A survey of artificial neural network applications in flow analysis of energy networks

Technical Report · Tue Nov 01 04:00:00 EST 1994

OSTI ID:10193191

Madhok, V

This report presents the results of a literature survey carried out to analyze the work done so far using artificial neural networks (ANNs) in tasks of flow assessment in networked systems. The survey revealed that current implementations developed so far may be classified as pattern analysis and optimization, with a parallel grouping for the types of ANNs used: the feedforward network trained using error backpropagation and the Hopfield network. Pattern analysis has been used in developing ANNs for prediction and system security monitoring, with the Kohonen feature map, as well as the backpropagation type of network, being useful in the latter application. Results have been promising, showing ANNs` potential for use in problems of load flow analysis in dense energy networks or in system monitoring in hazardous situations. Problems of the nature of resource scheduling, optimal path search, and so forth, have been handled in general by methods based on Hopfield dynamics. By and large, however, large-scale commercial applications of ANNs are rare. Industry has failed to exploit the technology fully, probably because of a lack of funding and the necessary multifaceted scientific workforce.

• Research Organization: Los Alamos National Lab., NM (United States)
• Sponsoring Organization: Department of Transportation, Washington, DC (United States)
• DOE Contract Number: W-7405-ENG-36
• OSTI ID: 10193191
• Report Number(s): LA--12855-MS; ON: DE95002452
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
990200
DATA-FLOW PROCESSING
INDUSTRY
MATHEMATICS AND COMPUTERS
NETWORK ANALYSIS
NEURAL NETWORKS
NONLINEAR PROBLEMS
POWER SYSTEMS