DSP-Based PSO Implementation for Online Optimization of Power System Stabilizers

Abstract

Real-time implementations of controllers require optimization algorithms which can be performed quickly. In this paper, a Digital Signal Processor (DSP) implementation of Particle Swarm Optimization (PSO) is presented. PSO is used to optimize the parameters of two stablizers used in a power system. The controllers and PSO are both implemented on a single DSP in a hardware-in-loop configuration. Results showing the performance and feasibility for real-time implementations of PSO are presented.

An LMI-SSI model based PSS design approach for a multi-machine power system

“…Through experimentation on all the 16 generators, it is found that the satisfactory system order for each generator is less than 12. Taking G 10 as an examp le, the identified system matrices are: LMI is a mathematical tool that is useful in poleplacement and H2/Hcn control [8]. Using the controlled system matrices, linear controller can be easi ly designed.…”

“…Though there are abundant LMI tools to solve those robust control problems, accurate linearized model is hard to obtain for an interconnected power system, calling for data-driven approaches. Heuristic optimization methods such as Particle Swarm Optimization (PSO) are adopted to tune PSSs [8]; yet its online effectiveness on an actual system is yet to be verified.…”

Damping inter-area oscillations using virtual generator based power system stabilizer

“…In literature, there exist a lot of papers on supplementary damping control; and a few relevant ones are discussed below [2][3][4][5][6][7][8][9][10][11]. An early approach is based on the analysis of power system oscillation modes using Prony Analysis (PA); local PSSs are designed as filters to get rid of specific modes [2].…”

“…An early approach is based on the analysis of power system oscillation modes using Prony Analysis (PA); local PSSs are designed as filters to get rid of specific modes [2]. A real-time implementation of a heuristic approach for online PSS tuning is presented in [3]. A coordinated design of local PSSs and AVRs is presented in [4].…”