Research On Adaptive Neural Networks Chaotification Of Robot Manipulators

Chaotification,also known as anti-control of chaos,refers to the creation of new chaos that people need by some means in a system that does not have chaotic behavior in nature.In recent years,it has been found that injecting chaotic behavior into the robot manipulator can promote its application in industrial,agricultural and household applications.The robot manipulator is a complex system with multiple inputs,multiple outputs,strong coupling and high nonlinearity.Therefore,it is difficult to achieve precise tracking control by traditional control algorithms due to uncertainty and environmental factors.In this paper,the algorithm research of anti-control of chaos combining with sliding mode control?adaptive neural networks?and DSC are carried out with the robot manipulator as research object.The paper mainly studied the following contents:1.For the robot manipulator with unknown parameters and bounded disturbance,a robust anti-control of chaos algorithm is designed by combining the sliding mode control with the adaptive neural networks(NNs).The terminal sliding mode control is used to suppress the unknown bounded disturbance,and the appropriate radial basis function neural networks(RBF NNs)is constructed to approximate the unknown nonlinear function of the system.The ultimately uniformly bounded(UUB)of all the signals in the closed-loop system is proved through Lyapunov stability theory and the effectiveness of the algorithm is verified by simulation experiment,which shows that not only the chaotic attractors are successfully observed,but also accurate speed tracking is realized.2.For the electric-drive manipulator with unknown parameters and unknown disturbance,an anti-control of chaos scheme combining DSC and RBF NNs is proposed in this paper,which solves the high nonlinearity caused by the introduction of the actuator during modeling,as well as avoids the "differential explosion" problem in traditional back-stepping method.In the second and third steps of the back-stepping method,the first-order low-pass filter is used to estimate the derivative of the virtual control variable.At the same time,RBF NNs is employed to approximate the unknown nonlinear function of the system,and the appropriate Lyapunov function is selected to prove the stability of the system.Finally,the effectiveness of the proposed algorithm is verified by simulation.3.Aiming at the uncertain robot manipulator with actuator saturation,an adaptive neural network controller combined with DSC is designed to realize its anti-control of chaos.The unknown dynamics of the system are approximated by RBF NNs,a suitable auxiliary system is constructed to solve the problem of input saturation,and the ultimately uniformly bounded of all the signals in the closed-loop system are proved by Lyapunov stability theorem.Then,the feasibility and effectiveness of the proposed method are verified by simulation with a two-link and a three-link electric-drive robot manipulator respectively.Finally,the content of this paper is summarized,and the research on related fields in the future is prospected.