Distributed Multi -Agent System Based Load Frequency Control for Multi- Area Power System in Smart G

Abstract

This paper presents an intelligent controller for “Load Frequency Control (LFC)” application in “smart grid (SG)” environment having changes in communication topology via multi agent system (MAS) technology. In study, network induced effects, time delay and change in communication topology (CT) have been addressed to examine the system performance in closed loop. An event triggered control method is used to reduce the communication burden in a network. An intelligent controller based on reinforcement learning consists of two levels estimator agent and controller agent in each multi-area system. Particle swarm optimization (PSO) is used to tune the controller parameters. Further proposed control strategy and system architecture as MAS for LFC in smart grid is analyzed in detail, verified for various load conditions and different network configurations. In addition, mean square error of power system states with CT is also analyzed.

Existing System

Event Triggering Sampling Scheme.

Proposed System

In this study, an intelligent controller for LFC in SG with CT changes using multi-agent system (MAS) technique is presented. This paper deals with an intelligent controller for LFC problem in a power system that has the frequency bias coefficient is as one of its functionalities and its implementation issues in smart grid are discussed in detail. An intelligent controller consists two agents which communicates with each other to provide complete information of the system. The first agent is the estimator agent which based on frequency bias coefficients provides the area control error (ACE) signal, and the second agent is controller agent provides signal according to ACE signal obtained from estimator agents, incorporating reinforcement learning RL algorithm to compensate the power imbalance between generations against the load demand. Further PSO technique is used to tune the controller gains. The network is modelled considering different network configurations, i.e. change in CT using MAS technique. Physically adopting such scheme indicates the performance of communication system in maintaining the exchange of information within the power system. The major contribution towards an intelligent control design for LFC problem in SG, considering change in communication topology are highlighted below: To improve the dynamic performance A MARL technique is applied in multi-area power system as CT changes in smart grid. A time varying communication matrix is included in power system to model the CT changes in smart grid. The LFC for power system with event triggered control method is used to reduce the amount of communication required for data transmission and hence improve the dynamic performance of system.

Architecture

Flow chart of modelling of multi area power system with MAS

Proposed multi agent model for ith area of thermal power system