ANALYSIS DATA SETS USING HYBRID TECHNIQUES APPLIED ARTIFICIAL INTELLIGENCE BASED PRODUCTION SYSTEMS INTEGRATED DESIGN

Abstract

The paper proposes a prediction model of behavior spindle from the point of view of the thermal deformations and the level of the vibrations by highlighting and processing the characteristic equations. This is a model analysis for the shaft with similar electro-mechanical characteristics can be achieved using a hybrid analysis based on artificial intelligence (genetic algorithms - artificial neural networks - fuzzy logic). The paper presents a prediction mode obtaining valid range of values for spindles with similar characteristics based on measured data sets from a few spindles test without additional measures being required. Extracting polynomial functions of graphs resulting from simultaneous measurements and predict the dynamics of the two features with multi-objective criterion is the main advantage of this method.

Keywords

• fuzzy logic
• artificial neural network
• genetic algorithms
• Matlab R2011b
• VGD 1.7