Forecasting Port Container Throughput with Deep Learning Approach

In order to improve the accuracy of port container throughput prediction, an improved ABC-BP prediction model based on Artificial Bee Colony (ABC) was proposed. With the advantage of global search ability and difficulty in local optimization, the weights and thresholds of Back Propagation (BP) neural network are optimized, and the optimal weights and thresholds in the network are finally determined. Referring to the existing research results, Qingdao 2014-2019 (24 quarters) GDP, port throughput, total foreign trade import and export volume were selected as network input, and Qingdao port 2014-2019 port container throughput statistics were used as network output, a total of 24 sets of data were constructed, so as to build a BP neural network prediction model. The first 20 groups of data were used as the training set, and the last 4 groups of data were used as the test set for instance verification. The results show that, compared with the traditional BP neural network and the BP neural network prediction model optimized by Genetic Algorithm (GA), the ABC-BP model can significantly improve the prediction accuracy of port container throughput.

As a nonlinear variation, the port container throughput is easy to get affected by market. There are many prediction methods such as regression analysis, exponential smoothing. However, each of these prediction methods has their own characters which all lead to low precision of container throughput forecasting. Constructed on the foundation of BP network, the combination forecast model performs well in time series forecasting and solves the problem excellently. Consequently, based on Elman neural network, this article builds a combination forecast model to improve the precision level of forecast. At last, with the empirical analysis of container throughput in Shanghai port, the reliability and accuracy of the combination model are proved.

Forecasting the demand for container throughput is a critical indicator to measure the development level of a port in global business management and industrial development. Time-series analysis approaches are crucial techniques for forecasting the demand for container throughput. However, accurate demand forecasting for container throughput remains a challenge in time-series analysis approaches. In this study, we proposed a mixed-precision neural architecture to forecasting the demand for container throughput. This study is the first work to use a mixed-precision neural network to forecast the container throughput—the mixed-precision architecture used the convolutional neural network for learning the strength of the features and used long short-term memory to identify the crucial internal representation of time series depending on the strength of the features. The experiments on the demand for container throughput of the five ports in Taiwan were conducted to compare our deep learning architecture with other forecasting approaches. The results indicated that our mixed-precision neural architecture exhibited higher forecasting performance than classic machine learning approaches, including adaptive boosting, random forest regression, and support vector regression. The proposed architecture can effectively predict the demand for port container throughput and effectively reduce the costs of planning and development of ports in the future.

Accurate forecast of container throughput is an important basis for reasonably planning port construction, making port operation plan and adjusting port development direction. Aiming at the complex nonlinear characteristics of port container throughput, a hybrid model based on selective deep-ensemble for container throughput forecasting (HMSD) is presented in this paper. Firstly, this model decomposes the original container throughput time series into several intrinsic mode functions and a residual by empirical mode decomposition (EMD). Considering highly nonlinear characteristics of each intrinsic mode function, the proposed model constructs three deep neural networks, namely, long short term memory (LSTM), gated recurrent unit (GRU) and convolutional neural network (CNN), as base learners to predict intrinsic mode functions. Then, this model establishes selective deep-ensemble forecasting model by improved group method of data handling (GMDH) on intrinsic mode functions and obtains their ensemble forecasting results. Furthermore, this model uses an autoregressive integrated moving average model to predict the linear residual. Finally, the forecasting result of the total container throughput is obtained by integrating the forecasting results of all intrinsic mode functions and the residual. The empirical analysis of container throughput data of Xiamen port and Shanghai port in China shows that the HMSD model has better forecasting effect than other models.

The statistical techniques based on compositional data are applied to investigate the evolution of the traffic share of the container throughput in a multi-port system. Compositional vectors are those which contain relative information of parts of some whole. The application of conventional statistical techniques to compositional data may lead to erroneous conclusions and spurious correlations. Therefore, compositional data (CoDa) should be treated taking into account their own mathematical structure. The so-called log-ratio approach provides a set of transformations that allow to apply conventional statistical techniques to the transformed compositional data samples. Thus, the objective of this paper is double. As a first stage it aims to introduce the CoDa formalism and highlight its potentiality in the port container throughput analysis as example of transport system providing an applied example: the container throughput evolution in the Spanish Mediterranean Ports system during the period 1976–2015. Second, based on the previous analysis, the aim is to characterize the container throughput in SpanishMed ports and its temporal evolution. The CoDa analysis clarifies the interpretation and data association of the container traffic throughput evolution in function of some selected change points: boom of containerization in 1990s and 2008 crisis. This contribution proves that the CoDa methodology is useful to investigate the complexity of the transport disciplines in order to understand and to manage the spatial integration that results from the movement of people and freight.

The X-Press Pearl disaster has the unfortunate hallmarks of other global container ship disasters that are becoming increasingly common and will continue to cause future environmental damage. Port container throughput is ~830 million twenty-foot equivalents (TEUs), projected to grow for the next four decades at minimum, and industry experts predict that ship sizes will continue to grow to 50,000 TEU. Misdeclared content, poor packing, inadequate stowage/lashing, ship-board fires, climate-change-driven intensity of storms, and increasing ship size all ensure increasing environmental damage from container ship maritime disasters. Industry consensus is that the number of containers lost at sea is conspicuously undercounted, and the impact of lost containers is unstudied. This paper reviews specific aspects of the container ship industry that contribute to environmental damage, and then addresses needs for risk, impact assessment and environmental management as pertains to industry operations and container ship disasters such as the X-Press Pearl.

