Port Throughput Forecasting Based on Broad Learning System with Considering Influencing Factors

With the development of the shipping industry, port throughput has increased significantly in recent years. Chinese ports have become increasingly important. The fluctuation of container throughput in ports is affected by many factors. The effective and accurate forecasting of port throughput provides a scientific reference for the development of the port. Based on the port throughput data of Lianyun Port in China from the first quarter of 2005 to the fourth quarter of 2016, this paper uses univariate linear regression, multiple linear regression, and broad learning system to forecast the container throughput of the port in each quarter of 2017 and 2018. Comparing the forecasting results with the actual throughput. The experimental results show that the broad learning system can consider two economic influencing factors about the throughput at the same time and forecast port throughput accurately and effectively. Therefore, broad learning system is more suitable for the forecasting of port throughput than other methods.