运筹与管理2024，Vol.33Issue(3)：22-27,6.DOI:10.12005/orms.2024.0073
基于实测数据的公路桥梁动态养修策略自适应优化模型
Adaptive Optimization Model of Highway Bridge Dynamic Maintenance Strategy Based on Measured Data
摘要
Abstract
With the rapid development of highway transportation industry of China,the number of highway bridges is also increasing,and bridge maintenance work is becoming increasingly important.Existing bridges have differ-ent degrees of aging,damage,and other issues that require timely repair and maintenance to ensure their safety and reliability.During the formulation of highway bridge maintenance strategies,it is necessary for management departments to effectively utilize limited resources during the operation period to ensure the service safety of existing bridges.Therefore,formulating scientific maintenance strategies is of great significance for the reasona-ble arrangement of bridge maintenance work and the reasonable allocation of maintenance resources. To construct an optimization model for the maintenance strategy of highway bridges,the performance detec-tion data of 208 highway bridges in coastal area and 176 highway bridges in inland area with historical mainte-nance work are analyzed from 1996 to 2020.Based on the distribution type test and maximum likelihood estima-tion of these long-term performance detection data,a probabilistic model of the training gain coefficient consider-ing the length of service and the frequency of training is established.Based on the improved inverse Gaussian process,an evaluation method for the remaining service life of the bridge and the reliability of the next detection time is proposed,and the maintenance decision vector and decision rule are established according to the evalua-tion results.The dynamic maintenance strategy model of bridge is established by embedding Markov chain in the improved inverse Gaussian deterioration process and combining Bayesian update method,and the prediction per-formance of the model is evaluated by using a large number of measured data.The results show that the average relative error of the model prediction is 11.1％,which can meet the needs of the project to a certain extent.The improved gray wolf algorithm is used to adaptively optimize the decision vector in the dynamic maintenance strategy model.The improved grey wolf algorithm is used to adaptively optimize the decision vector in the dynamic maintenance strategy model to find the decision vector with the lowest remaining life cumulative cost.The existence of the optimal solution of the decision vector and the effectiveness of the adaptive optimization model are verified by an actual bridge optimization example. Through the evaluation of model performance with a large number of measured data,the results show that when no data classification is performed,the prediction accuracy of the dynamic maintenance strategy model is relatively low.After classification by service region,the average relative error decreases by 50.9％,and the prediction performance is significantly improved.After further subdivision of the data by bridge size,the average relative error increases by 17.6％,especially for extra-large bridges,which increases by 33.2％.After evalua-tion,when the service region is divided,the total average relative error of prediction using the dynamic mainte-nance strategy model for 8 datasets is 11.1％,which is the lowest prediction error after exhaustive search.It can meet engineering needs to a certain extent.Finally,the improved gray wolf algorithm is used to adaptively optimize the decision vector in the dynamic maintenance strategy model,and find the decision vector that minimi-zes the cumulative cost of remaining life.An actual coastal small bridge is selected as an optimization example to confirm the existence of the optimal solution of decision vector.关键词
公路桥梁/逆高斯过程/养修增益系数/养修策略优化/贝叶斯更新/灰狼算法Key words
highway bridge/inverse Gaussian process/maintenance gain coefficient/maintenance strategy optimization/Bayesian updating/grey wolf algorithm分类
交通运输引用本文复制引用
王崇交,姚昌荣,赵思光,赵实达,强斌,李亚东..基于实测数据的公路桥梁动态养修策略自适应优化模型[J].运筹与管理,2024,33(3):22-27,6.基金项目
国家重点研发计划(2016YFC0802202) （2016YFC0802202）
