PREDICTION THE EVOLUTION OF TEMPERATURE AND VIBRATIONS ON SPINDLE USING ARTIFICIAL NEURAL NETWORKS AND FUZZY LOGIC

Abstract

Simulation spindle behavior in terms of temperature and vibration at higher speeds is more economical and more secure (avoid undesirable mechanical events) than testing practice. Testing practice has an important role in finalizing the product but throughout the course of prototype testing is more advantageous economic development simulation parameters based on data sets collected to dangerous speeds. In this paper we present an analysis mode hybrid (artificial neural networks - fuzzy logic) on prediction the evolution of temperatures and vibrations at higher speeds for which no measurements were made. The main advantage of the method is the simultaneous prediction of the dynamics of temperature and vibration levels.

Keywords

• spindle
• artificial neural networks
• fuzzy logic
• ANFIS