基于灰色马尔可夫模型的山东辖区海上险情事故量预测

王士鹏¹，周兆欣^1*，秦圻¹，马续仕¹，韩洋¹，邸光辉²

1.山东交通学院 航运学院，山东 威海  264200；2.烟台大学 继续教育学院，山东 烟台  264000

摘要：为确保山东海域辖区的水上交通安全，减少海上险情事故，提高水路运输的安全性，分析山东海域辖区内2010—2021年的海上险情事故量，采用灰色模型（grey model，GM）中的GM（1，1）模型与马尔可夫理论相结合的灰色马尔可夫模型预测山东海域辖区未来2 a的海上险情事故量，采用MATLAB软件编程求解。预测结果表明：灰色马尔可夫模型预测结果的平均相对残差比仅使用GM（1，1）模型提高5.240 8%，预测精度进一步提高。研究结果为改善山东海域辖区的船舶航行安全，建立海上交通预警机制提供可靠的数据参考。

关键词：山东海域；水上交通安全；GM（1，1）模型；马尔可夫理论；灰色马尔可夫模型

Forecast of amount of marine dangerous accidents in Shandong jurisdiction based on grey Markov model

WANG Shipeng¹,ZHOU Zhaoxin^1*,QIN Qi¹,MA Xushi¹,HAN Yang¹,DI Guanghui²

1.School of Navigation and Shipping, Shandong Jiaotong University, Weihai 264200, China；

2.School of Continuing Education, Yantai University, Yantai 264005, China

Abstract:In order to ensure the marine traffic safety, reduce dangerous maritime accidents at sea and improve the safety of waterway transportation, the grey Markov model combining grey model GM(1,1) and Markov theory is used to predict the amount of maritime dangerous accidents in Shandong sea area in the next two years by analyzing the amount of maritime dangerous accidents in Shandong sea area from 2010 to 2021. It is programmed and solved by MATLAB software. The prediction results show that the average relative residual of the prediction results after the revision of Markov chain is 5.240 8% higher than that of the GM (1,1) model, and the prediction accuracy is further improved. The results provide a reliable data reference for improving the ship navigation safety and establishing the maritime traffic early warning mechanism in Shandong sea area.

Keywords:Shandong sea area; marine traffic safety; GM(1,1) model; Markov theory; grey Markov model