【作者】 丁国勤；

【导师】 孟卫东；

【作者基本信息】 重庆大学 ， 技术经济及管理， 2008， 博士

【摘要】 未来信息化战争油料保障的精确性、高时效性对我军传统的油料保障指挥决策模式提出了挑战,纵观我军油料保障理论的研究,主要还是以定性研究为主,定量研究还处于起步阶段,能否实现以定性研究为主的油料保障理论到信息化油料保障的跨越,提高油料保障的精确性和时效性,做好当前我军“反台独”军事斗争油料准备,油料保障指挥决策模型化的研究就成为目前我军油料保障迫切需要研究的重要问题。油料保障指挥决策模型研究以油料保障活动为研究对象,以建立定量化的决策模型为主要目的,是实现油料保障指挥决策信息化重要的理论基础,也是当前油料保障领域有待深化和拓展的应用性研究课题,在军事上,具有重要的现实意义和理论意义。为了较为完整地反映油料保障指挥决策模型化的研究过程,论文紧密结合部队作战油料保障流程,主要研究了五个方面的问题:①油料保障需求预测优化模型研究“精确保障”是未来信息化战争对油料保障的要求,目前我军传统的“超量预储”的方式已经不适应未来信息化作战的“精确保障”要求。因此,精确预测部队油料需求,是提高油料保障能力的关键。论文根据战时作战部队油料消耗可能出现的三种主要情况,分析了每种情况下部队作战油料消耗特点,分别建立了有针对性的灰色预测模型、神经网络预测模型和基于向量夹角余弦的组合预测模型,并对模型进行了实证研究,验证了模型的有效性。②油料布局优化模型研究科学合理的油料布局,可以缩短油料调运的时间,提高油料保障时效,可以产生显著的军事效益。论文结合我军油料布局现状,以未来反台独作战为背景,针对战时油料布局主要问题的优化决策建立了军队油库站布局优化模型、油料储备布局优化模型和油库站选址优化决策模型,并进行了实际运用。③油料调拨运输优化模型研究油料调拨运输,是战时将油料的筹措、储存、供应、补给等油料保障活动融为一体的纽带。作为油料保障的一个极为重要的环节,对油料保障系统整体保障能力的形成与释放具有重大影响。部队战时油料调运主要强调其军事效益,与地方油料供应以经济效益为主有着本质区别,论文结合战时部队油料调运的特点,根据战时油料调运可能出现的情况,首先对传统确定性油料调拨模型进行了改进,建立了战时油料调运的模糊规划模型,并基于遗传算法进行了实现。其次,根据战时可能出现的单个油库担负一体化油料保障的情况,建立了单个油库有限量油车运输模型。最后,基于战时油料运输路径随时可能遭敌破坏的情况,建立了调运路径优化模型。建立的模型都进行了相应的实证研究。④战略支援油料保障优化模型研究针对未来可能发生的反台独作战背景,论文分析了我军东南沿海油料保障现状,发现战时,很可能会出现一线和二线油库不能满足部队作战油料的需求,这时必然要实施战略略支援。为有效提高战时战略支援油料保障能力,实现保障有力的目标,论文根据战时战略支援油料保障的两种决策情况,分别建立军队后方基地战略支援优化决策模型和地方油料保障资源战略支援优化模型,并进行了具体运用。⑤油料保障指挥决策系统研究鉴于未来信息化战争对油料保障指挥决策系统的需求,基于Multi-agent的自治性和协调性等特点,论文将Multi-agent技术引入到军队油料保障指挥系统,能够动态地协调海、陆、空及其他诸军兵种的油料保障,满足未来信息化战争油料保障一体化的要求,Multi-agent技术在军队油料保障指挥决策领域有着重要的应用价值。论文首先基于agent理论,将需求预测优化模型、油料布局优化模型、调拨运输优化模型、战略支援油料保障优化模型进行agent封装,然后基于multi-agent理论,建立了油料保障指挥决策系统的框架体系,为油料保障指挥决策系统研究提供了具体思路。论文主要创新点体现在三个方面:①在油料保障理论体系上有所创新。信息化条件下,对油料保障指挥的精确性提出了更高的要求,而我军缺乏信息化战争油料保障的成功经验,论文率先系统提出部队作战油料保障模型化研究,对提高我军未来信息化战争油料保障的精确性具有重要的军事意义。军事上,作战决定后勤,论文根据目前我军反台独军事斗争准备的战略方针,设计了反台独作战油料保障指挥链,对油料保障指挥链的主要活动建立了相应优化模型,最后根据未来信息化战争油料保障一体化的需求,基于multi-agent理论,给出了油料保障指挥决策系统开发和设计的具体思路,实现了油料保障理论体系的创新。②油料调拨运输,是战时油料保障指挥链最重要的环节,调拨运输的成败关系到整个油料保障系统保障力的释放。论文分析了战时油料调拨运输的特点,立足于军事上强调军事效益优先,经济效益其次的目标,首先基于战时油料调拨运输的模糊性和动态性,对传统的油料调运模型进行了改进,建立油料调运的模糊规划模型,并给出了基于遗传算法的求解思路。其次,基于战时可能出现的单个油库有限量运油车保障部队油料需求的情况,建立相应的模型,进行了遗传算法设计。最后,基于战时油料运输的特点,明确了油料运输路径评价的指标体系,建立了基于模糊三角数的油料运输路径优化模型,和传统的评价方法相比,该模型更能体现油料保障指挥首长的意图,贴近实战需求。③未来信息化油料保障一体化的需求,使得我军传统油料保障指挥决策系统“孤岛式”和“烟囱式”研究模式遇到了瓶颈问题:1)因系统由不同的程序员在不同操作系统及网络环境下、用各种语言在不同时间开发,其接口各不相同,致使软件之间协同作业困难,集成困难;2)由于战争形势发展的不可预测性,使得军队油料保障问题变得极为复杂,如油源的选择、保障环境的动态与不确定性等,因此,需要设计一个适应作战环境的油料保障指挥决策系统模型。论文基于agent的自主性和multi-agent的协调性等特点,将建立的油料需求模型、油料布局优化模型、油料调运模型、战略支援油料保障模型封装为单个agent模型,设计了基于multi-agent的油料保障指挥决策系统框架体系,建立了基于Multi-agent的油料保障问题求解过程,并进行了基于Multi-agent的油料保障指挥系统建模,为油料保障指挥智能决策系统的开发开辟了新的思路。更多还原

【Abstract】 The accuracy and time effect of the POL service in future informational war make a challenge to our traditional POL service command and decision mode. Our POL service theory is mainly qualitative study and quantitative study is in the first phase. Whether we can realize the transformation of qualitative study to quantitative study, increase the accuracy and time effect of the POL service, prepare the POL well for the current“anti-Taiwan-independence”military strive, the study of the POL command and decision model becomes an important problem urgently be solved at present. With POL service action being the study object and building quantitative decision model being the main aim, the study of the POL command and decision model is the theoretical base of realizing informational