Paper Details

Name: Debjani Chakraborti
Country: India

Year: 2025
Volume: Volume-12, Issue-2 (July-December)
Page Number: 75-87
DOI: https://doi.org/10.5281/zenodo.17097510

ABSTRACT
This study introduces an innovative method that employs evolutionary computation to
address interval-valued fractional bilevel programming (IVFBLP) problems. These involve
decision-making at two hierarchical levels, where some data is expressed as intervals rather
than fixed values, enhancing realism but increasing complexity. The proposed method
integrates mixed 0–1 programming, goal programming, and genetic algorithms to efficiently
address these challenges.
The genetic algorithm mimics natural evolutionary processes by generating multiple solution
candidates and iteratively selecting the best, allowing the model to explore a wide solution
space and converge towards optimal or near-optimal decisions. To ensure solutions closely
meet desired goals while minimizing adverse outcomes, termed regrets, the approach
employs two strategies—minsum and minmax—within a combined success-measuring
function.
The solution process follows two phases: first, it establishes optimal target intervals
representing achievable goal ranges; second, it identifies the best decisions for both leader
(upper) and follower (lower) roles. This structured approach delineates responsibilities
between decision-makers.
Utilizing evolutionary computing enables the model to handle uncertainties, nonlinearities,
and complex structures effectively. The paper demonstrates practical applicability through a
numerical example, showcasing its capability to solve real-world problems involving
interval-based objectives and constraints where decisions occur at multiple hierarchical
levels.
Keywords : Fractional bilevel programming, Goal programming, Evolutionary algorithm,
Interval programming, Interval-valued fractional bilevel programming, Multiobjective
decision-making.