On the Use of Evolutionary Optimization for the Dynamic Chance Constrained Open-Pit Mine Scheduling Problem

2026-04-16 · Source: cs.AI updates on arXiv.org · Field: Manufacturing & Industrial — Manufacturing Operations & Management, Artificial Intelligence & Machine Learning, Supply Chain & Logistics · Depth: Expert, extended

Summary

A study investigates the dynamic chance-constrained open-pit mine scheduling problem (DCC-OPMSP), which involves stochastic block economic values and dynamically changing mining and processing capacities. Researchers from Adelaide University propose a bi-objective evolutionary formulation that maximizes expected discounted profit while minimizing its standard deviation. To address dynamic changes, they introduce a diversity-based change response mechanism that repairs infeasible solutions and injects new feasible ones upon detecting changes. This mechanism was integrated into four multi-objective evolutionary algorithms (MOEA/D, NSGA-II, SMS-EMOA, and SPEA2) and evaluated against re-evaluation-based baselines. Experimental results on six mining instances, including Newman1, Zuck Small, and P4HD, demonstrated that the proposed diversity-based approach consistently outperformed baseline methods across various uncertainty levels and change frequencies, handling instances with up to 53,271 blocks.

Key takeaway

For AI Scientists and Research Scientists developing optimization solutions for complex real-world problems like mine scheduling, adopting a bi-objective evolutionary approach with a diversity-based change response mechanism can significantly improve robustness. Your models will generate high-quality, risk-aware schedules that adapt effectively to stochastic profits and dynamic resource constraints, eliminating the need for multiple runs at different confidence levels. Consider integrating hypermutation and random solution injection to maintain population diversity and feasibility in dynamic environments.

Key insights

A diversity-based evolutionary approach effectively optimizes dynamic, uncertain open-pit mine schedules by balancing profit and risk.

Principles

• Simultaneously optimize expected value and variance for risk-aware planning.
• Maintain population diversity to adapt to dynamic environmental changes.
• Bi-objective formulation eliminates the need for predefined confidence levels.

Method

The proposed method uses a bi-objective evolutionary formulation to maximize expected profit and minimize its standard deviation. A diversity-increasing change response mechanism repairs infeasible solutions via hypermutation and introduces new feasible solutions when dynamic changes occur.

In practice

• Apply MOEA/D-DIV or SMS-EMOA-DIV for robust mine scheduling.
• Use bi-objective optimization to explore risk-return trade-offs.
• Implement diversity-based repair for dynamic constraint changes.

Topics

• Open-Pit Mine Scheduling
• Dynamic Optimization
• Multi-Objective Evolutionary Algorithms
• Chance Constraints
• Stochastic Profits

Best for: AI Scientist, Research Scientist

Related on AIssential

• On the Use of Evolutionary Optimization for the Dynamic Chance Constrained Open-Pit Mine Scheduling Problem — Evolutionary optimization effectively manages dynamic, uncertain open-pit mine scheduling by balancing profit and risk.
• Adaptive mine planning under geological uncertainty: A POMDP framework for sequential decision-making — Sequential belief updating transforms geological uncertainty into an active component of value creation in mine planning.
• Diversity-Driven Offline Multi-Objective Optimization via Nested Pareto Set Learning — Offline multi-objective optimization can achieve diverse, high-quality Pareto sets by mitigating OOD issues and learning nested preferences.
• GOAL: Graph-based Objective-Aligned Diffusion Solvers for Dynamic Multi-Objective Optimization — GOAL is a graph-based diffusion solver for dynamic multi-objective optimization, achieving high feasibility and speed.
• Decoupling Numerical and Structural Parameters: An Empirical Study on Adaptive Genetic Algorithms via Deep Reinforcement Learning for the Large-Scale TSP — Dynamic structural adaptation is more critical than numerical fine-tuning for scalable evolutionary algorithms.
• Solving the Wrong Problem Works Better - Robert Lange — Shinka Evolve uses LLMs and evolutionary algorithms for sample-efficient program generation and scientific discovery.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.AI updates on arXiv.org.