Study On Highway Pavement Maintenance Management Based On Combining Forecasting And Fuzzy Theory

With the development of China's highway construction and network,highway maintenance management appears to be more and more prominent and important.At present, there still are not uniform technical standards and norms for the highway maintenance.In addition,the models of the highway maintenance management are very different all over the country.According to present situation of China's highway maintenance management,the paper studis pavement maintenance management decision of provincial highway network. Based on the pavement performance assessment and its forecasting studies,the thesis constructs the main frame of highway pavement maintenance management optimization decision,establishes the pavement maintenance fund allocation optimization model,discussed the model solution algorithm.This researchs have important theoretical significance in the context of sustainable development of highway itself and region socioeconomic.And they are also positive and practical for decision-making of the highway management department or its specific maintenance management.The key questions in the thesis includes:highway traffic forecasting,highway pavement performance assessment and its forecasting,optimization decision-making of pavement maintenance management.Highway traffic forecasting:On the basis of the characteristic and applicability of current traffic forecasting models,gray forecasting model and regression method have been selected and discussed systematically for highway traffic forcasting.Then,several combination forecasting models of the gray forecasting and regression are proposed to improve prediction accuracy based on combination forecast thought.And their algorithms are developmed in Matlab 8.The experiments show that the combination method can increase the forecasting precision and serviceability.Pavement performance assessment and its forecasting:The thesis discusses the measure and the assessment method about pavement performance sub-item indexes:Pavement Condition Index(PCI),Riding Quality Index(RQI),Pavement Structural Strength Index (PSSI) and Skidding Ressistance Index(SRI).Their fuzzy assessment functions are given, respectively.Then,a synthesis fuzzy assessment model is established for highway pavement performance assessment.The factors of highway pavement performance forecasting are analysed,and several present highway pavement performance forecasting models(such as: Neural Network,Markov model,et al.) are discussed about their character and applicabilities considered the present situation of our country highway maintenance management.Based on the combination forecast thought,the thesis proposes a combination forecasting model of the neural network and the Markov model to forecast pavement performance.Highway pavement maintenmance management decision-making optimization:The thesis discusses highway pavement maintenmance project decision-making and mainenmance fund allocation.Firstly,combined to existing pavement maintenmance technology standard,the requests and strategy on highway pavement maintenmance management are expatiated on. Secondly,depended on the engineering economy principle and method,the pavement maintenance cost-effectiveness is discussed and the decision-making scheme is choosed based on its cost-effctiveness analysis.Then the main frame of highway network level maintenance management optimization decision-making is proposed and analysed,and it mainly includes determination of research scope and scale,collection of basic data and examination data of path,forecasting pavement performance,choice of maintenance measure,optimization of maintenance decision-making and making projects.Finally,considered the key of pavement maintenance decision-making,three fuzzy mathematical programming models are proposed based on general standards,service level and fund restraint,respectively.