance. Compared with the traditional regression model, the random-coefficient regression model can better adapt to the complex and variable operating environment of wind turbines. Firstly, in order to quantitatively describe the effect of extreme low temperature on wind turbine blade performance, a quantitative model of extreme low temperature-blade performance was established based on the nonlinear random-coefficient regression model, and the function form of the quantitative model was determined based on the nonlinear least squares algorithm. Secondly, in order to improve the accuracy of prediction results, the expectation maximization algorithm and Bayesian updating algorithm are combined to realize the updating of random parameter of the model and the estimation of fixed parameters. Finally, the maximum blade stress is taken as the performance characteristic, and the simulation data of actual operation are used to carry out simulation tests to verify the superiority of the proposed algorithm. The results show that the maximum percentage of mean square error and mean absolute error of the proposed method are 0.0341 and 2.780% respectively, both of which are much lower than the model parameters estimation algorithm using only expectation maximization algorithm. Moreover, comparing the simulation results of the two methods, it can be seen that the proposed method has the advantages of high accuracy, fast convergence and strong robustness.