Active Disturbance Rejection Control For A Class Of Non-affine Nonlinear Systems Based On Event-Triggered Mechanism

Many modern civil engineering and national defense have non-affine nonlinear forms,such as fan-speed system and helicopter control system.These systems are susceptible to model uncertainties or external disturbances in actual working environment.Meanwhile,with the expansion of engineering application scale,the existing network communication resources are relatively limited.Aiming at the above problems,the study of active disturbance rejection control for a class of non-affine nonlinear systems based on event-triggered mechanism has theoretical value and practical significance.The main work is summarized in the following three aspects:First,the event-triggered adaptive fault-tolerant problem is studied for a class of uncertain non-affine nonlinear single input single output systems.Non-affine systems are transformed into affine systems by means of implicit function theorem and mean value theorem.Extended state observers and tracking differentiators are utilized to estimate the unknown dynamics and to simplify the high order derivative operation of the virtual control law,respectively.An adaptive law is designed to compensate the loss of effectiveness faults of the actuator in real time.A projection algorithm is introduced to estimate the parameters of time-varying bias faults online.In order to save communication resources,a time-varying threshold event-triggered mechanism based on tracking error and adaptive parameters is proposed.The tracking error is steered to converge to a bounded set with the help of a predefined performance function,and its transient performance is improved despite of faults.Furthermore,considering actuator faults and input hysteresis simultaneously,an event-triggered adaptive fault-tolerant control scheme is proposed.Due to the design conflict between the input hysteresis of the actuator and an event-triggered mechanism,a second-order filter is introduced in the controller to the actuator channel to obtain smooth and differentiable control signals,and then an adaptive fault-tolerant control scheme is designed to offset the influence of input hysteresis.A time-varying threshold based on hyperbolic tangent function is proposed for an event-triggered mechanism.According to the bounded property of hyperbolic tangent function,the threshold of control signal can be kept within a fixed boundary,thus limiting the size of triggered threshold.Therefore,the problems of jump and jitter of control signal are improved.The stability of closed-loop system is analyzed by using the input-to-state stability theorem and Zeno behavior is excluded.Second,for a class of uncertain non-affine nonlinear multi-agent systems,an active disturbance rejection output consensus tracking strategy based on practical-fixed-time and event-triggered mechanism is proposed.In the framework of backstepping method,non-affine nonlinear multi-agent systems are transformed into affine nonlinear multi-agent systems by combining active disturbance rejection control,implicit function theorem and mean value theorem.A distributed estimator is constructed to estimate the leader’s signal by utilizing the neighbor follower output information after triggered.The practical fixed-time technology provides an upper bound on the settling time,which is independent of the initial conditions.Extended state observers and tracking differentiators are employed to estimate the unknown dynamics and the derivatives of the virtual control law of each follower.The multi-agent system input-to-state practical stability is achieved by means of the proposed consensus tracking control strategy.Meanwhile,Zeno behavior is excluded.Third,an active disturbance rejection state consensus control scheme based on dual-terminal event-triggered mechanism is proposed for a class of uncertain stochastic non-affine nonlinear multi-agent systems.In order to improve the utilization rate of communication resources,a dual-terminal event-triggered mechanism is proposed.The dual-terminal event-triggered mechanism is based on the input of extended state observer and the control input signal.The event-triggered extended state observer is used to estimate the internal uncertainties and the external bounded random disturbances.An extended state observer with time-varying gain is proposed.The time-varying gain is designed as a piecewise function,the first part is a combination of constant and exponential function,and the second part is always a constant.From the form of the time-varying gain function construction,an additional constant is added to the improved time-varying gain function,which makes the initial estimation error become the product term of the adjustable parameter and the initial observation error.Compared to constant gain,time-varying gain increases the adjustability of the gain.The event-triggered extended state observer with time-varying gain effectively suppress the peaking phenomenon near the initial time caused by the constant gain method.Through the proposed state consensus tracking strategy,it is proved that the state consensus error and the observation error of the event-triggered extended state observer are bounded,and Zeno behavior is excluded.