Using Multiobjective Evolutionary Algorithms to Support Water Resources Planning
Water utilities undertaking long term planning must consider conflicting performance goals, e.g. trying to reliably meet growing demands, maintaining aging infrastructure, while avoiding costly new infrastructure and minimizing harm to the environment. Using traditional planning approaches, the utilities may only be able to develop and test a few alternative portfolios, and these few may not achieve the best performance or balance possible in the competing objectives. Multiobjective Evolutionary Algorithms (MOEAs) are an emerging decision support tool that can overcome the limitations in portfolio development and provide quantitative information about the tradeoffs between competing objectives, leading to insights about a water supply system and high-performing, innovative sets of planning decisions. Attendees of this workshop will:
- Learn what an MOEA is.
- Identify how it can be used in different ways- brief case study reviews.
- Participate in in-depth, hands-on, interactive session on fictitious but realistic western water utility to see how it can be used for long term water supply planning – (majority of workshop).
- Describe how MOEAs can support complex decision tradeoff analysis.
- See how visual analytics can help analyze difficult problems.
All participants must have Tableau Reader installed on their computers (downloadable for free
). The OS requirements are: Microsoft Windows 7 or later or Mac OSX 10.11 or later.
Download Tableau Reader
See this workshop in action - here's what attendees are saying:
“There’s a big trend to make decisions by weighting, and let the math make the decision for you, but we lose the ability for the human brain to weigh the tradeoffs. [The MOEA] is a really nice alternative … Combining human ability with model ability.”
“Is the impact that subtle changes in decisions can have on performance [water system] something that utilities miss in the way we currently do things?”
“It was really helpful to go through different layers of complexity. Once we got a basic understanding we moved into applied, interactive examples and had time for discussion.”