Feedback Control System in Dredging Engineering Based on Convolutional Neural Network Prediction

The control process of the cutter dredger plays a very important role in the dredging project. Currently, it mainly is performed by user experiences and basic working rules. Hence, the control process is not optimized, robust, realtime, and efficient. In this paper, after constructing a database of historical big data in the dredging engineering, the one-dimensional convolution neural network (CNN) is used to predict flow velocity and concentration, and the optimal outputs of CNN are selected from the database. According to the closed-loop control way, the control process is optimally carried out. Experimental results validate the proposed method in the actual dredging process.