Abstract: To avoid risks such as poor quality, construction period delay and higher management cost in construction projects, a construction subcontractor selection model was built oriented to system quality reliability maximization. To constrain the mutual influence of period, cost and quality between subcontractors, the penalty function was used to punish the excess value. Aiming at this model, MATLAB solving program was designed by using particle swarm optimization, and the optimal selection of construction subcontractor was realized. The proposed model was applied in subcontractor selected activity, and the feasibility and correctness were verified.

Key words: construction subcontractor selection, system general quality reliability, particle swarm optimization algorithm, series-parallel sub-projects, construction industry

摘要： 为规避工程项目分包时项目质量不合格、工期延迟和管理成本增加等风险，建立了以系统总质量可靠度最大为目标的分包商选择模型。为进一步约束分包项目之间工期、成本和质量等因素的相互影响，模型引入惩罚函数，以此对超出约束的值加以惩罚。针对该模型，利用粒子群算法的原理，设计了用于模型求解的MATLAB程序，实现了建筑分包商的优化选择。最后，将提出的模型和算法应用在建筑项目的分包商选择活动中，验证了所提方案的正确性和可行性。

关键词: 建筑分包商选择, 系统总质量可靠度, 粒子群优化算法, 串并联分包项目, 建筑行业