摘要：针对城市应急物流中心选址问题,建立了考虑覆盖率、总时间成本的多目标0~1整数模型和评价模型,并利用深度优先搜索法和模糊神经网络法来求解该模型。为验证所提出的模型及方法,以南京市江宁区为例,设计了算例实验,同时将多目标模型和单目标模型进行对比,分析多目标模型对选址方案的影响,并对选址方案及相关影响因素进行评价,最终实验结果验证了模型和方法的有效性,可为城市突发事件应急设施选址提供参考依据。

摘要：基于后装保障基地选址多目标特性,提出了保障水平—保障距离阶跃性、保障加权距离等概念,构建了多目标混合整数规划选址模型;针对模型目标函数多、维度大、难以计算等特点,设计了一种二进制多目标粒子群算法(BMOPSO),并进行算法效率分析;结合算例分析了具体选址问题,计算得到了一组帕累托选址方案组合和帕累托前沿。该方法可根据决策者偏好为其提供选址决策参考。

摘要：本文提出了一种高维集结算法来对多目标选址中的目标函数进行分析,并用其解决新能源设施选址问题。本文算法分为三个步骤:首先,将区域内所有场所定义为决策变量,计算出所有子目标函数。将决策变量及子目标函数投影为高维空间上的需求点,每一个维度对应一个子目标函数值,高维需求点之间的欧氏距离体现了不同场所在多目标优化环境下的差异。然后,设计高维集结算法n-DFPSA (n维设施点集结算法)寻找高维空间中的Pareto中位点,该点到所有高维设施点之间欧式距离之和最小,根据高维设施点与高维Pareto点之间的欧氏距离对设施进行赋权。最后,将所有场所根据其经纬度及权重信息投影为二维平面上的加权需求点,利用2-DFPSA (平面设施点集结算法)寻找二维平面中的加权Steiner点,该点即为新能源设施的合理选址位置。文章的最后,选用了一个工商储能站选址案例进行选址仿真实验,并与经典选址算法进行对比,对比结果证明了本文开发算法的优越性。The paper proposes a high-dimensional aggregation algorithm for analyzing the objective function in multi-objective location, which is applied to solve the problem of locating new energy facilities. The algorithm consists of three steps: Firstly, all facilities within the area are defined as decision variables and their corresponding sub-objective functions are calculated. These decision variables and sub-objective functions are then projected onto facility points in a high-dimensional space, where each dimension corresponds to a sub-objective function value. The Euclidean distance between these high-dimensional facility points reflects the differences among different facilities in the multi-objective optimization environment. Secondly, an n-DFPS (n-dimensional facility point seasonal algorithm) is designed to find the Pareto point in this high-dimensional space by minimizing the sum of Euclidean distances between that point and all other high-dimensional facility points. The facilities are weighted based on their Euclidean distances from both the high-dimensional facility point and the high-dimensional Pareto median point. Finally, all facility points are projected onto weighted facility points on a two-dimensional plane according to their latitude, longitude, and weight information;2-PGSA (2-dimensional facility point seasonal algorithm) is used to identify Steiner points on this plane as reasonable locations for new energy facilities. To validate its effectiveness, an industrial and commercial energy storage station location case study is conducted using simul