An Innovative Algorithm-Driven Optimization Framework for Landfill Mining: Aiming Sustainable Profitability Expeditiously

Abstract

Landfill mining (LFM) offers a potential solution to the environmental issues associated with landfilling. The current work aims to develop an efficient optimization framework for LFM that is sustainable, profit-yielding, and time minimizing at the same instance. The proposed framework involves a multi-objective multi-level solid transportation model (MOMLSTM). This model can be adapted by the organization across various geographies as it incorporates uncertainty of all the parameters of time, cost, and emission through pentagonal fuzzy numbers (PFN). The other crucial contribution of this work is the development of a genetic algorithm for offspring refinement (GAOR) that contributes in optimizing multiobjective optimization (MOO) problems. The GAOR's performance has been verified using the Congress on evolutionary computation (CEC) 2020 multi-objective benchmark test functions. GAOR is assessed against six robust MOO algorithms, including the multi-objective equilibrium optimizer slime mould algorithm (MOEOSMA), enhanced multi-objective particle swarm optimization (EMOPSO), multi-objective gorilla troops optimizer (MOGTO), adaptive crossover strategy enhanced NSGA-II (ASDMSGA-II), multiobjective slime mould algorithm (MOSMA), and multi-objective equilibrium optimizer algorithm (MEOA). GAOR delivered outstanding results across three crucial performance indicators. To rank these algorithms, a Friedman test was conducted, and GAOR achieved the highest ranking among the tested MOO algorithms. A case study is considered for real-life application of the model and solution technique GAOR. The outcomes of MOMLSTM from GAOR are compared to the epsilon-constraint method. The comparison revealed noteworthy improvements: a 0.14% increase in profits, a 1.29% reduction in carbon emissions, and a 3.81% decrease in the time required.