Application of fuzzy logic and neural network technologies in cone crusher control

摘要

Fuzzy Logic presents a robust technique for accommodating measurement uncertainty and error contaminated signals, and a proven technology for representation of heuristic knowledge and automation of subjective manual operations. Neural Network Technology, on the other hand, provides a valuable tool for modelling and prediction of non-linear and difficult processes. A brief summary of Fuzzy Logic and Neural Network principles is presented to provide a basis for the introduction of two applications, one in Fuzzy Logic and the other utilising a Fuzzy Neural Network. The applications are part of a major project aiming to develop a new generation of fully automated control systems for Autocone cone crushers. The Fuzzy application is used in conjunction with a number of novel wear sensors to predict the rates of liner wear under various operational conditions, in eluding feed- size, moisture content and crusher's setting. The Neural Network application has been developed as part of a Knowledge-Based. Condition Monitoring System and provides a novel technique for vibration analysis of the crusher as a fault diagnosis routine.