An Expert Network Process Control Application
Abstract
In this thesis, a computerized process control technique based on an artificial neural network (ANN) is presented. The process of concern is the wood dryer portion of an industrial retort used for the dehydration stage of wood waste pyrolysis in the making of charcoal. With the proposed control scheme, an ANN monitors the dryer system parameters, detects deviations from the norm, and applies corrective actions to counteract deviations should any occur. A standard backpropagation network structure is used to associate patterns of sensor readings with appropriate control responses that are based on process design specifications and operation heuristics. As heuristic programming is employed, the dryer controller is not unlike a typical knowledge base system, except that it is driven by an ANN rather than the more traditional rule base normally associated with expert system type applications
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1994
- Accession Number
- ADA283163
Entities
People
- Douglas M. Rodgers
Organizations
- Air Force Institute of Technology