Research On Network-level Pavement Maintenance Decision-making Technology Of Ningxia Ordinary National And Provincial Trunk Highway

The road network structure is gradually improved as the mileage of the Ningxia trunk route rises year after year,maintenance requirements are also rising.Nonetheless,the annual investment for maintaining trunk highways is modest.How to increase the standard of main highway maintenance while adhering to tight budgetary restrictions in order to save maintenance costs.A major problem in the Ningxia trunk highway maintenance project is how to achieve the scientific and rationalization of maintenance decisions.Due to the lack of a scientific decision-making model and the contradiction between capital efficiency in the process of highway maintenance and management.In this paper,the evaluation model and decision-making model in the process of highway maintenance and management are researched based on data from common national and provincial trunk highways in Ningxia.By examining the road network profile,pavement illnesses,level of pavement technical condition,and other factors,one may learn more about the technical state of the pavement on Ningxia trunk highways.Screening evaluation indexes that meet the actual situation of Ningxia,constructing evaluation models of common national and provincial trunk highways.In order to offer the decision basis for annual planning,mediumand long-term planning,etc.,the maintenance decision framework for the trunk highway is built,the algorithm is applied to maximize the decision outcomes,and ultimately the maintenance plan is clarified.The main research work and conclusions are as follows:Pavement damage rate(DR),pavement smoothness(IRI),and pavement rutting(RD)are the evaluation indexes that are appropriate for the Ningxia region.The classification prediction of the support vector machine is then utilized to classify and assess the pavement usage performance evaluation indexes.Converting the decision surface into a convex function using convex optimization and biasing the convex function yields the ideal decision surface.To assure the classification’s universality and rationality,mapping is employed to translate the 3D sample vector into an infinite dimensional vector.At the same time,the highway performance evaluation method of grid search,particle swarm algorithm search,and genetic algorithm search is more appropriate to Ningxia general national and provincial trunk routes.The results show that all three methods achieve a high correct rate,and the correct rate of the test set of the classification model obtained by using the genetic algorithm support vector machine model can reach 100%,indicating that the evaluation model established by the genetic algorithm optimization support vector machine has a high reliability.The disparities in traffic volume and surface layer thickness of national and provincial trunk highways were used to partition the regional highway pavement into nine subnetworks,and historical data on highway pavement performance was screened and analyzed.Using the gray prediction theory,the prediction model of pavement performance development law is created independently for each sub-network for the pavement technical condition index.The results demonstrate that for each sub-network,the disparities between the forecast results and the actual values determined by the gray theory are all placed between [-5,4] with just a tiny number of mistakes,essentially satisfying the prediction requirements.Based on the maintenance history data,the common maintenance types and typical maintenance schemes in Ningxia are summarized,the applicable conditions and costs corresponding to various maintenance measures are analyzed.The decision tree model is then used to select the highway maintenance countermeasures in accordance with the highway technical level division.To get the best pavement performance with the available maintenance resources,a multi-objective network-level maintenance decision model is solved using 0-1planning theory for single-year and multi-annual highway maintenance decisions.The nondominated ranking genetic algorithm(NSGA-Ⅱ)is used to solve the problem.The findings demonstrate that the multi-objective network-level choice model developed in this paper has solution outcomes that are more realistic and trustworthy,and can offer theoretical justification for the network-level decision of Ningxia highways.