Registration Number
0207U004085,
0105U000601
, R & D reports
Title
Neural Network Situational Feature Maps for Visualization of Modes in Electrical Energy Systems of Supplier Companies
popup.stage_title
Head
Rashkevych Yuriy Mykhaylovych,
Registration Date
26-01-2007
Organization
Lviv Polytechnic National University
popup.description2
In the final report on a theme "Neural Network Situational Feature Maps for Visualization of Modes in Electrical Energy Systems of Supplier Companies" the collection of primal problems of the analysis of the information in electrical energy industry has been defined. The setting of the task of compression and visualization of the multydimensional data describing modes of the operation of electrical energy system with the purpose of dimensionality reduction and selection of informative features has been realized; known statistical and neurocomputing methods of compression and visualization have been analyzed. The technique and program units of preprocessing intended for the analysis of archive data and for application in a real time owing to using of fast neurocomputing algorithms have been developed. The architecture of neural network program complex for compression and visualization of multydimensional data in the tasks of electrical energy industry has been created on the basis of the nonlinear analysis of principal components. Using of autoassociative feed-forward neural networks with nonlinear synapses on a basis of "a functional on set of tabular functions" paradigm with a nonterative basic training principle has been supposed. The program model includes units of training, operation and visualization with the two- and threedimensional situational feature maps. The technique of situational prediction of indices and diagnostics of electrical energy system states, intended for application in hardware-software complexes of dispatching controls of supplier companies has been created. The results of test researches of program complex on the database of JSC "Lvivoblenergo" have confirmed possibilities of visualization in two-dimensional space for data describing modes of operations of electrical energy systems with an error of reconstruction up to 1,5 %.
Product Description
popup.authors
popup.nrat_date
2020-04-02
