Entropy game-based information processing and Q learning molecular fuzzy optimization of second-life battery investments for recycling solutions
Computers and Industrial Engineering, vol.216, 2026 (SCI-Expanded, Scopus)
- Publication Type: Article / Article
- Volume: 216
- Publication Date: 2026
- Doi Number: 10.1016/j.cie.2026.111987
- Journal Name: Computers and Industrial Engineering
- Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, ABI/INFORM, Compendex, INSPEC, DIALNET
- Keywords: Ant colony optimization, Genetic algorithm, Molecular fuzzy sets, Recycling solutions, Second-life battery investment
- Azerbaijan State University of Economics (UNEC) Affiliated: No
Abstract
Second-life battery investments sit at the intersection of circular-economy strategy, battery recycling, and uncertainty-laden industrial decision making. Existing studies provide important techno-economic, lifecycle, and policy insights, but they less often support strategic prioritization when quantitative data are limited and expert judgments are heterogeneous. This study develops a hybrid decision-support framework that integrates entropy- and coalition-based decision-maker weighting, Q-learning-based evaluation calibration, genetic-algorithm-based criterion weighting, and ant colony optimization within a molecular fuzzy environment. The framework is applied to prioritize strategic alternatives for second-life battery recycling and repurposing investments. The results indicate that operational cost and technological readiness are the most influential criteria, while digital monitoring platforms and hydrometallurgical recycling emerge as the highest-priority alternatives. Additional sensitivity checks across different learning rates and molecular geometry shapes show stable ordering of criteria and alternatives. The contribution of the study is a structured framework that links uncertainty representation, expert-calibration, and ranking in a single architecture for emerging battery circularity decisions.