基于案例推理的航次决策方法及应用研究
|
摘要
世界经济全球化带动了物流、运输行业的快速发展,货物运输更多地依赖于海上运输,这给远洋运输事业的发展带来了极大契机。但时至今日,全球航运市场已趋于饱和,运力供过于求。日益激烈的竞争形势促使船舶运营者们通过各种途径不断地降低运输成本。在航次租船和航次期租方式下,船舶运营者欲在市场上揽取货源,必需准确估算航次成本和制定航次方案。航次决策的能力和针对市场需求的反应速度是竞争取胜的关键,简单地做出决策很有可能给企业带来巨大的经济损失。而大量的历史方案中蕴含了大量的经验性知识,将其合理地应用于当前决策中可以给船舶运营者提出指导性的建议,提高航次决策的准确性。
本文的研究是将有效利用历史航次决策方案中的经验知识为当前航次决策过程所用作为最终目标。因此,本文遵循基于案例推理理论的核心思想,详细研究了基于案例推理的航次决策方法。首先分析已有航次决策过程及数学模型,构建航次决策案例库及索引。其次设计有效的案例检索及案例推理算法,对航次决策模型的三类经验型输入进行推理。其中,直接经验型输入采用统计案例中的相关取值并进行数学计算的方法获取经验值作为对理论值的修正。燃油单价输入采用时间序列预测法做出预测。船吊使用情况,加油策略作为求解复杂的方案性输入,通过构造一定的子数学模型,设定约束条件单独求解,并进一步利用案例推理获得的经验值对于子模型求解获得的理论值进行修正。然后利用获得的所有输入求解航次决策模型获取最优的航次决策方案。最后,本文将研究的基于案例推理的航次决策方法与Uniwell(H.K.)公司的实际应用相结合,通过实际业务数据的操作,分析、验证决策方法的有效性。
本课题的研究为类似Uniwell(H.K.)这样的远洋运输公司提供了航次决策阶段科学的航次决策方法,能够帮助提升远洋运输公司的决策能力,使其在市场低迷、燃油成本大幅上涨的不利情况下降低风险提升收益。综上所述,论文的研究成果具有较高的理论意义和实用价值。
Globalization of world economy led to the rapid development of logistics and transportation industry. Cargo transportation mostly depends on maritime transport, which brings great chance to ocean shipping. But today, the global shipping market has become saturated with excess of supply of transport capacity over demand. Increasingly fierce competition prompts voyage charterers to reduce transportation cost in a variety of way. In a voyage charter mode, it`s necessary to accurately estimate costs and develop voyage program for charterers in order to canvass for cargo. Capability of voyage decision and speed of response for the market demand are the keys to competition victory. Simple decision may bring about huge economic losses to enterprises. Rational use of plentiful experiential knowledge among large numbers of historical cases can put forward directional proposals and improve the accuracy of voyage decision-making.
The study is taking utilizing experiential knowledge in historical voyage decision programme for current cases effectively as the ultimate goal. Therefore, following the theory of CBR, the thesis focuses on the research of CBR-based voyage decision method. Firstly, it analyzes the voyage decision-making process and existed model, builds case base and sets the index. Secondly, it designs effective algorithm of case retrieval and case reasoning, deducing three types of experiential input of the voyage decision model. The way of obtaining direct experiential input is to attain the experiential value by counting up and calculating related values in cases as the amendment to the theoretical value. The fuel price input is predicted by time series method. As complex programmatic input, whether to use gear and bunkering strategy are solved separately by constructing a sub mathematical model of certain constraints. The theoretical value from sub model is corrected by the experiential value in cases. Thirdly, all the values above are inputed to the voyage decision model to obtain the optimal scheme. Finally, the thesis integrates the CBR-based voyage decision method and the practical application in Uniwell (H.K.) Company, analyzing and ensuring the validity of the decision method by implementing the operation data.
Research of this subject provides ocean shipping companies like Uniwell (H.K.) with scientific voyage decision method in voyage decision phase; helps promote their decision-making ability and makes them increase revenue with a lower risk in bad status such as depressed market and high bunkering cost. Generally, the solutions have higher theoretic meanings and practice values.
