【作者】 雷定猷；

【导师】 陈治亚；

【作者基本信息】 中南大学 ， 交通运输规划与管理， 2005， 博士

【摘要】 从运输组织角度来看,货物有多种分类。货物按形态可分为成件货物和不成件货物。本文以铁路运输方式为背景研究成件货物的装运优化问题。我们把成件货物划为阔大货物和小件货物两类,而超限货物运输具有阔大货物运输的所有特征,所以论文从超限货物运输优化和小件货物装载优化两方面展开。 1.超限货物运输优化 (1)装载决策。分析超限货物运输基本特征和影响超限货物运输的主要因素,针对超限货物的特征构建超限货物运输评价指标体系,提出了评价模型和评价方法。超限货物装载决策是一个半结构问题,有一部分装载参数可通过建立数学模型求解得到,但多数装载参数由于受数学模型表达能力和求解条件的限制,它不能用结构化的方式表达,所以除使用数学模型之外,文章构建了超限货物装载决策的实例推理模型,提出了知识表达与推理机制,给出了装载推理算法和机器学习算法。 (2)路径决策。路径决策是超限运输的关键技术,文章把超限货物运输路径决策问题描述为双目标双容量的数学模型,用运输经济的思想把双目标模型化为单目标模型,提出了最大容量路算法和减少删除边的检查方法,把超限调度路径决策问题简化为容量限制下的最短路问题,提出了多项式算法。 (3)限界检算。限界检算是超限货物运输安全的保证,文章对铁路限界进行了凸性分析,凸性理论的分析结果极大地简化了货物超限检查的处理过程;文章提出桥隧限界综合算法和限界检算算法。 (4)超限货物运输智能决策系统的开发设计。运用知识推理理念和模型驱动方法构建了IDSS的框架,融合上述研究的系列模型与算法,开发出全路超限货物运输智能决策支持系统,解决了在繁忙路网上超限货物运输决策问题。 2.小件货物装载布局优化 小件货物装载优化属于带约束的布局问题,它是一个NP完备问题。论文构造了集重约束和重心约束条件下的一车多件货物装载布局优化模型;首次提出了s中心骨架,带已布物群约束的布局,形状的相似性等布局新方法,通过构造装载布局函数,中心骨架移动和带已布物群约束布局等算法,采用空间合并和融合技术,求解了一车多件货物装载布局优化问题。进一步把一车多件货物装载模型推广到多车多件装载的情形,构造了多车多件货物装载布局优化模型;提出类First Fit划分算法和多车多件货物装载布局算法,较好地解决了有货物集重约束和重心约束的多车多件装载布局优化问题。更多还原

【Abstract】 From the angle of transportation organization, there are many categories about goods. Goods can be classified into integrated goods and non-integrated goods according to their shapes. The optimizing problem of integrated goods loading and carrying is studied on the background of railway transportation manners in this paper. Goods is divided into long and large goods and small goods ,but out-of-gauge goods transportation has all characters of long and large goods transportation, so this paper is written from two aspects such as optimizing of out-of-gauge goods transportation and optimizing of small goods loading. 1. optimizing of out-of-gauge goods transportation(1)Loading decision. Analyze the basic characters of out-of-gauge goods transportation and the main factors affecting out-of-gauge goods transportation and construct evaluating index system on out-of-gauge goods transportation according to these characters in order to bring forward evaluating models and means. Out-of-gauge goods loading decision is a semi-structured problem and a part of loading parameters can be obtained through setting up and solving mathematical models, but most loading parameters can’t be expressed in structured means because of its constrains from expressing abilities of mathematical models and solving conditions, so this paper constructs example reasoning models on out-of-gauge goods loading decision and bring forwards knowledge expresses and deducing mechanisms and gives loading deducing algorithm and machine studying algorithm.(2)Trace decision. Trace decision is a key technique of out-of-gauge transportation. Out-of-gauge goods transportation is described as a mathematical model with double targets and double capacities in this paper. The model with double targets is changed into a model with a single target using thoughts of transportation economy. This paper brings forward maximal capacity route algorithm and checking means of reducing and deleting side and simplifies the problem of out-of-gauge attempering route decision into minimal route problem with capacity restricts and bring forwards multinomial algorithm.(3)Gauge checking and calculating. Gauge checking and calculating is aguarantee for out-of-gauge goods transportation security. This paper carries out protruding analysis of railway gauge and brings forward general algorithm for gauges of bridges and runnels and checking and calculating algorithm for gauges. The analyzing results of protruding problem greatly simplify disposing process of out-of-gauge goods checking.(4)Opening-up and design of intelligent decision system for out-of-gauge goods transportation. Construct frames of IDS S Using theory of knowledge reasoning and means of model drive and open up intelligent decision support system for the rail out-of-gauge goods transportation inosculating with a set of models and algorithms studied above so as to resolve decision problem of out-of-gauge goods transportation on the busy railway net. 2. loading layout optimizing of small goodsLoading optimizing of small goods belongs to layout problems with restricts. It is an NP complete problem. The paper construct loading layout optimizing model on a single vehicle and multiple goods under the condition of concentrated weight restricts and restricts of center of gravity and brings forward new layout means such as s centered framework, layout with restricts of goods laid out and shape similarities first. It resolves loading layout optimizing problem of a single vehicle and multiple goods through algorithms such as constructing loading layout functions mobile of central framework and layout with restricts of goods laid out and using techniques of space combination and amalgamation. The paper generalizes loading models on a single vehicle and multiple goods into conditions of multiple vehicles and multiple goods further so as to construct goods loading layout optimizing model on multiple vehicles and multiple goods. It brings forward similar First Fit partition algorithm and loading layout algorithm for multiple vehicles and multiple goods and resolves loading layout optimizing problem on multiple vehicles and multiple goods with restricts of goods concentrated weight and center of gravity better.更多还原