考虑多源不确定因素的建筑废弃物预测与控制方法研究

项目名称： 考虑多源不确定因素的建筑废弃物预测与控制方法研究

项目编号： No.71502008

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 管理科学

项目作者： 李亚帅

作者单位： 北京航空航天大学

项目金额： 17.5万元

中文摘要： 建筑工程中产生的数量巨大、种类繁多的废弃物受到其诸多不确定因素影响。为了更好地管理建筑废弃物，本课题以工程项目为单位探索在多源不确定因素影响下的建筑废弃物预测和控制方法。第一，利用物质流方法分析建筑材料的行为特性，探索建筑废弃物产生机理；通过文献调查和流程图分析方法识别影响建筑废弃物产生的不确定因素。第二，以不确定因素和待估计废弃物为节点，以它们之间的因果关联关系为有向边，利用专家知识和问卷调查方法获取先验概率和条件分布概率，构建贝叶斯网络模型来预测建筑废弃物的产生。第三，考虑不确定因素和建筑企业的废弃物管理偏好，以不确定性因素中的可控因素作为决策变量，以建筑废弃物管理偏好和控制成本为目标函数，构建多目标决策模型，采用启发式算法求解来控制建筑废弃物的产生。本课题以房屋建筑为例，通过调节可控不确定因素，控制建筑废弃物的产生，并通过实际工程项目予以验证和提高，对废弃物管理有启发意义。

中文关键词： 建筑废弃物预测和控制；多源不确定因素；物质流分析；贝叶斯网络；多目标优化

英文摘要： The significant amount of construction and demolition waste is heavily influenced by multiple uncertainties of construction projects. The aim of this proposal is to address construction waste estimation and control of one single project considering multiple uncertainties. First, based on the mass balance principle and work breakdown structure, material flow analysis is conducted for each terminal work package by considering characteristics of different materials to discover the mechanism of construction waste generation. At the same time, key factors influencing construction waste generation can be identified through literature review and workflow analysis. Second, a Beyesian network is constructed by taking uncertainties and amount of construction waste as nodes and their cause-effect relationships as arrows. The prior and conditional distribution can be obtained through expert opinion and structured questionnaire survey. An algorithm is also proposed to help the network forecast construction waste generation in uncertain environment. Third, preferences of construction companies towards construction waste management and waste generation estimation from the Bayesian network are incorporated to build a construction waste control model, which is a multi-objective optimization model taking management of construction waste & cost of management as objective functions and controllable uncertainties as decision variables. A heuristic method will be proposed to solve the problem to get the satisfied solution, which can help control construction waste. Finally, a case study will be conducted to verify the proposed models and facilitate its applications in the construction industry. In the future, a small number of projects will be selected for trial with the aim of improving the construction waste management performance.

英文关键词： Construction waste estimation and control;Multiple uncertainties;Material flow analysis;Bayesian network;Multi-objective optimization