Artificial immune system based algorithms for optimization and self-tuning control in power systems

Artificial immune systems (AIS) are new computational intelligence methods inspired by various mechanisms of the biological immune system. AIS are adaptive systems inspired by theoretical immunology and its functions, principles and models. The work depicted in this thesis centers on the applications of AIS based algorithms for optimization and self-tuning control in power systems. The optimization is carried out using an algorithm based on the clonal selection principle and the self-tuning characteristics of control for parameters are inspired by the humoral immune response of the human body. The work in this thesis is written in two papers as follows: Paper 1 - CSA is used to design multiple optimal power system stabilizers (PSS). The proper tuning of PSSs has a significant influence on its effectiveness in providing the required damping under different operating conditions and disturbances. CSA is used to determine the optimal parameters of four PSSs in a two area multi-machine power system. CSA optimized PSSs efficiently damp out the oscillations introduced in the system and its damping performance is slightly better than that of particle swarm optimization (PSO) optimized PSSs. The main contribution of CSA is that it converges faster and requires fewer computations than the standard PSO algorithm. Paper 2 - CSA is used for optimization of four benchmark functions in literature. It is then used to design an optimal synchronous machine excitation controller which reduces oscillations introduced in the terminal voltage during disturbances. Immune feedback law is used to incorporate self-tuning characteristics in the optimal controller. The self-tuning optimal excitation controller reduces overshoot and settling time of oscillations. It also reduces power losses in the field circuit, thus, enhancing its life --Abstract, page iv