基于加权平滑季节趋势时序分解法的船舶机损事故风险防控

Prevention of Marine Engine Fault by Combined Use of Weighted Smoothing and Trend Analysis Approach

摘要: 船舶机损事故严重威胁船舶运营安全，给航运公司带来重大经济损失，一直是航运公司风险防控的重点。为探究此类事故规律，控制和减少同类事故发生，利用基于局部加权平滑的季节趋势时序分解方法构建数学模型，对船舶机损事故的分布特征和时序规律进行描述和分析，发现船舶机损事故的发生具有季节性规律。根据机损事故的季节性规律，采取针对性防范措施。通过事故数据对比，验证该方法在船舶机务安全管理方面的实际效果，为船舶机务风险防控提供理论依据。

Abstract: Seasonal engine trend data is processed by a locally weighted smoothing algorithm.Time series decomposition is performed on the smoothed trend series and the mathematical model describing the distribution characteristics and the regularity of the time series is developed.Analysis shows that the risk of faults is season dependent.The countermeasures are suggested.