Development of a Stochastic Multicriteria Algorithm for Generating Waste Management Facility Expansion Alternatives

It is often beneficial to provide numerous quantifiably good solutions that provide distinct, contrasting perspectives when tackling waste management (WM) planning concerns. This is due to the fact that WM planning typically involves complicated challenges with a variety of performance objectives and design criteria that are difficult to define and describe when supporting decision models are needed. In the required decision space, the generated alternatives should meet all of the mentioned system criteria while being maximally dissimilar from one another. Modeling-to-generate-alternatives(MGA) is the process of generating as many diverse sets of solutions as possible. To handle computationally challenging, stochastic WM problems, simulation-optimization methodologies have been widely applied. This work describes a stochastic multicriteria MGA technique for WM planning that can produce sets of maximally diverse alternatives for any simulation-optimization method that uses a population-based solution algorithm. Because it provides the necessary number of maximally varied solution options in a single computational run of the technique, this algorithmic approach is computationally efficient. On a "real-world" waste management plant expansion case, the efficacy of this stochastic MGA approach is proved.