Target setting in the general combined-oriented CCR model using an interactive MOLP method

The traditional data envelopment analysis (DEA) model does not include a decision maker’s (DM) preference structure while measuring relative efficiency, with no or minimal input from the DM. To incorporate DM’s preference information in DEA, various techniques have been proposed. An interesting method to incorporate preference information, without necessary prior judgment, is the use of an interactive decision making technique that encompasses both DEA and multi-objective linear programming (MOLP). In this paper, we will use Zionts-Wallenius (Z-W) method to reflecting the DM’s preferences in the process of assessing efficiency in the general combined-oriented CCR model. A case study will conducted to illustrate how combined-oriented efficiency analysis can be conducted using the MOLP method.     

具有偏好锥的面向输出的超效率DEA模型分析

鉴于传统DEA模型无法区分有效决策单元,超效率DEA模型未考虑决策者的偏好,现提出面向输出的权重受限的综合超效率DEA模型及其投影概念,并讨论该模型与其他超效率DEA模型之间的关系.接着,分析模型的最优目标函数值与决策单元有效性之间的关系,并讨论面向输出的权重受限的综合超效投影与多目标规划问题的非支配解之间的关系.最后,通过对中国西部12个地区工业企业科技创新效率综合评价,并与原有方法进行比较研究,得出本文方法更具优势和合理性.     

直觉模糊多目标二人零和矩阵对策

研究矩阵对策是深入研究对策理论的一个基本途径和重要手段。根据直觉模糊多目标决策和模糊对策理论,研究了支付值为直觉模糊值的多目标二人零和矩阵对策。首先介绍了基于直觉模糊集的多目标二人零和矩阵对策模型,接着提出了求解直觉模糊多目标二人零和矩阵对策的线性规划方法。最后以数例说明本文提出的方法。结果表明该方法能方便地得到对策的均衡策略和均衡解。     

多储备点多品种应急物资配送的多目标优化模型

应急救援的社会化、应急物资需求的多样性、应急物资需求和补给的时变性,对应急物流的配送调度提出了挑战.应急物流的紧急性要求最大程度保障受灾点的物资供应;在经济上则要求应急物流的成本最小化.通过将时间离散化为阶段序列,在应急物资需求和补给是可预测的情况下,建立一个多储备点、多物资品种、单受灾点的应急物资配送的多目标优化模型,来最小化应急物资短缺次数和运输成本.仿真实例表明,该模型可以通过优化引擎快速求解,能够发现导致短缺的应急物资品种和加强供应补给的时间区间.     

基于NSGA-II的应急生产任务多目标优化模型及算法研究

重大突发事件发生后,若灾区的应急物资需求不能通过调用储备得到满足,则应急生产将成为灾区应急物资供应的重要保障手段。本文研究重大突发事件发生后应急物资生产任务的优化问题,重点关注原材料生产能力变化对完成应急生产任务的影响,以应急生产任务完成时间最短、完成成本最低为决策目标,研究了包含多个供应商、多个制造商以及单个受灾点的应急物资生产任务多目标规划模型。运用在求解多目标规划问题时具有众多优势的非支配排序多目标遗传算法(NSGA-II)对模型进行求解。通过算例分析,NSGA-II可以得到较好的Pareto前沿,并且可以根据不同情况给出最优的应急物资生产和原材料保障方案。本文的研究还表明,要想更快完成应急生产任务,需要做好原材料、资金、电力、交通等各种要素的配套保障工作。     

投资组合模型的改进研究:基于企业社会责任视角的实证分析

在尊重和借鉴前人对企业社会责任研究,尤其是在企业社会责任评价研究基础之上,本文从投资者的角度在投资组合过程中研究企业社会责任。在Markowitz(均值—方差)理论模型上添加企业社会责任的三个一级指标期望作为目标函数,由此将传统的投资组合模型扩展为五个目标函数的投资组合选择模型,而且我们根据经济学中经典的效用函数理论证明了此模型的正确性。本文引入主流的企业社会责任评价标准,并对一些典型公司进行打分量化。在此基础之上建立了以期望回报率、回报率的方差、核心利益相关者期望、蛰伏利益相关者期望和边缘利益相关者期望为目标函数的投资组合选择模型,在最小方差曲面上选取10个点构造投资组合,并以样本外的数据验证了模型的有效性。研究发现:根据此模型计算出来的部分投资组合回报率显著高于同期的市场指数。研究结果表明,这种关注企业社会责任的多目标投资组合选择模型,不仅让投资者可以直接控制企业社会责任,而且实际数据证明了此模型的优势之处,从而为关注企业社会责任的投资者提供一种投资的方法和思路。     

An application of interactive multi-criteria optimization to air pollution control

In this paper a problem of air pollution control is studied, posing it as a multi-objective control problem of partial differential equations. The original problem, dealing with the optimal management of a set of industrial plants inside a populated area, is formulated by means of the diffusion transport equation, including a linear reaction term and source terms modelled by Dirac deltas. Introducing adjoint state techniques, the problem transforms into a problem of multi-objective optimization in Banach spaces, where the large number of objective functions discourages the complete search of its Pareto front. Therefore, in order to solve the problem, two interactive methods of multi-objective programming are proposed: the VIA and the STEM algorithms. Finally, the paper illustrates how to combine both algorithms to solve in a more effective way a realistic problem posed in the Metropolitan Area of Guadalajara (Mexico).     

双目标DEA模型及其在商业银行评价中的应用

为了解决传统DEA模型所存在的判别能力弱和不现实权重两个缺陷,本文提出了一种双目标DEA模型。该模型在相同的约束条件下,同时以最大化各自效率和最小化所有效率值差异为目标。最后,通过某商业银行分支机构效率评价实例对所提出的方法进行了演示,说明了所提出方法的优点,着重说明了该方法可以很好地解决传统DEA方法所存在的两个缺陷。     

Tabu Search Based Procedure for Solving the 0-1 MultiObjective Knapsack Problem: The Two Objectives Case

We consider in this paper the solving of 0-1 knapsack problems with multiple linear objectives. We present a tabu search approach to generate a good approximation of the efficient set. The heuristic scheme is included in a redu tion decision space framework. The case of two objectives is developed in this paper. TS principles viewed into the multiobjective context are discussed. According to a prospective way, several variations of the algorithm are investigate. Numerical experiments are reported and compared with available exact efficient solutions. Intuitive justifications for the observed empirical behavior of the procedure and open questions are discussed.