节点文献

基于Lyapunov Drift-Plus-Penalty技术的施工队列多目标决策优化模型分析

Analysis of Multi Objective Decision Optimization Model of Construction Queue Based on Lyapunov Drift Plus Penalty Technology

【作者】 李伟

【导师】 何威;

【作者基本信息】 燕山大学 , 土木工程, 2021, 硕士

【摘要】 当前建设项目施工管理正随着前沿科技战略的发展和生态能源政策的覆盖发生多元化的改变,既是技术更新所带来的新机遇,亦是面临生态能源约束所带来的新挑战。目前人工智能技术正处于热化期,其主要用途是通过计算机结合智能算法实现对社会、工作和生活中所遇问题的决策,其中施工期多目标策略优化是重点研究的方向之一。施工期能耗和工期长久以来是业内专家学者关注的热点问题,由能耗和工期优化后所产生的收益更是业内管理者长期追求的目标,然而施工期难以避免的突发事件对能耗、工期及其收益会产生不同程度的影响。因此,在考虑施工期随机性事件发生的前提下,将施工过程作业转化为数据传输队列,并且基于Lyapunov优化理论设计了一种处理施工期多目标的寻优模型。首先,基于本文的研究方向,查阅国内外文献并进行系统性对比分析,重点研究施工期多目标之间的权衡关系;学习Lyapunov Drift-Plus-Penalty算法的适用环境、工作机理和参数调配;掌握计算模型的仿真流程、相关软件的操作环境和研究数据的预处理等前提基础。其次,构建施工期队列网络多目标优化模型的理论框架和过程指示性参数,描述施工期过程机理,以此构造施工期随机性事件的影响参数b_i(t),并定义其影响等级,随后构建施工期能耗、工期和收益数学模型,将施工期数据以单位t内的施工阶段为载体进行处理。在上述过程的基础上引入Lyapunov二次函数,经过漂移过程后进入惩罚阶段,引入非负惩罚参数V,构造虚拟施工队列,得到加权和最小化的目标函数。最后,本研究以秦皇岛市北戴河区某施工项目为试验对象,根据施工队列模型的内容进行案例仿真,为了清楚地表达施工期随机性的建模思想,以施工期天气因素为例进行分析,利用MATLAB和Python等软件进行计算机端的编程和运算,并且根据仿真结果进行详细的分析。

【Abstract】 At present,the construction management of construction projects is undergoing diversified changes with the development of cutting-edge science and technology strategy and the coverage of ecological energy policy.It is not only a new opportunity brought by technology update,but also a new challenge brought by ecological energy constraints.At present,artificial intelligence technology is in the hot period.Its main purpose is to realize the decision-making of social,work and life problems through computer combined with intelligent algorithm.Among them,multi-objective strategy optimization in construction period is one of the key research directions.Energy consumption and construction period have long been the hot issues concerned by experts and scholars in the construction industry.The benefits generated by the optimization of energy consumption and construction period are the long-term goals pursued by industry managers.However,the inevitable emergencies in the construction period will have different degrees of impact on energy consumption,construction period and its benefits.Therefore,under the premise of considering the random events during the construction period,the construction process operations are transformed into data transmission queues,and based on the Lyapunov optimization theory,a multi-objective optimization algorithm is designed to deal with the construction period.First of all,based on the research direction of this paper,consulting and sorting domestic and foreign literature and systematic comparative analysis,focusing on the trade-off relationship between multi-objective in construction period.Learn the applicable environment,working mechanism and parameter allocation of Lyapunov drift plus penalty algorithm.Master the simulation process of the calculation model,the operating environment of the related software and the preprocessing of the research data.Secondly,the theoretical framework and process indicative parameters of the queue multi-objective optimization model in the construction period are constructed to describe the process mechanism in the construction period,so as to construct the influence parameter b_i(t)of the random events in the construction period and define the influence level.Then,the mathematical models of energy consumption,duration and income in the construction period are constructed,and the data in the construction period are processed with the construction stage within unit t as the carrier.On the basis of the above process,Lyapunov quadratic function is introduced to enter the penalty stage after the drift process.The non negative penalty parameter V is introduced to construct the virtual construction queue,and the weighted and minimized multi-objective function is obtained.Finally,this study takes a construction project in Beidaihe District of Qinhuangdao as the test object,and carries out case simulation according to the contents of construction queue model.In order to express the modeling idea of randomness in construction period,the paper takes the weather factors in construction period as an example,and uses MATLAB and Python software to program and calculate the computer end,and carries out detailed simulation according to the simulation results analysis.

  • 【网络出版投稿人】 燕山大学
  • 【网络出版年期】2022年 01期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络