Load Identification of Non-intrusive Load-monitoring System Based on Time-frequency Analysis and PSO-SVM

Abstract

This paper presents a method for non-intrusive load monitoring (NILM) identification which is based on transient analysis and steady-state harmonic analysis. Each appliance has its own characteristics which results in a unique magnitude when it is switched on and have unique frequency spectrum in the steady-state. So based upon these analysis, frequency spectrum is used in combination with time domain analysis to identify loads. And the proposed NILM system employs the Particle Swarm Optimization (PSO) Algorithm with the Support Vector Machines (SVM) to perform load classification. The identification results confirm that the proposed system is suitable for identifying different loads.