中文摘要：

为解决非线性励磁控制律中出现的不直接可测变量的反馈问题，首先通过分析雅克比矩阵的秩，在理论上确定了不直接可测变量与直接可测变量及其导数之间函数关系的存在。然后，为克服该函数解析表达式难以获得的困难，将理论上存在的该函数代入非线性励磁控制律中，得到一个复合非线性控制律，进而采用神经网络来逼近该复合控制律。这样，最终得到的励磁控制器仅由直接可测变量及其低阶导数的反馈来实现。仿真分析验证了该方法的有效性。

英文摘要：

A method for directly immeasurable variables feedback in nonlinear excitation controller is proposed. Firstly, according to the rank of Jacobian, it is shown that there exists a nonlinear function relationship between the directly immeasurable variables and the measurable ones and/or their derivates in theory . Then, to overcome the difficulty in giving the above nonlinear function in an analytic way, this paper presents a compound nonlinear control law by applying the theoretical function into the excitation controller and then implements the compound control law by using artificial neural network. Thus, the obtained ANN nonlinear excitation controller only requires the feedback of the directly measurable variables and/or their derivatives. The simulation results show the validity of the method.