Real-time Accelerated Particle Swarm Optimization for Minimizing Energy Costs with Changing Demands and Fuel Price in Industrial Cogeneration System

Energy cost reduction in industrial cogeneration systems is critical for enhancing operational efficiency and minimizing environmental impact. However, achieving optimal load allocation across multiple boilers and turbines in real-time poses significant challenges, particularly under fluctuating demand and variable fuel costs. Traditional optimization methods often fail to adapt efficiently to these dynamic conditions, limiting overall system performance. This study develops a Real-Time Optimization (RTO) framework to minimize energy costs in cogeneration systems using Particle Swarm Optimization (PSO) and Accelerated PSO (APSO). The framework integrates Model Predictive Control (MPC) and PI controllers to optimize load distribution between boilers and turbines. The framework aims to minimize energy costs by dynamically adjusting operations in response to demand and fuel price fluctuations. Our findings show that APSO achieves a total energy cost reduction of 26% compared to PSO, lowering costs from RM 964,000 to RM 713,000. APSO delivers particularly significant savings for individual boilers, with reductions of 30% for Boiler 1 and 49% for Boiler 4, demonstrating its effectiveness in optimizing cost efficiency under variable operational conditions. Conversely, a 19% increase in energy costs for Boiler 2 highlights trade-offs focused on achieving system-wide efficiency. In summary, this work provides an adaptive and effective RTO framework that significantly enhances energy cost savings in large-scale cogeneration systems, proving APSO’s superior capability in managing dynamic, interdependent operations.