A Revised Deep Belief Network for Predicting the Slurry Concentration of a Cutter Suction Dredger

In order to predict the slurry concentration of a Cutter Suction Dredger (CSD), a revised Deep Belief Network (DBN) that contains two classifier models is proposed in this work. The two classifier models (i. e., a constant step model and a probability sampling model) are used to process the original data captured in a CSD during a dredging project. Then the classifier models are employed to build the revised DBN to predict the slurry concentration of a CSD. The simulated results show that the proposed approach can effectively extract the features of working data, and also predict the slurry concentration efficiently.