Neural Network With Sliding Mode Control For Robotic Manipulator

With the development of science and the need of human, robot technologyhas been used in various fields widely, and has been valued in the field ofindustry and science. The key of robots is its control system, the robot is anonlinear and uncertainty system. Recently, the intelligence control of robotshas been valued by scientists of all kinds of countries. This paper will put neuralnetwork control and sliding mode variable structure control together, used tocontrol the position of joint and the trajectory tracking.First of all, It describes the robot manipulators system’s composition andanalyses the dynamic model in detail, and points out that the system has notstable and nonlinear characteristics. Then, it sets up manipulator systemdynamic model by S-the function in the MATLAB environment, resolve theproblem that mechanical arm complex system modeling difficultly.Secondly, according to the sliding mode variable structure control systemparameters has nothing to do with the characteristics of the disturbance, designsSMC using equivalent switch control, and Lyapunov stability theorem is used todecide sliding mode switch gain, ensure robot manipulator system stability. Inthis paper, because of the uncertainty of the robot manipulator system, maketraditional sliding mode variable structure control need a great deal of switchgain, and cause large chattering.Finally, a kind of controller based on the radial basis function (RBF) neuralnetwork of sliding mode variable structure control has been designed in thispaper. It ensures that the system is stable using sliding mode variable structurecontroller, and RBF control approach to mechanical arm dynamic system,through the learning ability overcome system to reduce uncertainty switch gain,and reduce the chattering. This paper using sliding mode variable structure, RBF neural network sliding mode variable structure control of mechanical armsystem, and the simulation results show that the RBF neural network slidingmode variable structure control has better stability and anti-jamming.