本文的研究是将有效利用历史航次决策方案中的经验知识为当前航次决策过程所用作为最终目标。因此,本文遵循基于案例推理理论的核心思想,详细研究了基于案例推理的航次决策方法。首先分析已有航次决策过程及数学模型,构建航次决策案例库及索引。其次设计有效的案例检索及案例推理算法,对航次决策模型的三类经验型输入进行推理。其中,直接经验型输入采用统计案例中的相关取值并进行数学计算的方法获取经验值作为对理论值的修正。燃油单价输入采用时间序列预测法做出预测。船吊使用情况,加油策略作为求解复杂的方案性输入,通过构造一定的子数学模型,设定约束条件单独求解,并进一步利用案例推理获得的经验值对于子模型求解获得的理论值进行修正。然后利用获得的所有输入求解航次决策模型获取最优的航次决策方案。最后,本文将研究的基于案例推理的航次决策方法与Uniwell(H.K.)公司的实际应用相结合,通过实际业务数据的操作,分析、验证决策方法的有效性。
本课题的研究为类似Uniwell(H.K.)这样的远洋运输公司提供了航次决策阶段科学的航次决策方法,能够帮助提升远洋运输公司的决策能力,使其在市场低迷、燃油成本大幅上涨的不利情况下降低风险提升收益。综上所述,论文的研究成果具有较高的理论意义和实用价值。
Globalization of world economy led to the rapid development of logistics and transportation industry. Cargo transportation mostly depends on maritime transport, which brings great chance to ocean shipping. But today, the global shipping market has become saturated with excess of supply of transport capacity over demand. Increasingly fierce competition prompts voyage charterers to reduce transportation cost in a variety of way. In a voyage charter mode, it`s necessary to accurately estimate costs and develop voyage program for charterers in order to canvass for cargo. Capability of voyage decision and speed of response for the market demand are the keys to competition victory. Simple decision may bring about huge economic losses to enterprises. Rational use of plentiful experiential knowledge among large numbers of historical cases can put forward directional proposals and improve the accuracy of voyage decision-making.
The study is taking utilizing experiential knowledge in historical voyage decision programme for current cases effectively as the ultimate goal. Therefore, following the theory of CBR, the thesis focuses on the research of CBR-based voyage decision method. Firstly, it analyzes the voyage decision-making process and existed model, builds case base and sets the index. Secondly, it designs effective algorithm of case retrieval and case reasoning, deducing three types of experiential input of the voyage decision model. The way of obtaining direct experiential input is to attain the experiential value by counting up and calculating related values in cases as the amendment to the theoretical value. The fuel price input is predicted by time series method. As complex programmatic input, whether to use gear and bunkering strategy are solved separately by constructing a sub mathematical model of certain constraints. The theoretical value from sub model is corrected by the experiential value in cases. Thirdly, all the values above are inputed to the voyage decision model to obtain the optimal scheme. Finally, the thesis integrates the CBR-based voyage decision method and the practical application in Uniwell (H.K.) Company, analyzing and ensuring the validity of the decision method by implementing the operation data.
Research of this subject provides ocean shipping companies like Uniwell (H.K.) with scientific voyage decision method in voyage decision phase; helps promote their decision-making ability and makes them increase revenue with a lower risk in bad status such as depressed market and high bunkering cost. Generally, the solutions have higher theoretic meanings and practice values.
引文
[1] Si-Hwa Kim, Kyung-Keun Lee. An optimization-based decision support system for ship scheduling [J]. Computers & Industrial Engineering. 1997, 33: 689~692.
[2] Kjetil Fagerholt A computer-based decision support system for vessel fleet scheduling - experience and future research [J] . Decision Support System. 2004, 37: 35~47.
[3]屠琳桓.航次估算及应用思路的探讨[J].广州航海高等专科学校学报. 2008, 16(1): 29~31.
[4]姜红,王戈梁.浅析航次估算[J].世界海运. 2004, 27(3): 29~30.
[5]方芳,张俊晓.不定期船航次估算计算机辅助系统[J].水运管理. 1997(4): 10~13.
[6]孙绍伟,董妍慧,姜阳.一种航次货载最优选择的算法设计[J].水运管理. 2006, 28(8): 1~4.
[7]江焕平.远洋集装箱运输船舶航次运行费用的控制[J].中国水运. 2006, 6(10): 8~9.