基于面板数据港口集装箱吞吐量经济效应分析

Purpose – The purpose of this paper is to explore the relationship between port and city growth with the help of the port impact curve and relative concentration index (RCI). This is done from the two aspects of (1) clarifying the characteristics of the interaction between Ningbo port and city, and (2) clarifying the phased development of the relationship between Ningbo port and city. Design/Methodology/Approach – Based on the method of the port impact curve and RCI, we conduct a case study of Ningbo, a port-city in Southeast China, and we also make a meaningful comparison between Ningbo and other famous port cities in China. Findings – This paper finds that Ningbo port and city have been in a relatively loose relationship from 2010 to 2021. Over these 12 years, the relationship between Ningbo port and city has shown signs of changing from the stage of agglomeration and diffusion effect to the stage of urban selfgrowth effect. Also, we find that, compared with other port cities, the positive correlation between Ningbo port cargo throughput and urban economy development is relatively weak. However, Ningbo’s export-oriented economic characteristics form a certain boost for the development of port container throughput. Research Implications – In the management of this paper, this study revealed the flexibility and complexity of the port-city relationship, suggesting that the stage characteristics of the port-city relationship should be combined on the road of the sustainable development of ports and cities, which will also provide reference for follow-up research.

Seasonal container throughput forecasts at ports are immensely important to logistics companies, shipping lines, port authorities and shipyards. Such forecasts allow shipping lines and port operators to formulate appropriate short-to-medium strategies in order to maintain competitiveness. Seasonal autoregressive integrated moving average (in short, SARIMA) models can be employed for this purpose to provide reliable seasonal forecasts of container throughput at a given container port. This article explores the use of SARIMA models in forecasting container throughput at several major international container ports, while taking into consideration seasonal variations. First, the SARIMA model development methodology is described. Second, a database consisting of monthly container port traffic data between 1999 and 2007 for international container ports is developed. Short-term container demand forecasting models are then developed for each of the top 20 international container ports for the purpose of monthly container throughput prediction. Through the use of various performance metrics, the effectiveness of the developed SARIMA models for these ports is evaluated. It is found that SARIMA models can produce reliable throughput forecasts at major international ports. Qualitative insights are then drawn, thereby allowing shipping and port operators to make better tactical and operational decisions.

The aim of the present paper is to explain the Mediterranean Transhipment Rule by a mathematical model showing the relationship between the diversion distance and the transhipment volumes, as well as the relationship to the total container traffic volume. Thus, the formula was applied to the figures achieved in 1994, 1995 and 1996. To conform with 1994, 1995 and 1996 data the formula was slightly altered so as to forecast transhipment volumes in the Mediterranean ports. By applying the formula to the figures achieved by some Mediterranean ports, the estimations were done which describe reliably enough that the transhipment volumes depend on the diversion distance and the total traffic throughput. In short, the volume of cargo transhipped in a single port is an inverse linear function of the distance of the port from the main-line vessels' route transiting the Mediterranean, and a linear function of the port container traffic volumes.

A major goal for port authorities, operators, and investors is to achieve efficient operations and effective environmental protection. This is because the environmental performance of a container port is important for its competitiveness and sustainable development. However, the container ports along the Maritime Silk Road (MSR) have caused numerous problems with the rapid development, among which the most significant problem is environmental pollution. In this paper, we aim to measure and compare the environmental performance and operational efficiency of ten major container ports along the MSR, including the ports of Shanghai, Hong Kong, Singapore, Kelang, Laem Chabang, Colombo, Dubai, Barcelona, Antwerp, and Hamburg. We develop an improved, inseparable data envelopment analysis (DEA) model with slack-based measures (SBMs) to evaluate and compare the environmental performance and operational efficiency, and we incorporate the desirable output of container throughput as well as the undesirable output of CO2 emission. Our results show that. Overall. these container ports perform better in terms of operational efficiency than environmental performance. We also provide insights for management and policy makers for container ports with different levels of operational efficiency and environmental performance.

Investigating the learning approaches of students in nursing education

The relationship between transhipment incidence and throughput volatility in North European and Mediterranean container ports

We show in this paper that the throughput data for the top 300 container ports reported each year by the various authorities follows a simple truncated lognormal distribution. This surprising phenomenon repeats itself every year from 1982 to 2006, despite many tumultuous changes in the container shipping world. The empirical data suggests that Gibrat's Law of proportionate growth indeed holds for the world container throughput data. Unfortunately, the classical stochastic growth model and other variants often used to explain the origin of this law appears to be too simplistic for the container terminal industry. We use instead the perspective that the container terminal throughput data are essentially an aggregate measure of the number of visitations as each container circulates on the world shipping network, and use this to propose a Markov chain based container circulation model to explain the origin of this phenomenon. Simulation results show that our network-based model is able to replicate the behavior of the empirical data to a reasonable degree of accuracy, and does not contradict the law of proportionate growth. More importantly, this model is able to replicate the relationship between the degree of connectivity of a port (i.e. number of linkages with other ports) and its association with the container throughput data, an empirical regularity which could not be explained using classical approaches.

The operational performance of container ports has received more and more attentions in both academic and practitioner circles, the performance evaluation and process improvement of container ports have also been the focus of several studies. In this paper, Data Envelopment Analysis (DEA), an effective tool for relative efficiency assessment, is utilized for measuring the performances and benchmarking of the 77 world container ports in 2007. The used approaches in the current study consider four inputs (Capacity of Cargo Handling Machines, Number of Berths, Terminal Area and Storage Capacity) and a single output (Container Throughput). The results for the efficiency scores are analyzed, and a unique ordering of the ports based on average cross efficiency is provided, also cluster analysis technique is used to select the more appropriate targets for poorly performing ports to use as benchmarks.