Reduced Order Approximation and Performance Analysis of Renewable Energy Systems Using PSO and Its Variants

Power systems are large-scale, nonlinear randomly fluctuating operating conditions results in higher-order transfer function models makes it difficult to analyze, comprehend, and control design. Hence, a suitable simplified model of the same type is necessary for improved control design, performance analysis, and understanding. There exist different conventional and metaheuristic search based approaches in the literature for the model order reduction (MOR) of intricate systems. The inertia weight is a crucial parameter, which highly impacts the convergence and trade-off between the exploration-exploitation phases of PSO algorithm. Hence, this study is focused on application and performance analysis of the particle swarm optimization (PSO) algorithm with different inertia weights for the MOR of a linearized renewable power system network. The inertia weights considered are: (i) randomly varying inertia, (ii) linear time-varying with randomly changing inertia, (iii) nonlinear time-varying inertia, and (iv) linear time-varying inertia. The fifth order transfer function of the renewable energy system (RES) is reduced to third order. The time and frequency responses, time and frequency domain characteristics, and various system performance analysis measures are used to select the PSO algorithm with effective inertia weight.