Keywords
Optimisation; stakeholder input; evolutionary algorithm; water
Start Date
5-7-2022 12:00 PM
End Date
8-7-2022 9:59 AM
Abstract
Sustainable integrated water management (IWM) requires collaborative planning and design approaches. The complex engineering systems involved can have a large number of options and multiple benefits and costs to multiple stakeholders. Understanding trade-offs between options with respect to these diverse benefits and costs can increase decision-maker buy-in to negotiated outcomes on IWM performance requirements and engineering solutions. Optimisation frameworks using evolutionary algorithms (such as genetic algorithms) can be tailored to the level of stakeholder interaction required to encourage buy-in into the modelling process and adopted solutions. This paper presents a series of three (3) IWM planning and design evolutionary optimisation approaches with increasing levels of engagement with stakeholders. The first approach is a traditional designer-lead optimisation process. Stakeholder inputs are formulated as constraints within the optimisation problem formulation. This suits designs with strict performance requirements (e.g. stormwater pollutant removal for housing developments). In the second approach, scores from a multi-criteria analysis (MCA) for catchment-management projects are linked to an evolutionary algorithm to identify optimal portfolios of solutions with respect to multiple criteria. The scoring contributed by stakeholders forms the basis for the evaluation of individual projects, encouraging buy-in. This enhanced MCA approach is suited to situations where there is a single funding body for the works. In the third approach, the MCA optimisation approach is expanded to incorporate the individual values and costs of multiple stakeholders. Visual analytics enables an exploration of joint-optimal solutions that are optimal with respect to all stakeholder’s individual objectives. The joint-optimal solutions are compared with the Best Alternative To Negotiated Agreement (BATNA) for each stakeholder and against a wide range of performance objectives.
Including varying degrees of stakeholder input in the optimisation of integrated water management planning and design
Sustainable integrated water management (IWM) requires collaborative planning and design approaches. The complex engineering systems involved can have a large number of options and multiple benefits and costs to multiple stakeholders. Understanding trade-offs between options with respect to these diverse benefits and costs can increase decision-maker buy-in to negotiated outcomes on IWM performance requirements and engineering solutions. Optimisation frameworks using evolutionary algorithms (such as genetic algorithms) can be tailored to the level of stakeholder interaction required to encourage buy-in into the modelling process and adopted solutions. This paper presents a series of three (3) IWM planning and design evolutionary optimisation approaches with increasing levels of engagement with stakeholders. The first approach is a traditional designer-lead optimisation process. Stakeholder inputs are formulated as constraints within the optimisation problem formulation. This suits designs with strict performance requirements (e.g. stormwater pollutant removal for housing developments). In the second approach, scores from a multi-criteria analysis (MCA) for catchment-management projects are linked to an evolutionary algorithm to identify optimal portfolios of solutions with respect to multiple criteria. The scoring contributed by stakeholders forms the basis for the evaluation of individual projects, encouraging buy-in. This enhanced MCA approach is suited to situations where there is a single funding body for the works. In the third approach, the MCA optimisation approach is expanded to incorporate the individual values and costs of multiple stakeholders. Visual analytics enables an exploration of joint-optimal solutions that are optimal with respect to all stakeholder’s individual objectives. The joint-optimal solutions are compared with the Best Alternative To Negotiated Agreement (BATNA) for each stakeholder and against a wide range of performance objectives.
Stream and Session
false