MODEL-FOLLOWING CONTROL OF A VARIABLE INERTIA POSITIONER

The work presented here shows the results of a simulation made on a tw o-degree-of-freedom plant (robot joint) driven by an ideal model. The reference input is applied only to the model, and control effort on th e plant is generated by observing the deviations of its states from th e ideal norms issued by the model. The model thus constrains the plant to follow it. The design of this system is based primarily on the con struction of the model, which should have the same essential features but without the variable parameters found in the plant in question. It has been shown that satisfactory following of the model is possible w ith this approach, even under wide variations in the plant inertia (i. e., positive and negative rates of variation, as well as fluctuations in the load). The performance robustness of this arrangement is remark able, and its implementation is not complex.