海上人命损失规律及个人风险可接受标准研究

发布时间：2018-03-04 16:12

本文选题：海上人命损失　切入点：规律　出处：《大连海事大学》2016年博士论文　论文类型：学位论文

【摘要】：当前阶段海上交通安全特点已经发生显著变化,研究当下海上人命损失特点,对于提高海上人命安全防护能力具有重要意义。论文在既往研究的基础上,明确了相关的研究目标、思路和任务,围绕海上人命损失规律、趋势和可接受标准等开展了以下几方面的研究。首先,研究海上人命损失的分布、时间、形态和致死率特征,定量分析海上人命损失的总体规律。其次,利用Apriori算法对造成海上人命损失的事故属性进行了多维关联规则挖掘研究,发现了关于海上交通事故与船上操作事故间的重要关联规则,揭示了海上交通安全与船上操作安全具有紧密联系,提出了预防海上人命损失的对策。再次,研究了基于自回归滑动平均模型(Autoregressive Integrated Moving Average Model, ARIMA)和支持向量机(Support Vector Machine, SVM)的海上人命损失组合预测模型。预防海上人命损失一直是海上交通安全研究和实践的焦点。针对现有海上人命损失预测方法的不足,提出了一种基于ARIMA模型和SVM方法相结合的预测模型。首先采用ARIMA模型对时间序列线性部分建模分析,然后采用SVM模型对非线性部分进行建模。经验证,组合模型相对于单模型的预测具有更高的精度,该组合预测模型是一种有效的海上人命损失预测模型。最后,应用帕累托(Pareto)最优等理论研究了海上个人风险可接受标准(Marine Individual Risk Acceptance Criteria, MIRAC) o针对当前海上交通领域尚未建立一个被广泛接受的风险标准的现状,在定义MIRAC概念和阈值确定原则的基础上,基于Pareto最优原理确定了符合实际的MIRAC阈值,回答了海上交通安全系统“到底多安全才够安全”的经典问题。
[Abstract]:The characteristics of the current stage of maritime traffic safety has changed significantly, the current research of loss of life at sea, has an important significance for improving the safety of life at sea protection. Based on the previous research, the research target, ideas and tasks around the loss of life at sea, trend and acceptance criteria to carry out research in the following aspects. Firstly, the distribution of life at sea, the loss of time, morphology and mortality characteristics, quantitative analysis of overall loss of life at sea. Secondly, the accident caused loss of life at sea attribute of multidimensional association rule mining based on Apriori algorithm, found out about the operation of important association rules between the maritime traffic accident the accident with the boat, reveals is closely related to safe operation of maritime traffic safety and ship, puts forward some countermeasures for preventing the loss of life at sea. Again, the research on the autoregressive moving average model (Autoregressive Integrated Moving Average Model, ARIMA) and support vector machine (Support Vector Machine, SVM) the loss of life at sea, the combination forecasting model. To prevent loss of life at sea has been the focus of maritime traffic safety research and practice. Aiming at the shortage of the existing prediction methods of the loss of life at sea. Put forward a kind of ARIMA model and SVM method based on the combination forecast model. Firstly, by the analysis of ARIMA model of linear time series modeling, and then use SVM model to model the nonlinear part. After verification, compared to the single model prediction model combined with higher accuracy, the combination forecasting model is an effective sea loss of life prediction model. Finally, the application of Pareto (Pareto) optimal theory to study the risk of maritime personal acceptance criteria (Marine Individual Risk Acceptance Criteria, MIRAC o) according to the current status of maritime traffic field has not yet established a widely accepted risk criteria, based on determining the principles define the concept of MIRAC and the threshold, the Pareto optimal principle to determine the MIRAC threshold based on the practical, answered the classic problem of maritime traffic safety system is safe enough security ".

【学位授予单位】：大连海事大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：U698

【参考文献】