Particle Swarm Optimization of a universal inverter

Universal inverters will be the future interface of distributed generation to power systems based on active distribution grid where island operation should be allowed. Universal inverters should be able to work both connected to the grid both in island mode. A self-commissioning process based on advanced optimization algorithms should tune the parameters of the controllers to manage the different conditions and the transition among them. In this paper it is proposed to use the Particle Swarm Optimization that is based on social behavior. The method does not use gradients and the population update is based on simple dynamic trajectories.