【作者】 樊敏；

【导师】 李远富；

【作者基本信息】 西南交通大学 ， 项目管理， 2013， 博士

【摘要】 大型复杂工程项目无论开始阶段还是建设阶段,乃至项目运营阶段,最重要也是最频繁的活动之一是决策。决策包括信息采集、建模和分析等关键技术。决策就是从多个选择中选出具有相当合理性的一项,其过程通常是广义上的解决问题的过程,有关复杂系统(CS)的决策是各专业的决策者面临的一个共同问题。尽管各种决策的设置中包括典型的不完整、不明确的信息,专家们常常还是要据此作出好的决策,这也就是为什么需要跨学科的方法进行研究。与众多的技术、社会复杂性交织在一起,使本来就复杂的系统在决策过程中面临诸多困难,例如信息获取、决策模型制定,决策结果评价等等。多数现实生活中的大型工程问题都可看做是复杂系统,工程本身和管理过程拥有一些特定的特征,通过分析大型复杂工程项目的技术复杂性、社会人干系复杂性,我们得知,由于各种资源的限制、管理体制的僵化以及许多复杂的条例,重大工程项目通常在运营管理上受到严重的制约,为此我们需要寻求一种新的管理模式运用到实际工程当中。“重大工程科学决策、管理创新与可持续发展”成为当下本学科主要的议题,当前需要从重大工程管理实施中面临的实际问题出发,深入开展重大工程认识论与管理方法论、重大工程管理模式与决策机制、重大工程组织创新与精益建造、重大工程国际化战略、重大工程全寿命周期与可持续发展等主题研究。人工免疫系统(AIS)是一个自适应系统,受生物免疫学理论的启发,借鉴免疫功能、原则和模型,应用于解决实际问题的系统。在人工免疫算法中,被求解的问题视为抗原,抗体则对应于问题的解,改进的人工免疫算法与GA相似,一些人工免疫算法也是从随机生成的初始解群出发,采用复制、交叉、变异等算子进行操作,产生比父代优越的子代,这样循环执行,逐渐逼近最优解。本研究的主要贡献如下：(1)将生物学免疫原理运用到大型复杂项目决策过程中,分析了大型复杂项目决策信息获取存在争议的原因,隐喻免疫原理中抗体更新的过程,通过免疫系统识别自我非我的能力将决策者关心的信息从大量信息中提取,用以区分表示不同用户或不同时间段内决策者对信息的不同兴趣,起到信息过滤的作用。借鉴免疫过程中抵御有害抗原的算法,通过免疫计算估计人工免疫系统中的能量误差,排除运营维护期间同类解决方案中有较大风险的方案,改善了大型复杂工程项目面临决策复杂困难的机制,大型复杂工程项目的运营期多目标决策得到了生物免疫原理中免疫算法的成功隐喻,对各多目标决策优化问题的解决展现了免疫算法自适应的优势。(2)将多目标决策方法与免疫原理紧密结合,创新性的运用了免疫理论相关算法解决了大型复杂工程项目运营决策中涉及的资源受限优化问题,在任务工期预先指定最后期限内使得总资源成本最小化,计算方法和理念使得资源紧张的约束条件下项目调度问题的各种精确解和启发式程序得到了发展。(3)借鉴生物免疫的主动防御功能,运用危险理论解决了船舶撞桥的风险预测,根据获取的航行数据由危险模式的免疫算法给出危险信号,启动应急准备,改变了目前“应急”决策的被动模式,对大型复杂工程项目在突发事件状态下的应急决策问题,以实例给出了计算步骤,为运营管理者提供了主动的避险决策技术支持,转变了大型桥梁管理方被动接受损失的尴尬处境。(4)开发了基于多理论、多模型、多目标的大型复杂工程项目运营期智能辅助决策支持系统,并首次运用到实际大型跨海桥梁的运营维护决策管理工作中,根据养护管理需求建立的专业数据库,解决了数据来源多样性难题,研究建立了桥梁养护管理多决策模型的组合,突破了单一模型决策的局限,为决策者提供了技术支持,转变了运营管理工作模式,提升了管理效率,并在其他大型桥梁运营管理中得到了较好的应用推广,经专家鉴定,该项研究成果填补了国内的行业空白,成果达到了国际先进水平。最后,研究提出了关于免疫原理在实际应用中有待改进的建议和思路,为他人的研究提供参考。更多还原

