Neural Stabilizing Controller Based on Co-evolutionary Predator-Prey Particle Swarm Optimization

In this paper, an approach based on particle swarm optimization (PSO) and Lyapunov method to construct neural stabilizing controller is presented. The procedure to learn the value of neural network is formulated as min-max problem. And the problem is solved by the co-evolutionary predator-prey PSO which we newly propose. The PSO is able to generate an optimal set of parameters for neural controller. And then, the proposed neural controller can be satisfied the Lyapunov stability condition. The proposed method is validated through numerical simulations with power system stabilizing control problem comparing to the conventional control method.