Research On Improvement Of Expressway Pavement Maintenance Decision

As the construction of expressway in China is close to 1.18 hundred thousand kilometers, the planned mileage of National Highway Network Planning(Year 2013-2030), China’s development of expressway has transferred from construction period to maintenance and management period. From the year 1988 when construction of expressway reached zero breakthrough, expressway got rapid development pulled by economic growth, which caused relatively large base of maintenance demand, besides, influencing factors as rapidly increasing traffic volume, seriously overloading and transfinite transportation, severe weather changes et al speed up the expressway pavement damage, so, large base of maintenance demand and rapid growth of maintenance demand make expressway pavement maintenance heavy and urgent, and appear the treand of wide maintenance area, large maintenance workload, short maintenance cycle and large capital need. But, because of state system, shortage of maintenance experience, large scale of maintenance demand, some problems that resource of maintenance funds is short, perfect maintenance decision system and scientific decision methodology have not been formed, data as environmental information where the expressway exits in, funds division target expressway and technology choice, technology implementation, pavement usage has not implemented long-term detectin and analysis, and has not evaluated the long-term benefits of maintenance, and pavement maintenance notion and research lag, have appeared during the expressway pavement maintenance practices. As a result, expressway movement maintenance issues have got widespread concern and attention, and the issues that how to establish efficient and scientific method of highway pavement maintenance and management, improve operation efficiency, meet high quality demand of pavement maintenance, meet a variety of transportation have become important problems encountered by expressway management, which were explored and studied.From the perspective of system optimization, on the background of expressway pavement maintenance and management, aimed at using maintenance funds effectively, arranging maintenance planning scientifically, improving traffic efficiency in maintenance period, firstly, load factor was characterized by ESALs, and non-load factors were expressed quantitatively, then expressway pavement maintenance cost model was established with ESALs, freezing index, subgrade drainage coefficient, pavement age and temperature condition, whose parameters were calibrated by least square method, and contribution rate of each factor was calculated. Secongly, function relations between vehicle travel time cost savings and pavement performance, vehicle fuel consumption cost savings and pavement performance, vehicle tire wear tiring cost savings and pavement performance, vehicle safety cost savings and pavement performance, exhaust emission pollution cost savings and pavement performance, traffic noise pollution cost savings and pavement performance, respectively, which were used to compute benefits of expressway pavement maintenance. Thirdly, method of determining the optimal time of pavement preventive maintenance was established by ratio of benefits and costs under the framework of life cycle theory, and bi-level programming model was established for optimal pavement maintenance project selection of expressway on the base that maximum of the sum of vehicle travel cost savings and toll income increase and minimum of vehicle travel generalized cost are made as the objective function of the upper level model and the lower level model respectively, then search algorithm combined with dynamic traffic assignment was designed to solve the model. Lastly, discreteness of diversion vehicles and randomness of non-diversion vehicles were characterized by discrete intervals and probability density function, respectively, chance constrained programming model with interval parameter was established by the objective of shortest diversion time, then, ant colony algorithm was designed to compute the optimal and worst diversion plan.Main research results and conclusions can be generalized as:(1) expressway pavement maintenance cost model was established with load and non-load factors, and contribution rate of each factor was calculated, which shows that: contribution rate of load factor reaches 47%, contribution rate of temperature condition is 23%, contribution rate of subgrade drainage condition is 11%, the damage of seasonal freezing weather is also large, whose contribution rate is 19%.(2)benefit model of expressway pavement maintenance was established, calculation shows that in the total maintenance benefits, vehicle travel time cost saving is 30.111%, vehicle traveling cost saving is 49.768%, the most of which is the benefits caused by fuel consumption saving benefits, vehicle safety cost saving is 4.87%, exhaust emission pollution cost saving is 15.082%, traffic noise pollution cost saving is the smallest, 0.169%.(3)optimal time model of pavement preventive maintenance was established under the framework of life cycle theory, which shows that if preventive maintenance time is too earlier, total life cycle benefits are on the rise, total life cycle costs are downward, and ratio of benefits and costs is also on the rise, if preventive maintenance time is too late, total life cycle benefits are downward, total life cycle costs are on the rise, and ratio of benefits and costs is downward.(4)project choice model of expressway pavement was establish considering of ratio of benefits and costs, then search algorithm combined with dynamic traffic assignment was designed to solve the model, computation shows that optimal plan can make total traffic volume increase by 1.56%, average travel time cost saves as 18.96%, average fuel cost saves as 3.73%, vehicle tire wear tiring cost saves as 4.11%, average vehicle travel cost saves as 5.31%, average toll income increases by 1.24%.(5)traffic management model during expressway pavement maintenance period was established, then ant colony algorithm was designed to compute the model, calculation shows that the higher confidence level ?, the shorter the diversion time T, and because of influence by probability distribution of nondiversion vehicle volume, when there is any rounte whose traffic volume is large, there is no obviously positive correlation between the value of ? and diversion time.