Research And Application Of Pavement Prediction Models And Optimization Model

Pavement management system is an efficient tool that help highway administrative department to make full use of limited resource, in the hope of providing proper service level of road surface for road user. Pavement performance prediction models and optimization model is the important component of the pavement management system. The realization of performance prediction models and optimization model is the key of the development of the pavement management system.Pavement performance prediction model based on neural network is founded and realized in the pavement management system. The algorithm is described clearly in this paper. The model overcomes some traditional insufficiencies such as the model being apt to deviate and wrong because of the selection of parameter Three optimization models are used to make pavement maintenance need plan is founded. Three models have different restriction which include maintenance standard being specified, maintenance level of road network being specified and the maintenance cost being specified. Algorithms of three models are also described clearly in this paper. These algorithms resolve the optimum investment problem of pavement maintenance within somespecified area and specified years. Test results using practice data verify the feasibility of these algorithms.The author combine the idea that some sub territory can be ignored in the 0-1 integer programming algorithm and the algorithms own character to optimize the algorithm of specified investment. As a result, the run speed is improved greatly. The concept of group element is first used in variable keys and the referenced algorithm is presented as for integer programming algorithm with the specificrestricted function.