Intelligence based on neuro-control

In this paper, we consider a neuro-approach to motion control based on a PID (proportional plus integral plus derivative) control method to stabilize a pendulum. This controller requires the determination of PID control gains, but it is difficult to select the best gains theoretically. Here, we propose a method to use neural networks to tune the PID gains so that human operators tune the gains adaptively according to the environmental conditions and systems specification. As an example of self-tuning PID gains, we consider the torque control of an electric vehicle.