Optimum Design of Controller Parameters for Automatic Generation Control Employing Hybrid Statistically Tracked Particle Swarm Optimization Algorithm

Abstract

In the electrical energy system, automatic generation control (AGC) is a mechanism for changing the energy output of number of generators at different power plants in response to the load variances. Since the power grid needs generation and load to be carefully balanced from time to time frequent changes to the generator output are required. It is necessary to develop an efficient control system for AGC and mitigate to such type of problem, and this paper proposed an optimum design of controller parameters for automatic generation control with employing particle swarm optimization (PSO) and statistical tracked particle swarm optimization (STPSO). The proposed algorithm uses the properties of statistical parameters to accelerate the velocity of the particle to search the best possible value with faster and improved convergence speed. The effectiveness of the proposed algorithm for frequency control of the system is analyzed by comparing its performance with other evolutionary algorithms such as ant colonies, chaotic PSO, cuckoo search, bacteria foraging optimization (BFO), and teaching learning-based optimization (TLBO). The result shows that STPSO is much better than other algorithms when the frequency control of AGC system is observed.

Appendix

$$ \begin{gathered} {\text{System parameter for four area AGC system }}[2,\bar{4}] \hfill \\ k_{g,1} = k_{g,2} = k_{g,3} = k_{g,4} = 1, \hfill \\ T_{g,1} = T_{g,2} = T_{g,3} = T_{g,4} = 0.08, \hfill \\ k_{t,1} = k_{t,2} = k_{t,3} = k_{t,4} = 1, \hfill \\ T_{t,1} = T_{t,2} = T_{t,3} = T_{t,4} = 0.3, \hfill \\ K_{r,1} = K_{r,2} = K_{r,3} = K_{r,4} = 0.5, \hfill \\ T_{r,1} = T_{r,2} = T_{r,3} = T_{r,4} = 10, \hfill \\ K_{p,1} = K_{p,2} = K_{p,3} = K_{p,4} = 120, \hfill \\ T_{p,1} = T_{p,2} = T_{p,3} = T_{p,4} = 20, \hfill \\ \beta _1 = \beta _2 = \beta _3 = \beta _4 = 0.872, \hfill \\ R_1 = R_2 = R_3 = R_4 = 2.2568, \hfill \\ 2\pi T_1 = 2\pi T_2 = 2\pi T_3 = 2\pi T_4 = 0.05. \hfill \\ \end{gathered}$$

Parameters for PSO algorithm:

Initial population = 30; Maximum iteration = 100; W_max = 0.9; W_min = 0.4; C₁ = C₂ = 1.5.

Parameters for CPSO algorithm:

Initial population = 30; Maximum iteration = 100; W_min = 0.1

Parameters for DE algorithm:

Initial population = 30; Maximum iteration = 100; F = 0.5;

R = 0.98;

Parameters for B B-BC algorithm:

Initial population = 30; Maximum iteration = 100; $\beta$ = 0.7;

$\alpha$ = 0.5

Parameters for TLBO algorithm:

Initial population = 30; Maximum iteration = 100.