Surveillance of power plants - Värmforsk

Together with Mälerdalens Högskola, Västerås, Sweden, The Machine Intelligence Group was part of the project ”Värmforsk” (2003-04). Together with ELSAM we had a project on establishing a monitoring system giving early warnings when a production plant may be heading into serious production disturbances. A special property of this domain type is that there is (usually) a lack of both knowledge and data concerning possible faults, and it is therefore infeasible to learn/construct a standard classification model for doing fault detection. As an alternative, fault detection can be based on a model of normal system operation only. This means that it does not rely on information about the possible faults, and discrepancies between sensor readings and the model are then taken as an indication of a possible fault.