DEVELOPMENT OF A MAINTENANCE PLAN OPTIMISATION SYSTEM FOR BRIDGE NETWORK

摘要

The authors have been developing a Bridge Management System (J-BMS) for single bridge, which integrated with the Concrete Bridge Rating Expert System (BREX) that can be used to evaluate the serviceability of existing concrete bridges. The J-BMS will be able to predict the deterioration process of existing bridge members, construct a maintenance plan for repair and/or strengthening based on minimizing maintenance costs and maximizing quality, and estimate the maintenance costs in single bridge. In this system, the Genetic Algorithm (GA) technique was used to search for an approximation of the optimal maintenance plan. In the present study, a comprehensive decision support system is developed for maintenance strategies with/without annual budget limitations based on life cycle cost analysis of an entire bridge inventory, which form part of a highway network. Since formulation of an optimum bridge maintenance program for an entire stock of bridge structures is usually impeded by a lack of sufficient information on their existing conditions, the capability of the system was tried out, with a view to enhancing its capability, by applying it to a limited number of individual existing bridges, categorized as belonging to a typical bridge inventory. Genetic Algorithm is adapted, in order to obtain the semi-optimal solutions in this problem.