In this article, an adaptive nonlinear controller is developed for a class of stochastic systems, whose inputs are uncertainly nonlinear and virtual control gains (simplified as VCGs) include unknown and known items. A novel auxiliary function of boundedness and smoothness is constructed for handling the unknown items of VCGs. Aiming at the challenges of control laws without enough differentiability and uncertainties caused by deadzone and saturation of inputs, the novel control signals are proposed, which are tested on a nonsmooth system, SISO robot and MIMO quadrotor. The superiority and effectiveness of the designed control strategy are proved via the strict stability analysis and simulation comparisons.

This paper addresses the adaptive control problem for a one‐link manipulator system driven by a motor with uncertain virtual control coefficients (VCCs) and nonlinearities. The unknown VCCs are composed by the certain and uncertain items, where the first part is decided by the moment of inertia of the manipulator and another is generated by that of the load. An auxiliary variable and the Nussbaum gain technique are invoked to handle the uncertainties of VCCs and input nonlinearities. Moreover, a finite‐time prescribed performance function is introduced to design the adaptive control scheme for the manipulator system, which guarantees the tracking error converges to a fixed interval in a setting time and all signals in the manipulator system are bounded.

Aim at an air-handling unit (AHU), this article describes a direct design scheme for humidity and temperature tracking control. The nonlinear dynamic model of the AHU is treated directly by backstepping strategy, which can reserve the modeling qualifications and make better use of known information than proportion integration differentiation (PID) and other non-model-based schemes. Simultaneously, the linearization technique is not adopted, so that the model structure is not changed, which can retain the original system model restrictions to variables and does not lose useful conditions about the system model. These two points ensure the developed controller to achieve higher accuracy and better tracking performance, which is verified by stability analysis and comparisons of simulation results.