Assisting decision-makers select multi-dimensionally efficient infrastructure designs - Application to urban drainage...

Area of Science:

• Engineering
• Environmental Science
• Systems Design

Background:

• Multi-objective design approaches balance engineering, environmental, and societal goals for infrastructure systems.
• Assessing trade-offs among optimal solutions is crucial for infrastructure planning.
• Interpreting large sets of efficient solutions from multi-objective optimization can be challenging.

Purpose of the Study:

• To reduce the complexity of Pareto-fronts in multi-criteria infrastructure system design.
• To aid decision-making by simplifying the set of efficient infrastructure designs.
• To preserve the multi-dimensional structure of solutions on the efficiency frontier.

Main Methods:

• Application of a soft clustering algorithm to partition the Pareto-front.
• Identification of similarities between system design solutions.
• Selection of representative solutions and introduction of post-optimization metrics.

Main Results:

• The method effectively simplifies Pareto-fronts containing thousands of solutions.
• Identified representative designs highlight crucial trade-offs.
• The approach aids in interpreting the performance of simulated infrastructure systems.

Conclusions:

• The developed method simplifies complex multi-objective optimization outputs for infrastructure design.
• This simplification facilitates better understanding of trade-offs for decision-makers.
• The approach is applicable to real-world case studies, such as urban drainage networks.