Maintenance strategy optimization of bridge decks using genetic algorithm

With the rapidly increasing requirements of bridge maintenance and the limited budget available for this maintenance, the cost optimization of long-term maintenance strategy considering the network-level bridge system under an allowable deterioration level is becoming an important problem. However, the total maintenance cost is a function of the number of bridges and their deterioration degrees, the planning period, and the maintenance methods. The selection of the optimization algorithm is important to get the optimal solution of this problem. In this research, using the bridge deck as an example, the search procedure and optimization technique of the long-term maintenance cost of a network-level bridge system are presented using a genetic algorithm. A population of maintenance strategies evolves from one generation to the next generation by applying the principles of natural selection and survival of the fittest. The feasibility of this method is demonstrated using a real-world example, and the results are compared with the conventional maintenance planning. In addition, several maintenance policies are suggested and compared.