P. Mariño, C. Sigüenza, F. Poza, F. Machado, and F. Vásquez (Spain)
Sensors, dynamic reconfiguration, data warehouse, scala bility, filters.
Power transformers' failures carry great costs to electric companies since they need resources to recover from them and to perform periodical maintenance. To avoid this prob lem in four working 40 MVA transformers, the authors have implemented the measurement system of a failure prediction tool, that is the basis of a predictive mainte nance infrastructure. The prediction models obtain their inputs from sensors, whose values must be previously con ditioned, sampled and filtered, since the forecasting algo rithms need clean data to work properly. Applying Data Warehouse (DW) techniques, the models have been pro vided with an abstraction of sensors the authors have called Virtual Cards (VC). By means of these virtual devices, models have access to clean data, both fresh and historic, from the set of sensors they need. Besides, several charac teristics of the data flow coming from the VCs, such as the sample rate or the set of sensors itself, can be dynamically reconfigured. A replication scheme was implemented to al low the distribution of demanding processing tasks and the remote management of the prediction applications. VCs and the modular architecture proposed make the system versatile and scalable, respectively. The system is currently installed and working in four power distribution plants of a Spanish electric company.
Important Links:
Go Back
