Sistema de suporte a decisão para analise e previsão de carga por barramento / Decision support system to the analysis and bus load forecasting

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

In a power operation system there are many variables that can Influence the execution of t ho control routines. The attendance of these variables is very important to aid the processes in the safe operation of the system. In the electric system the bus load level is an important variable in operation planning step. Accurate Information on the load level In each bus provide a better control In the power flow, safe and security analysis, operation programming, planning and others. Thus, this work presents two contributions to the bus load forecasting area. The first contribution is the development of a bus load forecast methodology that executes and provides the forecast results for several buses in an aggregate way. The second contribution is a computational system for bus load forecasting named SysPrev - Support System for Bus Load Analysis and Forecasting- developed to facilitate the bus load forecasting process. The aggregate bus load forecasting model is composed by two steps: In the first phase a clustering algorithm is used to Identify buses clusters with similar dally load profile and In the second phase Is proposed an aggregate structure for to foresee each bus using a conventional prediction model The SysPrev is a software composed of a data manager subsystem, a model subsystem and graphical interfaces subsystem.The data manager subsystem is composed of a database that contains bus load time series and other relevant Information. The model subsystem is composed of forecasting and clustering techniques. To facilitate the user s interaction, the graphical interface has packages of graphic windows that provide a friendly environment facilitating the Integration of tlie SysPrev and their users. An excellent characteristic of the SysPrev is its construction based on components plug-and-play that allows the Insert of new functionalities With this architecture, each component Is produced on an independent way. could be Inserted In the software without alterations in the main system This factor Is Important, because It supplies the user with the option of customizing and creatlng new analysis tools, facilitating the updating and the maintenance of the system. The SysPrev was applied to a real power system data composed of 73 buses. The easiness of creating multiple sceneries made possiblethe fast understanding of the data providing good results In the forecasts, making possible larger precision on the power system operation step

ASSUNTO(S)

electric power systems sistemas de energia eletrica - distribuição de carga cluster analysis inteligencia artificial artificial intelligence decision support systems time series analysis sistemas de suporte a decisão analise de series temporais analise de cluster

ACESSO AO ARTIGO

Documentos Relacionados