Waste management with recourse: An inexact dynamic programming model containing fuzzy boundary intervals in objectives and constraints

• Published in: Journal of environmental management 2010-09, Vol.91 (9), p.1898-1913
• Main Authors: Tan, Q., Huang, G.H., Cai, Y.P.
• Format: Article
• Language: English
• Subjects: Algorithms; Animal, plant and microbial ecology; Applied ecology; Biological and medical sciences; Boundaries; Coefficients; Computational methods; Conservation, protection and management of environment and wildlife; Decision Support Techniques; Dynamic programming; Fundamental and applied biological sciences. Psychology; Fuzzy; Fuzzy Logic; Fuzzy set theory; General aspects; Interval; Intervals; Linear programming; Mathematical models; Modelling; Municipal solid waste; Optimization; Programming, Linear; Random; Risk; Solid waste management; Uncertainty; Waste disposal; Waste management; Waste Management - economics
• ISSN: 0301-4797; 1095-8630
• DOI: 10.1016/j.jenvman.2010.04.005

The existing inexact optimization methods based on interval-parameter linear programming can hardly address problems where coefficients in objective functions are subject to dual uncertainties. In this study, a superiority–inferiority-based inexact fuzzy two-stage mixed-integer linear programming (SI-IFTMILP) model was developed for supporting municipal solid waste management under uncertainty. The developed SI-IFTMILP approach is capable of tackling dual uncertainties presented as fuzzy boundary intervals (FuBIs) in not only constraints, but also objective functions. Uncertainties expressed as a combination of intervals and random variables could also be explicitly reflected. An algorithm with high computational efficiency was provided to solve SI-IFTMILP. SI-IFTMILP was then applied to a long-term waste management case to demonstrate its applicability. Useful interval solutions were obtained. SI-IFTMILP could help generate dynamic facility-expansion and waste-allocation plans, as well as provide corrective actions when anticipated waste management plans are violated. It could also greatly reduce system-violation risk and enhance system robustness through examining two sets of penalties resulting from variations in fuzziness and randomness. Moreover, four possible alternative models were formulated to solve the same problem; solutions from them were then compared with those from SI-IFTMILP. The results indicate that SI-IFTMILP could provide more reliable solutions than the alternatives.