Search

Accurate prediction of sea level height is critically important for the government in assessing sea level risk in coastal areas. However, due to the nonlinear, time-varying and highly uncertain characteristics of sea level change data, sea level prediction is challenging. To improve the accuracy of sea level prediction, this paper uses a new swarm intelligence algorithm named the sparrow search algorithm (SSA), which can imitate the foraging behavior and antipredation behavior of sparrows, to determine optimal solutions. To avoid the algorithm falling into a local optimal situation, this paper integrates the sine–cosine algorithm and the Cauchy variation strategy into the SSA to obtain an algorithm named the SCSSA. The SCSSA is used to optimize the parameter values of the CNN-BiLSTM (convolutional neural network combined with bidirectional long short-term memory neural network) model; finally, a combined neural network model (named SCSSA-CNN-BiLSTM) is proposed. In this paper, the time series data of seven tidal stations located in coastal China are used for experimental analysis. First, the SCSSA-CNN-BiLSTM model is compared with the CNN-BiLSTM model to predict the time series data of SHANWEI Station. With respect to the training and test sets of data, the SCSSA-CNN-BiLSTM model outperforms the other models on all the evaluation metrics. In addition, the remaining six tide station datasets and five neural network models, including the SCSSA-CNN-BiLSTM model, are used to further study the performance of the proposed prediction model. Four evaluation indices including the root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and coefficient of determination (R2) are adopted. For six stations, the RMSE, MAE, MAPE and R2 of SCSSA-CNN-BiLSTM model are ranged from 20.9217 ~ 27.8427 mm, 9.4770 ~ 17.8603 mm, 0.1322% ~ 0.2482% and 0.9119 ~ 0.9759, respectively. The experimental analysis results show that the SCSSA-CNN-BiLSTM model makes effective predictions at all stations, and the prediction performance is better than that of the other models. Even though the combination of SCSSA algorithm may increase the complexity of the model, indeed the proposed model is a new prediction method with good accuracy and robustness for predicting sea level change. [ABSTRACT FROM AUTHOR]

Most data collected during routine monitoring of railway station buildings are categorized as in good health, with only a small percentage of samples corresponding to a damaged state. Standard deep neural networks (DNNs) process target samples regarded as normal samples. When typical DNNs are confronted with imbalanced samples, namely, samples with a low proportion of abnormal samples (e.g., damaged stage), this can lead to a high rate of misdiagnosis. To solve this problem, this paper suggests a novel damage monitoring model for railway station buildings, called cluster-based undersampling and Siamese artificial neural network. In this model, normal samples are undersampled based on clustering, and the clustering center and damage samples are combined in a training dataset to reduce the imbalanced rate of damage samples. SANN maps the original space to a feature space where imbalanced damage states are easier to distinguish. Based on SANN, a data-pair selection strategy based on a feedback mechanism is designed to sufficiently mine sample characteristics. A loss function that combines similarity calculation and multi-classification cross-entropy is proposed to improve the representation and classification performance of the model. Six popular algorithms are compared through experiments based on the actual monitoring datasets obtained at the Shijiazhuang Railway Station in China. For different imbalance ratios, the proposed model displayed the highest performance in all indicators. The recall rate for certain imbalanced classifications increased by over 30% compared to the baseline method, verifying the effectiveness and superiority of the model in identifying imbalanced damage states. [ABSTRACT FROM AUTHOR]