Predicting alarms in supermarket refrigeration systems using evolved neural networks and evolved rulesets
David Taylor, David Corne, Daniel Taylor, John H. Harkness · 2003
Supermarkets suffer serious financial losses owing to problems with their refrigeration systems. Most refrigeration units have controllers which output "high-temperature" and similar alarms. We describe a system developed to predict alarm volumes from this data in advance, and compare evolved and backpropagation-trained neural networks, and evolved rulesets for this task.