[8] Ian Watson, Applying Case-Based Reasoning: Techniques for Enterprise Systems [M], Morgan Kaufmann Publishers, 1997.
[9] B. Porter, R. Bareiss. PROTOS: An Experiment in Knowledge Acquisition for Heuristic Classification Tasks [A], In: Proc. of First International Meeting on Advances in Learning (IMAL) [C], France: 159~174.
[10] K. Ashley. Modeling Legal Argument: Reasoning with Cases and Hypothetical’s Cambridge [M], MIT Press, 1990.
[11] C. Skalak, E. Rissland. Argument and Cases: An inevitable twining [J]. Artificial Intelligence and Law 1(1):3~48.
[12] D.W Aha, Ian Watson, Case-based reasoning research and development [A]. In: Proceeding of ICCBR` 2001.
[13] Pal K, Campell JA. A hybrid system for decision-making about assets in English divorce case[A].In: Proc of First United Kingdom Workshop on CBR [C], 1995: 152~156.
[14] Lorcan Coyle, Dónal Doyle, Pádraig Cunningham. Representing similarity for CBR in XML [C]. Advances in Case-Based Reasoning (Proc of ECCBR-04). Madrid, Spain: Springer, 2004: 119~127.
[15]张光前,邓贵仕,李朝晖.基于事例推理的技术及其应用前景[J].计算机工程与应用. 2002: 38(20): 52~55
[16]张英菊,仲秋雁,叶鑫,裘江南,曲晓飞. CBR的应急案例通用表示与存储模式[J].计算机工程. 2009, 35(17): 28~30.
[17]李玲娟,汤文宇,王汝传.基于XML的案例表示和案例库构造方法[J].计算机应用研究. 2007, 24(11): 70~73.
[18]张鹏程.基于语义的复杂产品案例库和推理技术研究[D].南京:南京理工大学.2008.
[19] Rissland E L, Ashley K D. Credit assignment and the problem of competing factors in case-based reasoning [C]. In: Kolodner ed. Case-Based Reasoning, Proceedings from a Workshop, Clearwater Beach, Florida, Morgan-Kauffman Publ, 1988
[20] Surma J. Enhancing similarity measures with domain specific knowledge [C]. In: Preprints Second European Workshop on Case-Based Reasoning, Acknosoft Press, 1994:365~372
[21]张光前,邓贵仕,王瑾.变权值下的最近相邻检索策略[J].计算机工程与应用. 2003:39(11): 40~42.
[22] Dubois D, Esteva F, Garcia P. Fuzzy set modelling in case-based reasoning [J]. International Journal of Intelligent Systems, 1998,13: 345~373
[23] Purvis L, Athalye S. Towards improving case adaptability with a genetic algorithm [C]. In: Proceedings of the second International Conference on Case-based Reasoning, Lecture Notes in Artificial Intelligence 1266, 1997: 403~412
[24]李海芳,魏晓艳,陈俊杰.多维优化案例推理检索算法研究[J].计算机工程与应用. 2008, 44(25): 157~160.
[25]赖大荣.基于人工神经网络的案例推理研究[D].合肥:合肥工业大学. 2006.
[26] Aamodt A. Plaza E. Case-based reasoning: foundational issues, methodological variations and system approaches [J]. AI Communications, 7(1): 39~59.
[27] Kolodner J L. Improving human decision making the rough cased-based reasoning techniques [J]. AI Magazine, 1991:12(3):52~59.
[28] Hinrichs T.R., Kolodner J. The roles of adaptation in case-based design [C]. In: Proceedings of Case-based Reasoning Workshop, Washington, 1991: 121~132.
[29] Hanney K, Keane M T. The adaptation knowledge bottleneck: How to easy it by learning from cases [C]. In: Proceedings of the Second International Conference on Case-based Reasoning, Lecture Notes in Artificial Intelligence 1266, Springer-Verlag, 1997: 359~370.
[30] Leake D B, Kinley A, Wilson D. A case study of case-based CBR [C]. In: Proceedings of the second International Conference on Case-based Reasoning, Lecture Notes in Artificial Intelligence 1266, Springer-Verlag, 1997: 371~382.
[31] Schweizer B, Sklar A. Associative functions and abstract semi-groups [M]. Publications Mathematic Debrecen, 1963-10: 69~81.