【Abstract】 Decision-making is one of the most important and frequent active in large and complex projects with its Start-up phase, construction phase, and the project operational phase. Decision-making’s key technologys include the information gathering, modeling, and analysis. Decision-maker selected one from the multiple choice with considerable justification, the process is usually on the generalized problem-solving process, complex systems (CS) decision is a common problem of professional policy makers. Despite the various policy settings include typical incomplete and unclear information, but experts want to make good decisions, which is why the need for an interdisciplinary approach to research. With many intertwined technical, social complexity, complex systems is faced many difficulties in the decision-making process, such as access to information, decision-making model developed, result assessment, and so on.Mega project question may be as a complex system in real life, project itself and the management process has some specific characteristics, through analysis large-scale complex engineering project technical complexity, and the social stakeholder complexity, we knew, the important engineering project usually receives the serious restriction in the operation management, because of result of resources limited, the management system ossification as well as many complex rules. We need to seek one kind of new management pattern to utilize the actual project. Science decision-making of important engineering, management innovation and sustainable development immediately become a main subject in this field. We need face the actual problem of the important project management implementation, develop the important project epistemology and the management methodology, and the important project management pattern, important project subject research and so on organization innovation and fine profit construction, important project internationalization strategy, important project total life cycle and sustainable development.Artificial immune system (AIS) is an adaptive system, inspired by the biological theory of Immunology, immune function, principles and models for reference, should be used to solve practical problems of the system. Artificial immune algorithm, solving issues as antigens, antibodies was correspond to the solutions of problem, improved artificial immune algorithm similar to GA, some artificial immune algorithm is also starting from the randomly generated initial solution group, executing operation through copy, crossover and mutation, The new population surpasses the father generation, and is gradually approach the optimal solution.This research main contribution is as follows:(1) Utilized the biology immunity principle in the project complex project decision-making process, analyzed the reason of decision information existence dispute in its gather progressing, in metaphor the antibody update progress of immunity principle, extract the information which decision-makers concerned with, from the large amount of information by the immune system to recognize the ability of self-nonself, to distinguish between the different interests of different users or different periods, and decision-makers, play the role of information filtering, draw against harmful antigens in the immune process algorithm to calculate the estimated energy error of artificial immune system, excluded the solution which withing greater risk in similar solutions of operation and maintenance. Improved the situation of difficult decision-making of large-scale complex engineering project. the target decision of operation of mega projects drew on immune algorithm of biological immunity principle, problem solving of optimized for multi-objective decision making demonstrated the advantages of immune algorithm.(2) Closely integrated multi-objective decision-making methods and immune principle, innovatively used the immune theory algorithm to solve the optimization problem of limited resources involved in mega projects operational decisions, so that the total minimum cost of resources within the pre-specified deadline task duration, the heuristic procedures of project scheduling problem under tight resource constrained have been developed with calculation methods and philosophy(to see Chapter6section2and3).(3) First learn from active defense capabilities of biological immune system, used the Danger Theory to solve the the emergency decision problem of large and complex projects in the state of emergency, send the danger signal according to the gain navigation data by the immunity algorithm dangerous pattern, changed the passive mode of emergency decision, given the calculation procedure of avoid the ship collision with bridge, provided proactive hedging decision support for operational managers, give solution of the risk of preventing ship collision with bridges, change the awkward position of passive acceptance of loss for the manager of large bridges.(4) Developed an intelligent decision support system based on multi-theory, multi-model, multi-objective for operation period of large and complex projects, and for the first time applied to actual large-scale cross-sea bridge’s management of operation and maintenance decision-making, according to the conservation and management needs to establish professional database and solve the diversity problem of data sources to study the establishment of a decision-making model of bridge maintenance and management of multi-combination, broke through the limitations of a single model of decision-making, provided technical support for the decision-makers, changes the mode of operation and management, and enhance management efficiency, and better promote the application of the operational management of other large bridges, the experts identified the research filled the domestic blank, the results reached the international advanced level.Finally, the researchers put forward suggestions and ideas on the improvement of immune principle, to provide a reference for others in practical applications.更多还原