论文部分内容阅读
为预防桥梁施工过程中安全事故的发生,完善施工风险预警机制,提升施工风险评估的可靠性,结合大跨径斜拉桥施工风险特点,提出了基于MCMC法与CCRAA法的大跨径斜拉桥施工风险预警模型。依据施工工序,对大跨径斜拉桥施工过程进行风险源识别,建立了施工风险层次评价体系。由于大跨径斜拉桥施工风险分析的信息较少,为最大限度地利用已有信息,采用MCMC方法中的Gibbs抽样,通过数值模拟直接从后验分布中生成参数向量的仿真样本,并针对这些仿真样本进行了统计推断,求解了复杂后验分布高维积分,进而得到了一级风险源的施工风险概率。运用CCRAA法首先确定风险基值,计算了相对海明距离,按比例对各一级风险源的风险值进行了补偿,得到风险区间后再进行风险聚合,求得二级风险源的施工风险概率,避免了聚合风险出现极值或者对风险概率产生依赖。基于以上两种方法,并结合大跨径斜拉桥施工实际,构建了施工风险概率区间,进而建立了大跨径斜拉桥施工风险预警模型,并实施了施工风险二次评估,以确保桥梁施工安全。通过工程实例分析表明:该预警模型对大跨径斜拉桥实际施工风险能做到有效评估,对于保证桥梁施工安全具有一定的实用价值。
In order to prevent the occurrence of safety accidents in bridge construction, improve early warning mechanism of construction risk and improve the reliability of construction risk assessment, the long-span cable-stayed bridge based on MCMC method and CCRAA method is proposed in combination with the construction risk characteristics of long-span cable- Bridge construction risk early warning model. According to the construction procedure, the risk source identification of the construction process of the long-span cable-stayed bridge was established, and the construction risk hierarchy evaluation system was established. In order to make full use of the existing information, the Gibbs sampling in the MCMC method is used to generate the simulation samples of the parameter vector directly from the posterior distribution through numerical simulation, because there is not much information on the construction risk analysis of the long-span cable-stayed bridge. The simulation samples are statistically inferred and the complex posterior distribution high-dimensional integrals are solved, and the probability of construction risk of the first-level risk sources is obtained. CCRAA method is used to determine the base value of risk firstly, the relative Hamming distance is calculated, and the risk value of each level of risk source is compensated according to the ratio. After the risk interval is obtained, the risk aggregation is carried out to obtain the construction risk probability of the second-level risk source , To avoid the extreme aggregation risk or risk dependence on risk. Based on the above two methods, combined with the actual construction of large-span cable-stayed bridges, the construction risk probability interval is constructed, and then the early-warning model for construction risk of long-span cable-stayed bridges is set up and the second assessment of construction risk is implemented to ensure that the bridge construction safety. The case study shows that the early warning model can effectively evaluate the actual construction risk of the long-span cable-stayed bridge and has certain practical value for ensuring the construction safety of the bridge.