中文摘要：

针对多输入多输出（MIMO）非线性强耦合的两电机变频调速系统,对系统数学模型进行广义逆存在性分析,推导出系统的广义逆数学表达式,进一步构造神经网络广义逆系统串联在两电机系统之前,组成基于广义逆的伪线性复合系统,实现MIMO系统的线性化与解耦。在神经网络逆系统离线训练的基础上提出在线训练的控制方法,在电机的运行过程中对网络进行在线训练,不断修正网络权值,使网络适应环境的变化,增强其鲁棒性,更精确地逼近其逆系统。实验证明在用PLC作为主控制器的控制过程中,神经网络不断进行自我调整,增强了神经网络的适应性,提高了系统的稳定性和鲁棒性。

英文摘要：

Two motor variable frequency speed-regulating system is a multiple input multiple output, nonlinear, and high coupling control system. The generalized reversibility of this two motor system was testified. Consequently, a pseudo-linear system was completed by constructing a neural network generalized inverse （NNGI） system and combining it With two motor system. The generalized inverse can transform the MIMO nonlinear system into a number of single input single outple linear subsystems. In order to approach the inversion exactly in operation of the motor, the control method on-line based on artificial neural network （ANN） inverse system is proposed, in which connection value can be amended continuously on-line to make the ANN adapt to the changes of environment to strengthen its robustness. Experiment results have shown that ANN can be adjusted in the control process using PLC as the main controller. The good applicability of ANN along with the strong stability and robustness of the system can be achieved by using the proposed method.