| The national road network has taken shape through the development in the past several years, while the road industry continues its rapid growth, along with a sharp increase in workload of road maintenance. Due to huge investment and complicated technologies of the road project, using scientific and efficient management approaches becomes the only way to achieve maximum cost savings, ensure quality and meet the needs of modern road network construction. However, despite the inevitable trend of modern engineering project management has been widely recognized, the degree and frequency of information management system utilization by road management department is still at a low level with a narrow range of related applications, not to mention the decision support tools. When it comes to the current situation, poor design ability for the road management system which leads to lack of system authority, out of touch with the real condition and neglect of the system maintenance are all major factors responsible for it. In order to develop a real practical road management system, an intelligent pavement information system was built with practicability, accuracy, authority, versatility and ease of operation.First of all, a pavement condition evaluation model was established by means of improved AHP method, in which the consistency and compatibility of AHP were optimized to figure out the weight vector of each evaluation index after the pairwise comparison matrixes were given by experts, together with variable fuzzy comprehensive evaluation method to estimate the comprehensive pavement condition level, followed by an example which demonstrated the operation process of the pavement condition evaluation system. Not only did the model keep the experts'original opinions to the maximum extent without personal subjective preferences, but the convergence speed of the iterative algorithm turned out to be acceptable as well. A dynamic determination method for index weights and membership degrees instead of traditional static method could reflect the real pavement condition more objectively and accurately.Second, the idea that analogical reasoning can be used for new projects in terms of the existing data and successful experience of similar cases was introduced and applied into a number of road prediction models, such as road condition prediction model and road risk analysis model. Given the assumption that certain solution can be reused in the same situations, similar cases were retrieved from the case database to provide accurate predictions by simulating human cognitive experience. Also, the system was able to conduct other data mining activities to explore the potential objective laws behind the data after obtaining information from similar cases. This model was more related to human's primitive way of thinking about cognition and problem solving, both qualitative data and quantitative data could be taken into account, in addition to its ability of active learning and incremental learning.Third, a 3D-WebGIS platform was built. Related software, scripts and other technologies, along with its application in intelligent pavement information system were introduced. The 3D virtual simulation map can vividly perform the geographical information and surroundings as a strong visualization tool, and map users can login on the platform, search or-share data via Internet. Moreover, the 3D-WebGIS platform can also show real-time working process and results of pavement condition evaluation system, road risk analysis system and so forth.Finally, an example was performed in the intelligent pavement information system to demonstrate each function of the system. The combination of pavement management system, intelligent transportation system and 3D-WebGIS platform proved to be practical and feasible.
|
PDF Full Text Request