Conceptual

PID Neural Network Control for DC Servo Motor Speed Regulation

A speed-control method for DC servo motors in which a small feedforward neural network with proportional, integral, and derivative hidden neurons (a 3-3-1 PID-neuron network) replaces a fixed-gain PID controller and adapts its weights online. Students learn how the classical PID control law can be embedded in a neural network so its gains retune in real time to nonlinear and time-varying loads, improving settling time, steady-state error, and overshoot over a conventionally tuned PID.