Design of PID Controllers using the Particle Swarm Algorithm

P.B. de Moura Oliveira, J. Boaventura Cunha, and J. Paulo Coelho (Portugal)


PID control, Particle Swarm Optimisation, Neural Networks.


Particle swarm optimisation is proposed as an alternative technique to the controller design for single-input single output systems. The evolutionary scheme proposed is applied to design the most popular controller within the process control industry: the proportional, integral and derivative (PID) controller. The particle swarm optimisation algorithm is applied to the design of both linear and non-linear PID controllers. In the non-linear control configuration a feedforward neural network is deployed to generate a non-linear gain that acts as a filter to the tracking error inputted to the PID controller. This evolutionary approach is illustrated by simulation examples in which the PID is used to control a delayed second order system.

Important Links:

Go Back