“…Among the control approaches using the input-output model are adaptive control [10], [16], H ∞ control [17], variable structure control [18], and neural network control [11]. For the dynamic models constructed from static models, sliding-mode control [19], PID control [20], [21], saturated adaptive robust control [22], active disturbance rejection control [23], and nonlinear predictive control [24] have been proposed for different kinds of robotic manipulators with PMs in different configurations. For most of the aforementioned works, due to the lack of rigorous stability analysis, the performance of the closed-loop system is not theoretically guaranteed.…”