POL service command and decision, and also is a application study task awaiting deepening and widening which has important practical and theoretical meaning in military.In order to inflect the study process of the POL service command and decision model completely, the thesis mainly does research from five aspects combining with the POL service flow during a battle:①the optimized model study of military POL service demand forecast“Precise service”is the demand of future informational war which our present traditional mode of“store too much in the first”hasn’t meet. So, forecasting military POL need is the key point of increasing the POL service ability. The gray forecast model, neural networks forecast model and combined forecast model based on vector angle cosine, which are built in the thesis adequately combine the characteristics of military POL consuming, and proved to be effective to solve the problems of forecasting of military POL need.②the optimized model study of military POL layoutScientific and rational distribution can shorten the time of POL transportation, save cost, and then increase the time effect of POL service, bring remarkable military and economic benefit. It has great instructional effect for the POL service of future integrated war. Combining with our military POL service actuality and based on future anti-Taiwan-independence campaign, the thesis builds oil-warehouse layout model, military POL depot distribution model and oil-warehouse location choosing. At the same time, the models are applied in practice. ③the optimized model study of military POL transportationPOL transportation is the ligament which connecting POL collecting, storing, providing and filling. As a very important factor of POL service, POL transportation is not only related to the forming and releasing of the continuous service ability of POL system, but also brings great influences to the forming and the whole service ability of the POL system. The military POL transportation is mainly emphasized on military benefit which is ultimately different. According to the possible instance during wartime, the author firstly builds POL transportation fuzzy programming through improving the traditional confirmed model. Secondly, the thesis builds the model of single oil-warehouse with finite oil transportation vehicles. At last, the thesis builds model of transportation route optimized. Every model is applied in practice.④the optimized model study of strategic support POL serviceAccording to the background of anti-Taiwan-independence, when the POL depot of the first and second level cannot meet the need, it is necessarily to carry out strategic support. In order to improve the ability of strategic support POL service and realize the target of effective service, the thesis build optimized models of military strategic rear base support POL service and society oil resource support in according to possible two instances in wartime. Every model is applied in practice.