The possibilities of applying of artificial neural networks in automatic diagnostic system of machine cutting process
Source abstract
Abstract
The causes of failures in contemporary automatic diagnostic systems working with use of neural networks in industrial conditions are given. It has been stated that one of the basic causes was the lack of opportunity to perform logical operations by the artificial neuron model of McCulloch and Pitts which makes the core of artificial neural networks programs. A new method of data gathering and initial preparation has been offered here, based on the methods derived from non-linear dynamics, fuzzy sets and the theory of chaos. The data processed in such a way are transferred to the neural networks of Feed Forward Back Propagation type, which allows shortening the network's training, increasing the calculations speed and shortening the reaction inertness time of the FFBP network (on-line mode).
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Engineering Technology and Methodologies · Engineering Diagnostics and Reliability · Advanced machining processes and optimization
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References
Reference metadata was not supplied for this record.