[32]段军,戴居丰.案例修正方法研究[J].计算机工程. 2006, 32(6): 1~3.
[33]吴晓,种玉珍,倪红波,王海鹏.一种CBR与RBR相结合的智能家庭推理系统[J].计算机应用研究. 2009, 26(3): 977~979.
[34]倪志伟,蔡庆生,贾瑞玉.用神经网络来实现基于范例的推理系统[J].计算机工程. 2002, 28(7): 12~14.
[35]高慧颖,颜志军.基于CBR的信息系统案例检索多Agent模型研究[J].计算机工程与设计. 2008,29(5): 1226~1228.
[36]王君,潘星,李静,刘鲁.基于案例推理的知识管理咨询系统[J].清华大学学报(自然科学版). 2006, 46(s1): 990~995.
[37]于江有,张艳青,王知衍.结合CBR和工作流的地铁应急演练系统[J].微计算机信息(管控一体化). 2009, 25(3-3): 18~20.
[38]姜丽红,徐博艺,席俊红.基于案例推理的过滤算法及智能信息推荐系统[J].清华大学学报(自然科学版). 2006, 46(s1): 1074~1077.
[39] Aamodt A, Plaza E. Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches [J], Artificial Intelligence Communications. 1994, 7: p39~p59.
[40] Gavin R. Finnie, Zhaohao Sun. R5 model for case-based reasoning [J]. Knowledge-Based System 16(1): 59-65(2003).
[41] Ramon Lopez de Mantaras. Case-based Reasoning [M]. Springer-Verlag Heidelberg, 2001
[42]王凯.基于案例推理的应急管理案例库构建方法研究[D].上海:上海交通大学.2009
[2] Kjetil Fagerholt A computer-based decision support system for vessel fleet scheduling - experience and future research [J] . Decision Support System. 2004, 37: 35~47.
[3]屠琳桓.航次估算及应用思路的探讨[J].广州航海高等专科学校学报. 2008, 16(1): 29~31.
[4]姜红,王戈梁.浅析航次估算[J].世界海运. 2004, 27(3): 29~30.
[5]方芳,张俊晓.不定期船航次估算计算机辅助系统[J].水运管理. 1997(4): 10~13.
[6]孙绍伟,董妍慧,姜阳.一种航次货载最优选择的算法设计[J].水运管理. 2006, 28(8): 1~4.
[7]江焕平.远洋集装箱运输船舶航次运行费用的控制[J].中国水运. 2006, 6(10): 8~9.
[8] Ian Watson, Applying Case-Based Reasoning: Techniques for Enterprise Systems [M], Morgan Kaufmann Publishers, 1997.
[9] B. Porter, R. Bareiss. PROTOS: An Experiment in Knowledge Acquisition for Heuristic Classification Tasks [A], In: Proc. of First International Meeting on Advances in Learning (IMAL) [C], France: 159~174.
[10] K. Ashley. Modeling Legal Argument: Reasoning with Cases and Hypothetical’s Cambridge [M], MIT Press, 1990.
[11] C. Skalak, E. Rissland. Argument and Cases: An inevitable twining [J]. Artificial Intelligence and Law 1(1):3~48.
[12] D.W Aha, Ian Watson, Case-based reasoning research and development [A]. In: Proceeding of ICCBR` 2001.
[13] Pal K, Campell JA. A hybrid system for decision-making about assets in English divorce case[A].In: Proc of First United Kingdom Workshop on CBR [C], 1995: 152~156.
[14] Lorcan Coyle, Dónal Doyle, Pádraig Cunningham. Representing similarity for CBR in XML [C]. Advances in Case-Based Reasoning (Proc of ECCBR-04). Madrid, Spain: Springer, 2004: 119~127.
[15]张光前,邓贵仕,李朝晖.基于事例推理的技术及其应用前景[J].计算机工程与应用. 2002: 38(20): 52~55
[16]张英菊,仲秋雁,叶鑫,裘江南,曲晓飞. CBR的应急案例通用表示与存储模式[J].计算机工程. 2009, 35(17): 28~30.
[17]李玲娟,汤文宇,王汝传.基于XML的案例表示和案例库构造方法[J].计算机应用研究. 2007, 24(11): 70~73.
