【摘要】 围绕建立基于FMEA、GCA、PCA与PSO-BP相结合方法的公共工程施工过程风险控制的审计评价模型,在分析相关文献的基础上,结合专家咨询和对工程建设的调查研究,明确施工人员风险、施工技术风险、施工组织风险和施工环境风险4个影响公共工程施工风险的决定因素,构建了公共工程施工风险控制的审计评价的指标体系;采用FMEA与专家经验相结合的方法,将各定性指标定量化,确定了各评价对象的专家评价结果,通过灰色关联分析,确定了各个风险指标与专家评价结果的关联程度;结合主成分分析、粒子群优化算法、BP神经网络对某高速公路工程的承建企业的施工风险进行了审计评价。实证结果表明FMEA、GCA、PCA与PSO-BP神经网络相结合的方法,提高了审计评价的科学性和客观性。更多还原

【Abstract】 The audit model is established to control the construction process risk of public works based on the conjoint methods of FMEA,GCA,PCA and PSO-BP.Firstly,Based on the analysis on concerned literatures of the audit and evaluation to control the construction process risk of public works,combining with expert advice and research on the construction,the 4determinants including construction workers risk,construction technology risk,construction organization risk and construction environment risk,which affect public works construction process risk are ascertained,and the audit and evaluation criteria system to control construction process risk of public works is established.Then,the method combining the FMEA and expert experience is used to quantify various qualitative indicators,the evaluation results of every evaluation object are ascertained by experts,and the correlation degree between each risk indicator and the expert evaluation results is determined through grey correlation analysis.Next,combing principal component analysis,particle swarm optimization and BP neural network,the construction process risk of contracting enterprises of a highway construction project is audited and evaluated.The empirical results show that the conjoint methods of FMEA,GCA,PCA and PSOBP neural network improve the scientific and objective of the audit and evaluation results.更多还原