基于GA的建设项目设计规模的仿真优化

Applying computerized simulation and genetic algorithms for design scale optimization of construction project

建设项目设计规模的确定是项目决策阶段的一项重要内容.设计规模确定的科学与否,直接影响到项目技术方案、设备方案、投资估算等的准确性,进而影响到项目决策和项目建成后的经济效益.针对建设项目设计规模决策的特点,提出了基于遗传算法(GA)的建设项目设计规模的仿真优化算法,使用计算机仿真技术处理随机的决策环境,利用GA进行最优方案的搜索.解决了传统方法没有考虑随机决策环境的弱点,同时又提高了算法的全局搜索能力,大大地提高了求解效率,增强了方法的实用价值.在充分考虑主要影响因素及其变化规律的基础上,在满足市场需求及项目经济目标的前提下,能够高效地找到最优设计规模,从而为决策者作出准确的决策提供科学全面的基础数据.

Selecting appropriate design scale is one of the most important tasks of construction project decision-making. Decision of the design scale of a construction project relates closely to the accuracy of the technical proposal, equipment programme and project investment estimates, and a scientific decision can improve the economic benefit of the project. According to the inherent stochastic characteristics of design scale decision of a construction project, a new mechanism that integrates simulation with genetic algorithms （GA） was presented, which could find the optimal solution to the design scale of a construction project. This new hybrid approach applies the computerized simulation technique to dealing with the stochastic decision circumstance, and adopts GA to search optimal solutions. The method has recently been modified to make it more efficient by considering the main influencing factors and their changing trends, It can meet the needs of the market and the economic target of a project. The process will be repeated until the optimal solution is found. The proposed approach provides the decision-makers an efficient tool of analyzing and optimizing the design scale of construction projects.