Real-Time Prediction of Large-Scale Ship Model Vertical Acceleration Based on Recurrent Neural Network

In marine environments, ships are bound to be disturbed by several external factors, which can cause stochastic fluctuations and strong nonlinearity in the ship motion. Predicting ship motion is pivotal to ensuring ship safety and providing early warning of risks. This report proposes a real-time ship vertical acceleration prediction algorithm based on the long short-term memory (LSTM) and gated recurrent units (GRU) models of a recurrent neural network. The vertical acceleration time history data at the bow, middle, and stern of a large-scale ship model were obtained by performing a self-propulsion test at sea, and the original data were pre-processed by resampling and normalisation via Python. The prediction results revealed that the proposed algorithm could accurately predict the acceleration time history data of the large-scale ship model, and the root mean square error between the predicted and real values was no greater than 0.1. The optimised multivariate time series prediction program could reduce the calculation time by approximately 55% compared to that of a univariate time series prediction program, and the run time of the GRU model was better than that of the LSTM model.

The normal distribution fitting method for frequency distribution characteristics of peak arrival time of earthquake

The physical meaning of phase difference controls the occurrence time of peak values of triangular series, which can be abbreviated as peak arrival time. This article describes a method for deducing the correspondence between the energy distribution characteristics in the time domain and the peak arrival time. The variation laws of acceleration time–history curve of natural seismic wave can be transformed into its energy distribution characteristics in time domain. Based on the needs of seismic engineering, this article proposes a conversion formula to describe the relationship between the peak arrival time and the energy distribution characteristics in the time domain. To ensure that the variation laws of acceleration time–history curve of artificial seismic waves are more consistent with those of natural seismic waves, a normal-fitting annealing algorithm is proposed based on the normal-fitting method. The proposed method not only considers the frequency distribution characteristics of the peak arrival time comprehensively but also optimizes the fitting parameters according to the actual situation. The results of all the experimental cases verify the rationality and reliability of the proposed method.

Investigating the Performance of Viscoelastic Dampers (VED) Under Nearfield Earthquakes with Directivity Feature

One of the most important factors that make structures vulnerable to earthquakes is the short distance between structures and epicenter. Near-field earthquakes have special properties, such as increasing acceleration applied to the structure, which distinguishes them from far-field earthquakes. Therefore, the absorption of input energy for structures located near the faults is very important. Hence, by rotating the earthquake acceleration time history and comparing the resulting spectral acceleration response, the angle which applies the greatest force to the structure on the earthquake directivity side is obtained, and then the performance of a steel structure with viscoelastic dampers (VED) under near-field earthquakes with directivity feature is investigated. After analyzing the structure using nonlinear time history analysis, it was observed that the directivity phenomenon leads to significant increase in the force applied to the structure, but the viscoelastic dampers showed an acceptable performance in both states of with and without directivity.