ARTIFICIAL NEURAL NETWORK BASED DECISION SUPPORT SYSTEM FOR ROAD MAINTENANCE

摘要

Development of a Decision Support System (DSS) for road maintenance involves choosing optimal times, places, and maintenance and repair (M & R) actions to be carried out on a road network. Due to very large number of available alternatives and complexity involved in the decision making process, it is difficult to rely solely upon the experience and judgement of experts and field engineers. Further, road maintenance programming has an extremely large solution space and involves solving a highly constrained, multi-objective optimization problem which is difficult to solve by traditional methods. Artificial Intelligence (AI) techniques like Neural Networks and Genetic Algorithm (GA) are evolving as promising new technologies for assisting planners in decision making process. Artificial Neural Network (ANN) provide efficient and optimal solutions for complex problems involving realistic data with advantage of faster implementation and easier updating than with other traditional techniques. The present work aims at development of ANN based modules for identifying distressed road segments which need maintenance and the most appropriate M & R actions to be carried out on these segments.