摘要
风机叶片多点静力加载试验存在各加载点之间的交叉耦合情况,影响试验效果。对此,研究了叶片多点加载耦合基本理论,分别设计了离散变速积分PID和神经网络变速积分PID两种解耦控制器。仿真表明,这两种解耦控制器都能够实现多点解耦,且神经网络变速积分PID解耦控制器可以适应较大范围内叶片刚度的变化。现场试验进一步证明,神经网络变速积分PID解耦控制方案在保证一定加载精度的情况下,可以适应不同刚度,不同长度的叶片,而离散变速积分PID解耦控制方案在叶片刚度相近,特定长度的情况下可以获得良好的加载精度。
The wind turbine blade multi-point static loading test has the cross coupling between the loading points,which affects the test results. In this regard,Two decoupling controllers are designed,respectively,PID of variable speed integral and neural network. Simulation results show that these two kinds of decoupling controller areable to achieve multi-point decoupling,and neural network decoupling controller can adapt to a wide range ofblade stiffness. The experiments proved that neural network decoupling control scheme with certain loading precision,can adapt to different stiffness,blade of different length,and PID of variable speed integral control schemeis similar in blade stiffness,the specific length of the case can obtain good loading precision.
出处
《装备制造技术》
2017年第11期24-27,共4页
Equipment Manufacturing Technology