DocumentCode
1596203
Title
EXDAFS-an Expert System for Dynamic Allocation of Facilities at Stations
Author
Sinha, Rahul ; Rajput, V.S. ; Asthana, R.G.S.
Author_Institution
Centre for Railway Inf. Syst., New Delhi, India
fYear
1991
Firstpage
419
Lastpage
425
Abstract
EXDAFS is a rule-based system designed to facilitate the allocation of facilities to trains arriving and/or departing from a railway station. EXDAFS develops a static allocation for all trains, referred to as the master berthing chart, based on the train schedules and their berthing requirements. However, the trains get delayed due to unavoidable circumstances causing deviations from the normal schedule, thereby necessitating a dynamic reassignment of the facilities. A technique based on transformational synthesis to control the computational complexity of the reassignment process has been devised and implemented. This requires the problem to be decomposed into subtasks, which can be solved independently and then synthesized into a final solution. The system takes decisions globally in a real-time environment. EXDAFS has an explanation capability which provides a better understanding of its reasoning directions. This, along with a visual capability and interaction through a user-friendly interface, results in increased user acceptance of the system
Keywords
computational complexity; expert systems; explanation; rail traffic; real-time systems; resource allocation; scheduling; user interfaces; EXDAFS; computational complexity; delays; dynamic facilities allocation; dynamic reassignment; explanation capability; global decisions; inherent knowledge; master berthing chart; railway station; real-time environment; reasoning directions; rule-based system; subtasks; train schedules; transformational synthesis; user acceptance; user-friendly interface; visual capability; Artificial intelligence; Delay; Dynamic scheduling; Expert systems; Information systems; Knowledge based systems; Rail transportation; Real time systems; Resource management; Safety;
fLanguage
English
Publisher
ieee
Conference_Titel
Artificial Intelligence Applications, 1991. Proceedings., Seventh IEEE Conference on
Conference_Location
Miami Beach, FL
Print_ISBN
0-8186-2135-4
Type
conf
DOI
10.1109/CAIA.1991.120903
Filename
120903
Link To Document