Data-mining experiments on a hydroelectric power plant

This study presents some data-mining experiments applied to electric power systems with the aim of extracting knowledge from historical data produced by the supervision, control and data acquisition system of a hydroelectric plant in Brazil. In the first experiment, statistical analysis is performed on discrete events such as Boolean events, alarms, commands, set-points and analogue quantities as electrical frequency, to display relevant aspects of the electrical system operation. Next, the results of an experiment performed on discrete events from associations describing relationships patterns among items in a database are presented. In the third experiment, a decision tree is used to reveal relationships among several analogue variables as: the relationship between the downstream water level and generated power, for example. In the fourth experiment, a decision tree is designed to detect if the hydro generator operation is violating any constraint imposed by its capability curve, also indicating which limit is extrapolated. These experiences contribute to successfully show the data-mining applicability to power systems, to improve the management of hydroelectric power plants operation, maintenance and planning, besides also contributing to establish a culture of its usage in the electrical industry.

[1]  Q. Henry Wu,et al.  Association Rule Mining-Based Dissolved Gas Analysis for Fault Diagnosis of Power Transformers , 2009, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[2]  Vipin Kumar,et al.  Introduction to Data Mining , 2022, Data Mining and Machine Learning Applications.

[3]  Ian H. Witten,et al.  Data mining: practical machine learning tools and techniques with Java implementations , 2002, SGMD.

[4]  Stuart A. Boyer.  Scada: Supervisory Control and Data Acquisition , 1993 .

[5]  L. Miclea,et al.  A data mining experiment on a SCADA system’s historical acquired data , 2008, 2008 IEEE International Conference on Automation, Quality and Testing, Robotics.

[6]  Lexin Li,et al.  Statistical Feature Selection From Massive Data in Distribution Fault Diagnosis , 2010, IEEE Transactions on Power Systems.

[7]  Jiawei Han,et al.  Data Mining: Concepts and Techniques , 2000 .

[8]  Sarasij Das,et al.  Understanding Power System Behavior through Mining Archived Operational Data , 2009 .

[9]  Goran Andersson,et al.  Voltage dependent reactive power limits for voltage stability studies , 1995 .

[10]  Marko Čepin.  Introduction to Power Systems , 2011 .

[11]  S.M. Shahrtash,et al.  A decision tree based method for fault classification in transmission lines , 2008, 2008 IEEE/PES Transmission and Distribution Conference and Exposition.

[12]  Georges Hébrail.  Practical data mining in a large utility company , 2001 .