Knowledge-based system for condition monitoring of cone crushers

Paper presented at a meeting on: Artificial intelligence in the minerals sector, held in Nottingham, UK, 20 April 1993 (original title: Knowledge-based condition monitoring of cone crushers). A large proportion of the electricity used by the quarrying industry is consumed in the comminution and sizing process. In most cases aggregate processing plants operate with minimum control features resulting in limited operational efficiency. Improved control is required to increase operational efficiency with concomitant savings associated with reduction in re-circulating load and improved resource utilisation. Development work has suggested that a combination of adaptive control and process condition monitoring could improve control of comminution plants. Many commercial process control and condition monitoring systems are available, but none were deemed suitable due to the domain specific characteristics of the application. An assessment of the particular characteristics of cone crushing concluded that the application was best suited to the use of small domain specific process control combined with bespoke condition monitoring. The processing power of the control system for the project is a function not of the academic requirement, but rather the economic and practical constraints imposed by the application. As a result of the constraints a high-end PC system has been designed in preference to a workstation. Within the research project work packages are sub-divided into novel wear sensor development, knowledge based control/condition monitoring and the investigation and mathematical modelling of the wear mechanisms inherent in cone crushing. Numerous strategies were considered to implement the control system outlined. The system currently under development is a modular design which utilises intelligent novel sensors to provide a variety of data inputs