[18]张鹏程.基于语义的复杂产品案例库和推理技术研究[D].南京:南京理工大学.2008.
[19] Rissland E L, Ashley K D. Credit assignment and the problem of competing factors in case-based reasoning [C]. In: Kolodner ed. Case-Based Reasoning, Proceedings from a Workshop, Clearwater Beach, Florida, Morgan-Kauffman Publ, 1988
[20] Surma J. Enhancing similarity measures with domain specific knowledge [C]. In: Preprints Second European Workshop on Case-Based Reasoning, Acknosoft Press, 1994:365~372
[21]张光前,邓贵仕,王瑾.变权值下的最近相邻检索策略[J].计算机工程与应用. 2003:39(11): 40~42.
[22] Dubois D, Esteva F, Garcia P. Fuzzy set modelling in case-based reasoning [J]. International Journal of Intelligent Systems, 1998,13: 345~373
[23] Purvis L, Athalye S. Towards improving case adaptability with a genetic algorithm [C]. In: Proceedings of the second International Conference on Case-based Reasoning, Lecture Notes in Artificial Intelligence 1266, 1997: 403~412
[24]李海芳,魏晓艳,陈俊杰.多维优化案例推理检索算法研究[J].计算机工程与应用. 2008, 44(25): 157~160.
[25]赖大荣.基于人工神经网络的案例推理研究[D].合肥:合肥工业大学. 2006.
[26] Aamodt A. Plaza E. Case-based reasoning: foundational issues, methodological variations and system approaches [J]. AI Communications, 7(1): 39~59.
[27] Kolodner J L. Improving human decision making the rough cased-based reasoning techniques [J]. AI Magazine, 1991:12(3):52~59.
[28] Hinrichs T.R., Kolodner J. The roles of adaptation in case-based design [C]. In: Proceedings of Case-based Reasoning Workshop, Washington, 1991: 121~132.
[29] Hanney K, Keane M T. The adaptation knowledge bottleneck: How to easy it by learning from cases [C]. In: Proceedings of the Second International Conference on Case-based Reasoning, Lecture Notes in Artificial Intelligence 1266, Springer-Verlag, 1997: 359~370.
[30] Leake D B, Kinley A, Wilson D. A case study of case-based CBR [C]. In: Proceedings of the second International Conference on Case-based Reasoning, Lecture Notes in Artificial Intelligence 1266, Springer-Verlag, 1997: 371~382.
[31] Schweizer B, Sklar A. Associative functions and abstract semi-groups [M]. Publications Mathematic Debrecen, 1963-10: 69~81.
[32]段军,戴居丰.案例修正方法研究[J].计算机工程. 2006, 32(6): 1~3.
[33]吴晓,种玉珍,倪红波,王海鹏.一种CBR与RBR相结合的智能家庭推理系统[J].计算机应用研究. 2009, 26(3): 977~979.
[34]倪志伟,蔡庆生,贾瑞玉.用神经网络来实现基于范例的推理系统[J].计算机工程. 2002, 28(7): 12~14.
[35]高慧颖,颜志军.基于CBR的信息系统案例检索多Agent模型研究[J].计算机工程与设计. 2008,29(5): 1226~1228.
[36]王君,潘星,李静,刘鲁.基于案例推理的知识管理咨询系统[J].清华大学学报(自然科学版). 2006, 46(s1): 990~995.
[37]于江有,张艳青,王知衍.结合CBR和工作流的地铁应急演练系统[J].微计算机信息(管控一体化). 2009, 25(3-3): 18~20.
[38]姜丽红,徐博艺,席俊红.基于案例推理的过滤算法及智能信息推荐系统[J].清华大学学报(自然科学版). 2006, 46(s1): 1074~1077.
[39] Aamodt A, Plaza E. Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches [J], Artificial Intelligence Communications. 1994, 7: p39~p59.
[40] Gavin R. Finnie, Zhaohao Sun. R5 model for case-based reasoning [J]. Knowledge-Based System 16(1): 59-65(2003).
[41] Ramon Lopez de Mantaras. Case-based Reasoning [M]. Springer-Verlag Heidelberg, 2001
[42]王凯.基于案例推理的应急管理案例库构建方法研究[D].上海:上海交通大学.2009