Radial-Interval Linear Programming for Environmental Management under Varied Protection Levels

Abstract

In this study, a radial-interval linear programming (RILP) approach was developed for supporting waste management under uncertainty. RILP improved interval-parameter linear programming and its extensions in terms of input reasonableness and output robustness. From the perspective of modeling inputs, RILP could tackle highly uncertain information at the bounds of interval parameters through introducing the concept of fluctuation radius. Regarding modeling outputs, RILP allows controlling the degree of conservatism associated with interval solutions and is capable of quantifying corresponding system risks and benefits. This could facilitate the reflection of interactive relationship between the feasibility of system and the uncertainty of parameters. A computationally tractable algorithm was provided to solve RILP. Then, a long-term waste management case was studied to demonstrate the applicability of the developed methodology. A series of interval solutions obtained under varied protection levels were compared, helping gain insights into the interactions among protection level, violation risk, and system cost. Potential waste allocation alternatives could be generated from these interval solutions, which would be screened in real-world practices according to various projected system conditions as well as decision-makers’ willingness to pay and risk tolerance levels. Sensitivity analysis further revealed the significant impact of fluctuation radii of interval parameters on the system. The results indicated that RILP is applicable to a wide spectrum of environmental management problems that are subject to compound uncertainties.