An Application of Artificial Intelligence Theory to Reconfigurable Flight Control.
Abstract
A knowledge-based flight control system capable of detecting, identifying, and reconfiguring for a wide range of aircraft failures has been designed. Combining modern control theory, statistical hypothesis testing, and artificial intelligence techniques, this research addresses the question of whether or not an intelligent computer could assist a pilot during a failure. Analytical redundancy techniques, including a Generalized Likelihood test, are used for failure detection. Failure diagnosis is performed by an expert system. Utilizing knowledge of cause-and-effect relationships between all aircraft components and the statistical results of a Multiple-Model algorithm, the expert system decides which aircraft component has failed and how to reconfigure for the failure. Preliminary tests on an 8-bit microprocessor system were conducted and are summarized, and plans to expand to a 16-bit multi-microprocessor system are outlined. Keywords: Cybernetics; Systems analysis; FDIR (Aircraft, Failure Detection Identification and Reconfiguration). (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 30, 1984
- Accession Number
- ADA165082
Entities
People
- David A. Handelman
Organizations
- Princeton University