An extended nonlinear disturbance observer (ENDOB)-based fuzzy sliding mode control approach is proposed for single-input and single-output systems with matched and mismatched uncertainties/disturbances, and it is extended to the general nth order systems. By integrating the mismatched disturbance estimation and its derivative into the sliding mode surface, an ENDOB-based SMC method is developed for these systems. A fuzzy control system is designed to adjust the gain of the sliding mode dynamically. Compared with the nominal SMC method, the proposed method has a stronger anti-disturbance ability in the presence of matched and mismatched uncertainties/disturbances. It can ensure a satisfactory system performance and reduce the chattering. The stability of the system is proved by using Lyapunov theory. The final simulation results of two examples are provided to verify the effectiveness of the proposed control method. INDEX TERMS Matched and mismatched, fuzzy sliding mode control (FSMC), extended nonlinear disturbance observer (ENDOB), permanent magnet synchronous motor (PMSM).

The surface permanent magnet synchronous motor (SPMSM) speed regulation system is easily affected by the inner parameter perturbation and the external load disturbance at run time. To solve this problem, the H ∞ robust control strategy is proposed in this paper. First, given the systematic uncertainties, the H ∞ robust current controller based on the Hamilton-Jacobi Inequality is designed to ensure the robustness of current control under the SPMSM nominal mathematical model. This model is expressed as the port-controlled Hamiltonian with the dissipation form; second, the linear matrix inequality-based H ∞ sliding surface and the sliding control law are designed under the extended state space expression of the SPMSM motion equation. Thereby, the robust H ∞ sliding mode speed controller is acquired, thus realizing the robustness of speed control and improving the dynamic characteristics of the system. Finally, the effectiveness and availability of the proposed control strategy are verified by the hardware-in-the-loop simulation experiment.INDEX TERMS Surface permanent magnet synchronous motor (SPMSM), H ∞ robust control, sliding mode control, linear matrix inequality (LMI), Hamilton-Jacobi Inequality (HJI).

Considering that the nominal dynamics model or numerous parameters of robotics are usually unsuitable for real applications, a model-free adaptive sliding mode control with an adjustable funnel boundary is proposed for robot manipulators with uncertainties. First, time delay estimation (TDE) technique is utilized to estimate the unknown dynamics of the control system, which ensures an attractive model-free advantage. Furthermore, a modified funnel function is introduced to transform the trajectory tracking error fall within an adjustable funnel boundary strictly. Then, based on the transformed error variable, a novel funnel nonsingular fast terminal sliding mode control scheme is developed to enhance the transient and steady-state tracking performance of the closed-loop control system. To cope with the TDE error, an adaptive update method is designed with only one adaptive parameter, which is adaptively tuned according to the sliding surface. Finally, the simulation and experimental results are presented to illustrate the superiority and high-precision tracking performance of the proposed approach.

Robust and precise control of robot systems are still challenging problems due to the existence of uncertainties and backlash hysteresis. To deal with the problems, an adaptive neural sliding mode control with prescribed performance is proposed for robotic manipulators. A finite-time nonsingular terminal sliding mode control combined with a new prescribed performance function (PPF) is developed to guarantee the transient and steady-state performance of the closed-loop system. Based on the sliding mode variable, an adaptive law is presented to effectively estimate the bound of system uncertainties where the prior knowledge of uncertainties is not needed. To approximate nonlinear function and unknown dynamics, the Gaussian radial basis function neural networks(RBFNNs) is introduced to compensate the lumped nonlinearities. All signals of the closed-loop system are proven to be uniformly ultimately bounded (UUB) by Lyapunov analysis. Finally, comparative simulations are conducted to illustrate superiority and reliability of the proposed control strategy.