Research On Decision Model Of Pavement Preventive Maintenance Based On Multi-Objective Particle Swarm Optimization Algorithm

The stock of highway facilities in China has increased year by year,and themaintenance and management needs of pavement facilities are particularly prominent.The scientific use of limited funds as well as the efficient and reasonable maintenancemanagement programs are the main problems facing the highway managementdepartments in China.The micro-surfacing,the representation of preventivemaintenance technology with quick construction and obvious effects on input,has beenwidely used at home and abroad.However,there are various problems in themaintenance decision-making model of core functions,and it is necessary to construct amore scientific and rational model.For the division of road segmentation,traditional road section method is employedin the existing studies with the small application scope,complicated operation,and lowprecision.For the preventive maintenance benefit of expressway pavement,consideration is solely taken on the economic cost and performance of the pavementduring the maintenance process,with little attention paid to the environmental pollution.Meanwhile,most of the existing researches taking much consideration to the economicbenefits,lose sight in the internal link of the three factors of social benefits,environmental benefits and environment benefit.For the decision-making model ofpreventive maintenance of expressway pavement,the traditional manualdecision-making method or simple quantitative model are mostly applied in the existingresearch,and little innovation is made in the construction it.Accordingly,it is urgent toconsider the advantages of artificial intelligence algorithm in decision-making problemsand make use of it in the maintenance decision-making process.Based on probabilistic neural network theory and multi-objective barebonesparticle swarm optimization algorithm,taking advantages of artificial intelligencealgorithm in solving the decision-making problem of preventive maintenance ofpavement,this paper constructs a new preventive maintenance management model,anda specific case analysis is given,analyzing the economic,social and environmentalbenefits in the process of preventive maintenance,The main contents and results of thisthesis are as follows:(1)Based on the theory of probabilistic neural network,the model of road segmentdivision has been established and the influence factors of road segment division hasbeen analyzed.The contribution degree of each influence factor has been given,which can automatically classify the road segments according to their damage degree.(2)The benefits of preventive maintenance of pavement were divided into three aspects:economic cost saving,social cost saving and environmental cost saving.The evaluation models of economic,social and environmental benefits were established as the objective function of multi-objective decision-making model.(3)Based on multi-objective barebones particle swarm optimization,a preventive maintenance decision-making model has been constructed,and an optimal maintenance scheme selection method based on fuzzy function has been given by using the objective function constructed.The model can automatically evaluate the quality of maintenance decision-making results,and provide the basic parameters with high stability.(4)A case study has been carried out by using the above-mentioned decision-making model.The results showed that the best maintenance plan could be’rendered by the preventive maintenance decision-making model constructed in this study.