Robot trajectory tracking control with an improved neural network adaptive sliding mode control

作者：Liu Tianhua^*; Yin Shoulin; Li Hang

来源：Journal of Computational Information Systems, 2015, 11(12): 4401-4411.

DOI：10.12733/jcis14666

• Approximation errors
• Hidden-layer units
• Network adaptive
• Network structures
• Nonlinear mappings
• Particle swarm optimization algorithm
• Trajectory tracking
• Unknown nonlinear systems

In order to improve the performance of the robot trajectory tracking control, this paper proposes an improved neural network adaptive sliding mode control method based on the neural network sliding mode control. This method takes neural network as a controller and uses its nonlinear mapping ability to approximate various unknown nonlinear systems. At the same time, it uses robust control law to eliminate the approximation error. Taking hidden layer units and network structure parameter into consideration, this paper's optimum aim is to reduce chattering. It optimizes network structure parameter by particle swarm optimization algorithm. Finally, it conducts simulation experiments under the Matlab/Simulink platform and makes a comparison with other control methods. Simulation results show that the improved control system has a good control accuracy and robustness based on the improved method. It also effectively weakens the chattering and reduces the effect of hidden layer units and network structure parameter on neural network mapping.

• 出版日期2015