Dynamic crude oil fouling prediction in industrial preheaters using optimized ANN based moving window technique

► ANN model was applied for online prediction and monitoring of crude oil fouling in heat exchangers. ► The model is automatically online updated at any time when a new data block can be achieved. ► The results revealed the model could well predict (multi step-ahead) the ascending processes. ► Generalization of the model was proved by proximity of errors for training and prediction subsets.