Maintenance Planning of Deteriorating Bridges by Using Multiobjective Optimization

Cost-effective bridge maintenance planning requires balanced consideration of long-term bridge performance and life-cycle maintenance cost. Many of the existing methodologies determine an optimal maintenance planning solution based solely on life-cycle cost minimization while enforcing constraints on bridge performance. The resulting single planning solution, however, may not always satisfy bridge managers' specific requirements for bridge performance over an intended time horizon. In response, the life-cycle maintenance planning of deteriorating bridges is formulated as a multiobjective optimization problem and is solved by a genetic algorithm. The visual inspection-based condition state, structural assessment-based safety state, and cumulative life-cycle maintenance cost are all treated as competing criteria. A group of different maintenance strategies is considered. A multilinear computational model is adopted to predict time-varying deterioration processes under no-maintenance and maintenance interventions. Relevant parameters in this computational model are described as random variables to account for sources of uncertainty associated with the complex deterioration process. Monte Carlo simulation is performed to estimate sample mean values of performance indicators and maintenance cost on the basis of data collected in the United Kingdom. Application examples are presented for maintaining a group of reinforced concrete crossheads that have been undergoing deterioration in both condition and safety. It is demonstrated that an overall desirable maintenance planning solution can be obtained only by balancing all necessary merit measures (e.g., condition, safety, and cost) through an explicit trade-off analysis.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01002484
  • Record Type: Publication
  • ISBN: 0309093813
  • Files: TRIS, TRB
  • Created Date: Jul 28 2005 5:30PM