ABSTRACT
At present, China has increased investment in inland waterway locks. More and more locks have occurred in China. However, the safety monitoring of cofferdams in locks remains in the traditional way, especially the displacement prediction of cofferdam in the lock. When the cofferdam displacement exceeds the warning value, the safety of the cofferdam would be seriously threatened. But the cofferdam displacement is affected by various factors simultaneously so that the system state and data change trend cannot be accurately predicted. The adaptive Kalman filter method is used to effectively fuse the surface horizontal displacement and the deep horizontal displacement closest to the surface at the same position whose displacement changed fastest. Besides, the grey GM (1, 1) model is used to predict the displacement of the cofferdam, and the relevant indicators of the model are tested to evaluate the prediction accuracy of the model. Then, the prediction results are appropriately corrected according to the state transition probability in the Markov chain. The results show that after the data processing of the combined model, the prediction data are better estimation for the displacement of cofferdam compared with the data of a single measurement point. At the same time, the prediction accuracy is significantly improved compared with linear prediction, indicating that this combined data processing model can be effectively used for cofferdam displacement prediction.
