Neural network adaptive sliding mode control and its application to SCARA type robot manipulator

A synergistic combination of neural networks with sliding mode control is proposed. As a result, the chattering is eliminated and error performance of SMC is improved. In such an approach, the determination of the structure of NN, i.e. number of layers, number of neurons at each layer, etc. does not come up as a problem because these are directly related to the SMC. A Lyapunov function is selected for the design of the SMC and gradient descent is used for weight adaptation of the neural network. The criterion that is minimized for gain adaptation is selected as the sum of the squares of the control signal and the sliding function. This novel approach is applied to control of a SCARA type robot manipulator and simulation results are given.