土方调配优化与自动分区研究

李迟典

摘 要

伴随着建筑行业的快速发展，建筑行业的竞争日益激烈，这对于建筑行业的发展水平提出了更高要求。土方工程是建筑工程的主要分部工程之一，对土方分区施工方案和土方调配施工方案进行研究，有助于提高建筑行业的发展水平。然而目前土方分区施工方案依靠项目经理经验确定，缺乏具体评价优化指标；土方调配施工方案通过线性规划模型进行确定，影响因素考虑不够全面。针对上述问题，本文以土方调配优化和自动分区为研究对象，主要包括以下工作：

首先，在现有研究基础上，本文通过分析土方调配过程，提出了土方调配成本模型并对模型进行了分析，利用模型计算了两个算例的土方成本，计算结果与商业软件结果对比误差分别为0.89%和0.22%，验证了模型的可用性。其次，根据土方调配成本模型的特点将土方调配优化划分为线性规划模型和非线性规划模型，前者利用单纯形法求解，求解结果与商业软件对比误差为0.79%；后者利用改进模拟退火算法进行求解，并通过与其他启发式算法的对比验证了改进模拟退火算法的求解性能。然后，分析人工划分土方施工分区的经验并建立土方自动分区问题模型，利用改进遗传算法对模型求解。当限制分区数量为实际分区数量时，通过与其他启发式算法对比说明改进遗传算法的优越性，相较人为划分结果可以节省4.56%的成本；在限制分区数量在5~30条件下，是否考虑管理成本会对最优结果有较大影响，相较于人工分区方案，不考虑管理成本情况可以降低10.89%的成本，考虑管理成本情况下则可以降低5.60%的成本。最后，利用虚幻引擎将土方分区施工方案及调配施工方案进行仿真建模，在满足可视与直观条件下，利用交互实现了更多信息表达。

结果表明，本文提出的土方调配优化方法和土方自动分区方法以土方成本最小为目标，实现了对土方分区施工方案和土方调配施工方案的对比寻优，有助于对土方工程成本进行控制，为建筑行业施工方案智能生成打下了基础。

关键词： 土方调配；模拟退火；土方工程分区；遗传算法；虚幻引擎

Abstract

With the rapid development of the construction industry, the competition in the construction industry has become increasingly fierce, which puts forward higher requirements for the development level of the construction industry. Earthwork is one of the main sub-projects of construction engineering. The study of earthwork zoning construction plan and earthwork allocation construction plan will help to improve the development level of the construction industry. However, the earthwork zoning construction plan is determined by the project manager's experience, lacking specific evaluation and optimization indicators; the earthwork allocation construction plan is all determined by a linear programming model, the consideration of factors is not comprehensive enough. In view of the above problems, this paper takes the optimization of earthwork allocation and automatic partitioning as the research object, and the following work is carried out:

Firstly, based on the existing research, by analyzing the earthwork allocation process, this paper puts forward the earthwork allocation cost model and analyzes the model. The earthwork cost of two examples is calculated by using the model. The errors between the calculation results and the results of commercial software are 0.89% and 0.22% respectively, which verifies the availability of the model. Secondly, according to the characteristics of earthwork allocation cost model, the earthwork allocation optimization is divided into linear programming model and nonlinear programming model. The former is solved by simplex method, and the error between the solution result and commercial software is 0.79%; The latter is solved by the improved simulated annealing algorithm, and the performance of the improved simulated annealing algorithm is verified by comparing with other heuristic algorithms. Thirdly, the experience of manual division of earthwork construction zoning is analyzed, and the earthwork automatic zoning problem model is established, which is solved by improved genetic algorithm. When the number of partitions is limited to the actual number of partitions, the superiority of the improved genetic algorithm is demonstrated by comparing with other heuristic algorithms. Compared with artificial partitioning results, the cost can be saved by 4.56%. Under the condition of limiting the number of partitions to 5 ~ 30, whether to consider the management cost will have a great impact on the optimal result. Compared with the manual partition scheme, the cost can be reduced by 10.89% without considering the management cost, and 5.60% when considering the management cost. Finally, the Unreal Engine is used to simulate and model the earthwork zoning construction plan and the earthwork allocation construction plan. Under the condition of visual and intuitive, the interaction is used to express more information.

The results show that the earthwork allocation optimization method and the earthwork automatic zoning method proposed in this paper aim to minimize the earthwork cost, realize the comparison and optimization of the earthwork zoning construction plan and the earthwork allocation construction plan, help to control the cost of earthwork, lay the foundation for the intelligent generation of construction plans in the construction industry.

Key words: Earthwork Allocation, Simulated Annealing, Earthwork Zoning, Genetic Algorithm, Unreal Engine