论文部分内容阅读
建立人工神经网络、径向基函数网络和支持向量回归机三种近似模型,结合蒙特卡洛方法与表征粗糙度参数随机特性的概率模型对风力机翼型气动性能进行不确定性分析。结果表明,支持向量回归机具有最佳预测精度。对于风力机翼型FX 63-137,最大升力系数对吸力面前缘粗糙度的敏感性明显高于压力面;对于吸力面或压力面,前缘粗糙带厚度对最大升力系数的影响稍大于其覆盖长度的影响。研究工作为风力机翼型的鲁棒性设计优化奠定了理论基础。
Three approximate models of artificial neural network, radial basis function network and support vector regression machine are established. Combining the Monte-Carlo method with the probabilistic model which characterizes the stochastic properties of roughness parameters, the aerodynamic performance of wind turbine airfoil is analyzed indefinitely. The results show that the support vector regression machine has the best prediction accuracy. For wind-wing type FX 63-137, the maximum lift coefficient is significantly more sensitive to the leading edge roughness of the suction surface than to the pressure surface. For the suction or pressure surface, the effect of the leading edge roughness on the maximum lift coefficient is slightly greater than its coverage Effect of length. The research laid a theoretical foundation for the robust design and optimization of wind turbine airfoils.