Hibrid Method for parameter estimation of photovoltaic power plants

• Grant number: 17/09208-4
• Support Opportunities: Regular Research Grants
• Start date: August 01, 2017
• End date: July 31, 2019
• Field of knowledge: Engineering - Electrical Engineering - Power Systems
• Principal Investigator: Elmer Pablo Tito Cari
• Grantee: Elmer Pablo Tito Cari
• Host Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated research grant(s): 17/50389-2 - Optimization and prediction modeling of solar module considering environmental parameters, AP.R SPRINT

Abstract

Phovoltaic generation is increasing around the world as an important alternative of energy production. Brazil has a great solar potential and, thus, the government has supported for the inclusion of this kind of energy by creating laws to facilitate the approval for the distributed utilities. In addition, there are also supports by agencies such as Development National Bank. Therefore, studies to foresee the impact of inclusion of this type of energy in the network become necessary. For this goal the representation of those photovoltaic generators by equivalent model and the estimation of their parameter become necessary. Usually manufactures of photovoltaic modules provide some information to obtain model parameters of those panels using standard test conditions of temperature, irradiance, and wind conditions. However, in real situation, rarely those conditions are similar of those where the photovoltaic modules are installed. Thus, differences between the outputs of simulation model and real measurements can appear that make a power forecast unsuccessful. In this research a hybrid method for parameter estimation of equivalent photovoltaic power plant is proposed using real measuremets. The hybrid method is based on Mean Variance Mapping Optimization and Trajectory Sensitivity Methods. This hybrid approach use the advantagens of both methods heuristic and nonlinear and yields to the estimation process some special characteristic such as robutstness regarding to uncertainties in initial guess and availability of range of parameter search, robustness regarding noise and speed to converge, therefore, low computation burden.The proposed method will be used to model and validate with real measurements in a photovoltaic system installed in laboratory of electrical engineering and computing department of Sao Paulo University. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

LOPES, SOFIA M. A.; CARI, ELMER P. T.; HAJIMIRZA, SHIMA. A Comparative Analysis of Artificial Neural Networks for Photovoltaic Power Forecast Using Remotes and Local Measurements. JOURNAL OF SOLAR ENERGY ENGINEERING-TRANSACTIONS OF THE ASME, v. 144, n. 2, p. 11-pg., 2022-04-01. (17/50389-2, 17/09208-4)

LANDGRAF, TAYLON G.; CARI, ELMER P. T.; ALBERTO, LUIS F. C.. An analysis of structural and practical identifiability applied to a transient generator model. Electric Power Systems Research, v. 206, p. 11-pg., 2022-01-28. (18/20104-9, 17/09208-4)