⑤research in POL support command decision system design and exploit According to the requirement of the informational war to POL support command decision system, Introducing Multi-agent technique into POL transportation command system can assort with navy, army, air force and other arm’s POL support entity dynamically. It can also satisfy the future POL support demand in the information-based war. Therefore, Multi-agent technique has important application value in POL support field. Firstly, encapsulate the optimized forecast requirement’s model, the optimized POL storage position’s model, the optimized transportation’s model and the optimized strategic patronized POL support’s model based in agent. Secondly, bring forward the detailed thought in exploiting the POL support command intelligent decision system.The major innovations of the paper:①The innovation of POL service theory system is made. The transformation in the information-based war requires higher accuracy in POL support command. But my army is at a primary moment in research of modeling the POL support as the reason of lack of successful experience in the POL support of the information-based war. The thesis firstly puts forward to the study of POL service model. The study is important military significance in improving the accuracy of POL service in the informational war. The paper based in the anti-independence of Taiwan designs the corresponding POL support command chains, sets up the model system in POL support command, bring forward the detailed thought in exploiting the POL support command decision system based on multi-agent. The innovation of POL service theory is realized.②The oil transportation is the most important tache of POL support command chain in wartime. Its fail or success refers to the release of POL service system ability. The paper analyzes the characteristic of oil transportation in wartime. Aiming at firstly improving military effect and secondly improving economic benefit, the paper firstly builds fuzzy programming model through modifying traditional oil transportation based on the fuzzy and dynamic trait of oil transportation. Secondly, the paper sets up model of single oil warehouse supporting the army, designs the corresponding genetic algorithm. At last, the paper sets up guide lines of evaluating transportation path based on the trait of oil transportation in wartime, builds optimized model of appraising oil transportation path based on fuzzy triangle figure superior to traditional appraising model, and the model is better in the person of commander’s intention, presses close to the requirement of the real war.③Aiming at the difficulties in the exploitation of the POL support command intelligent decision system of the island mode and chimney mode, the problems is appeared.1) The software’s cooperation and integration is very difficult for the system is exploited by different programmer and adopting different exploiting flat.2)The POL support problem becomes very complex for the war is unpredicted, such as choosing oil resource and dynamic service circumstance. The paper has set up the POL support command model which automatically adapts different war-circumstance based on agent and multi-agent theory. The paper firstly packs requirement forecasting model, oil layout model, oil transportation model and oil strategic support model as single agent. The paper designs the frame of POL service command decision system based on multi-agent, sets up the solving process of POL service problem based on multi-agent, builds model of POL service command decision system. The study provides a new way for exploiting POL service command decision system.更多还原