| Abstract |
In the conventional AC position servo drives there exist many drawback, such as the structure of the system is complicated, and its control performances are significantly influenced by various uncertainties. To overcome these demerits, a model algorithmic control (MAC) scheme for AC position servo drives is proposed by synthesizing the advantages of MAC in the model prediction, rolling optimization and feedback tuning in real time on-line. The first order ARMA-model of the AC servo drives is used as the parameter-model in the predictive control, which greatly simplifies the control algorithm that meets the requirement of the real-time control for the split-second dynamics of the motor. Moreover, the optimum control of the servo drives can be insured by the dynamic optimization and feedback tuning based on the predictive model in real-time. The analysis of the system’s stability and robustness shows that the control system has very good robustness in any cases. Experimental results verify the excellent dynamics and static performances of the proposed AC position servo drive. |
