Robust Multi-Objective Production Optimization Under Demand Uncertainty, Machine Availability and Dynamic Energy Prices Using Real Industrial Data

Abstract

Production systems increasingly operate under simultaneous uncertainty in customer demand, machine availability, and energy-market conditions, creating a need for optimization frameworks that can coordinate operational efficiency with resilience and energy-related economic performance. This study develops a robust multi-objective production optimization framework for manufacturing environments in which production decisions must remain effective despite variations in demand, machine availability, and dynamic electricity prices. The proposed framework integrates production planning, machine-capacity allocation, operational flexibility, and time-dependent energy costs within a unified optimization structure. Demand uncertainty is represented through scenario-based uncertainty sets, while machine availability is incorporated as a dynamic operational constraint and electricity prices are treated as time-dependent parameters. The optimization model simultaneously considers production cost, energy expenditure, production completion time, and service-related performance, allowing the trade-offs among competing operational objectives to be explicitly evaluated. A robust solution mechanism is developed to identify production plans that maintain acceptable performance across alternative operating conditions rather than optimizing exclusively for a single deterministic scenario. The empirical analysis uses production, capacity, machine-availability, and electricity-price observations to construct a multi-period manufacturing decision environment. Several operational scenarios are evaluated to examine the sensitivity of the optimal production plan to demand fluctuations, machine unavailability, and energy-price volatility. The results indicate that incorporating multiple uncertainty sources into the production optimization process can substantially alter production allocation and scheduling decisions compared with deterministic planning. The proposed approach also demonstrates that greater operational robustness can be achieved without necessarily imposing proportional increases in production cost, particularly when production flexibility is strategically allocated to periods with more favorable energy prices. The framework provides a quantitative basis for balancing economic efficiency, operational resilience, and energy-aware production decisions and offers a scalable decision-support structure for complex manufacturing systems characterized by interacting operational uncertainties.