Application of an adaptive control system to a gold processing plant

摘要

Installation and operation of a commercially available adaptive process control system in a gold processing plant in Nevada has proven technically and financially successful. Test results show that the system paid for itself through increased throughput within two weeks of commissioning. The control system uses an expert rule base written with both crisp and fuzzy logic rules integrated with continually adapting neural network models used by predictors, and optimizers. This allows the control system to provide optimum process performance under continually varying feed conditions. This paper describes the installation, rule and predictor configuration, and statistically verifiable testing of the control system in an operating plant.