基于IPSO-BP的受损钢结构改造施工安全预警评估研究

为研究受损钢结构改造施工安全预警状况，建立受损情况下钢结构改造施工安全预警指标体系，并针对BP神经网络算法易陷入局部最优的问题，提出了用改进粒子群算法（IPSO）对BP神经网络权重及阈值进行调整的IPSO-BP安全预警评估模型。通过分析某单层重钢厂房受损现状，针对其结构损伤情况和已构建的安全控制指标体系进行数值模拟分析。研究结果表明:与传统的BP模型相比，IPSO-BP模型具有更好的预测能力，构建的安全预警指标体系及预警模型可以很好地对受损钢结构改造施工过程安全状况进行综合评估，对受损钢结构改造施工安全控制具有一定的参考价值。

Study on safety early warning assessment of damaged steel structure reconstruction based on IPSO-BP

In order to study the safety early warning status of damaged steel structure reconstruction,a safety early warning index system of damaged steel structure reconstruction was established.Aiming at the problem that the BP neural network algorithm is easy to fall into the local optimum,an IPSO-BP safety early warning assessment model was put forward,which adopted the improved particle swarm optimization algorithm (IPSO) to adjust the weights and thresholds of BP neural network.By analyzing the damage situation of a single layer heavy steel plant,the numerical simulation analysis was carried out based on its structural damage conditions and the built safety control index system.The results showed that the IPSO-BP model had better predictive ability compared with the traditional BP model,and the safety early warning index system and early warning model could assess the safety status of the damaged steel structure reconstruction process comprehensively.It has a certain reference value for the safety control of damaged steel structure reconstruction.