Load Model Identification Through a Hybrid Approach

Author(s)

Jollies, Gabriel J. N. ; Lemes, Francisco R. ; Cari, Elmer P. T.

• Document type: Journal article
• Source: 2019 IEEE CANADIAN CONFERENCE OF ELECTRICAL AND COMPUTER ENGINEERING, CCECE; v. N/A, p. 4-pg., 2019-01-01.

Abstract

This paper proposes a hybrid approach to estimate the parameters of Z-IM Load Model from on-line measurements obtained from tap changes of transformers. For this aim, the model equations were linearized to use the information provided by small disturbances. The approach proposed applies a heuristic (Mean Variance Mapping Optimization) and a nonlinear (Trajectory Sensitivity) methods in cascade to optimize convergence. Simulation results shows the adequacy of the hybrid method for load model. The entire process took, in average, 24 seconds to converge. All the codes used for this work were written in Python 2.7. (AU)

• FAPESP's process: 17/09208-4 - Hibrid Method for parameter estimation of photovoltaic power plants
• Grantee: Elmer Pablo Tito Cari
• Support Opportunities: Regular Research Grants