【作者】 张笛；

【导师】 严新平；

【作者基本信息】 武汉理工大学 ， 载运工具运用工程， 2011， 博士

【摘要】 作为世界上少有的季节性大型河流,长江受季节因素以及通航枢纽下泄流量等诸多因素的综合影响,枯水期航道水深明显降低,多处浅险航道出现险情后常常导致航道拥挤、船舶压港、航道出浅等碍航事件发生,加之少数船舶超载运营或违规操作,船舶搁浅、碰撞等水上交通安全事故时有发生,直接影响到了内河航运的畅通与安全,同时也给我国航运的可持续发展带来了不利影响。通航风险的评价分析与预测预警技术,对预防、减少枯水期航道碍航事件以及其他水上交通安全事故有着至关重要的作用。文中围绕长江枯水期水路交通运输中的安全问题,通过开展通航风险评价和预测预警技术等研究,识别枯水期长江通航风险的关键因子,分析不同区域的风险分布,提出风险控制的对策建议；通过进一步将航道碍航作为枯水期长江通航风险的典型,研究碍航风险的评价预测模型和成本效益对策,实现碍航风险的预测预警决策支持。具体研究工作内容如下：1)枯水期长江通航风险的识别研究。通过建立基于多风险影响因素的长江水上交通安全评价层次模型,运用离散模糊集和层次分析法对枯水期长江通航风险的关键要素进行了识别研究,并基于控制风险要素的综合效用排序提出了最优的风险控制方案。2)枯水期长江通航风险的评价研究。在已建层次评价模型的基础上,运用基于模糊规则库的证据推理方法处理多层次、多指标模型的综合评价问题,通过对长江上、中、下游相关主客观数据的收集,分析了枯水期长江不同区域的通航风险分布特征,指出了风险控制的重点区域。3)航道碍航风险的建模研究。针对长江航道碍航的成因提出了基于事故特征分析的研究思路,在收集历史水上交通安全事故数据的基础上运用相关性分析和贝叶斯网络对碍航风险进行建模,通过情景分析指出了造成航道碍航的关键要素。4)碍航风险对策的成本效益分析。提出了针对碍航关键要素的风险控制方案,建立了多方案的成本效益评价模型,运用连续模糊集和逼近理想解的排序方法对各备选方案进行成本效益分析,研究不同碍航风险状态下的碍航对策成本效率排序。5)碍航风险的预警决策支持研究。在已建碍航风险贝叶斯网络模型的基础上提出了碍航风险指标CRI (Congestion Risk Index)的计算方法,建立了基于CRI的碍航预警等级评价指标体系,结合碍航风险控制方案的成本效益分析结果提出了各预警等级的碍航对策建议,解决了碍航风险预测预警决策支持系统的关键问题。更多还原

【Abstract】 As one of the few seasonal large-scale rivers worldwide, the waterway depth of Yangtze River drops dramatically during its dry season caused by seasonal factors as well as the release quantity from reservoirs, thus congestion incidents have often occurred in shallow channels of Yangtze River particularly when marine accidents such as groundings and collisions happen during the dry season. This may not only influence the efficiency and safety of inland waterway transportation, but also affect the sustainable development of Chinese shipping industry.The techniques of navigational risk assessment and precaution play a vital role in the prevention and reduction of congestion incidents and other marine accidents. This PhD thesis focuses the navigational risk in the dry season of Yangtze River, reviews various methods for navigational risk assessment and precaution, identifies the critical safety elements, analyzes the spatial risk distribution and proposes the corresponding risk control options (RCOs). Furthermore, the congestion risk is chosen as the typical navigational risk during the dry season of Yangtze River that a congestion model has established and a cost-benefit analysis for RCOs has carried out. More specifically, the main content of this research is as follows:1) Navigational risk identification in the dry season of Yangtze River. A hierarchical model involving multiple influencing factors is established for navigational risk assessment of Yangtze River by which the safety critical elements in the dry season of Yangtze River are identified utilizing discrete fuzzy sets and an Analytic Hierarchy Process, AHP. A utility approach is presented that the best RCO is selected by ranking the utility values of various RCOs.2) Navigational risk estimation in the dry season of Yangtze River. Based on the established hierarchical model, the Evidential Reasoning (ER) approach and a fuzzy rule base are combined to cope with the multiple criteria analysis. By collecting both subjective and objective data of upstream, midstream and downstream of Yangtze River, the spatial distribution of navigational risk in the dry season is analyzed that the focal area for risk control is highlighted.3) Congestion risk modeling. An accident based idea is proposed for congestion risk study based on the cause of congestion incidents in Yangtze River. By collecting the historical data of marine accidents, a congestion risk model is established using a correlation analysis and the Bayesian Network (BN) approach. The congestion critical elements are further analyzed by scenario analysis.4) Cost-benefit analysis for congestion RCOs. Corresponding congestion RCOs are proposed based on the identified congestion critical elements. By establishing a multiple alternative cost-benefit analysis model, the cost effectiveness of each RCO is calculated using continuous fuzzy sets and the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) approach that the preferences of RCOs under various risk conditions are discussed.5) Congestion risk precaution and decision support. A Congestion Risk Index (CRI) is presented based on the proposed congestion risk BN model by which a novel evaluation system for congestion warning is established. RCOs for every warning level are selected referring to the proposed cost-benefit analysis. Therefore, the risk precaution and decision support of congestion incidents can be achieved.更多还原