Keywords: bridge management, optimization, computer application, infrastructure management, genetic algorithms, shuffled frog-leaping, evolutionary algorithms.

Bridges are vital links in road networks and it is important to keep them well maintained despite of the difficult operating conditions and the limited repair budgets. Various bridge management systems (BMS) have therefore been introduced to support decision makers in allocating the limited repair funds to top-priority bridges. The prioritization task, however, is extremely challenging and involves decisions related to project-level selection of least costly repair strategies as well as network-level selection of top-priority bridges that bring the highest return on the repair budget. These decisions represent a complex and large-scale optimization problem that traditional optimization techniques are often unable to solve.

In this paper, a simplified framework for bridge deck management system (BDMS) that considers both network-level and project-level decisions is presented. To deal with the primary challenge of large problem size, two evolutionary-based algorithms; genetic algorithms (GAs) and shuffled frog leaping (SFL) are introduced and various experiments are conducted to maximize their effectiveness. A comparison between the GAs and the SFL is then presented for case studies with different numbers of bridges to investigate their suitability for infrastructure management applications.

For practicality, the proposed BDMS includes various aspects: (1) detailed models for deterioration prediction, repair cost analysis, and after-repair condition analysis; (2) ability to consider user defined, project-level, and network-level constraints such as a yearly budget limit and a desirable overall condition; and (3) a decision support module for condition assessment, and life cycle cost optimization. The proposed BDMS was implemented on a commercial spreadsheet program (Microsoft Excel). Using the VBA Macro programming language of Excel, the GA and the SFL algorithms were coded and integrated into the developed BDMS. The proposed BDMS takes as inputs data related to each bridge, including: construction year; initial cost; deck type (steel or concrete); highway type (interstate or other); average daily traffic (ADT); width; length; and inspected condition (current condition). The output of the BDMS optimization is the set of bridges decided to be repaired at each year in the planning horizon, along with the associated types of repairs.

Initial experiments using the developed GA and SFL algorithms were then conducted on a network of fifty bridges but the results were unsatisfactory. The large number of variables involved even in a fifty-bridge network took a lot of time to optimize. Also, while the sum of yearly expenditures is getting closer to the desired budget limit, the yearly distribution of the expenditures often violates the yearly budget limit. In addition, the best solutions obtained still violated the minimum desirable condition for some individual bridges.

Due to this poor performance, various additional approaches were used to improve the performance of both algorithms on this typical infrastructure problem:

• Examining other objective functions and deciding the most proper one to use;
• Developing a pre-processing function to avoid violating the minimum condition constraint;
• Examining the effect of changing the initial solutions used in the optimization;
• Determining the best values for the parameters of each algorithm; and
• Investigating the use of a year-by-year optimization st

The effects of these measures on the performance of the GAs and the SFL algorithms are analysed on networks of 50, 100, 200, and 400 bridges, and the results are compared. The results show that each algorithm can perform better, given proper setup for its parameters. Also, the best results show that the two algorithms could allocate the repair funds efficiently and produce results that correspond to the objective function used. Based on the experiments made and the various approaches used to improve the optimization performance, the best optimization strategy to use for this typical infrastructure problem is determined.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £105 +P&P)