Motion and balance neural control of inverted pendulums with nonlinear friction and disturbance

In this paper, a motion and balance control scheme is introduced for inverted pendulums using artificial neural network (ANN). The control strategy uses a trade-off strategy to achieve motion tracking and balance control simultaneously with a single controller. Unlike other neural control strategies, no offline learning or a priori system´s dynamics knowledge is required. The controller is trained online to learn the nonlinear inverted pendulum system´s dynamics. Simulation results for different situations highlight the performance of the proposed controller in compensating for friction nonlinearities and for external disturbance. Furthermore, ANNs´ inherent parallelism makes them a good candidate for real-time implementation.