Robust decentralized sliding mode control for the Tennessee-Eastman Process: a plant-wide approach

Abstract This study introduces a robust decentralized multiloop single-input-single-output (SISO) control architecture for the Tennessee-Eastman Process (TEP) utilizing Sliding Mode Control (SMC) and Higher-order Sliding Mode Control (HOSMC). Sliding Mode Controllers are employed due to their robustness against uncertainties and disturbances, effectively managing TEP’s complex nonlinearities and dynamic challenges. However, the chattering problem remains a significant drawback of these controllers. Various techniques exist to mitigate this issue, with Higher-order SMC being one of the most effective methods. In this paper, both SMC and HOSMC are tested and compared with traditional PID controllers. To achieve optimal performance, the controllers’ coefficients are optimized using the Particle Swarm Optimization (PSO) technique. The methodology systematically designs production rate and inventory controls using HOSMC. The results demonstrated that, despite the increased complexity and the difficulty of implementation and parameter tuning in HOSMC, the output fluctuations did not significantly improve. This suggests that the proposed SMC method is simple yet effective in achieving optimal production rates and maintaining stability under significant disturbances, such as feed composition variability and reaction kinetics shifts. A notable feature of the system is its ability to autonomously override standard operations during severe disruptions, such as the complete loss of a reactant feed. Compared to other control methods, the SMC-based architecture exhibits superior accuracy, simplicity, and reliability in managing critical process variables, even under stringent operational constraints. This work underscores the potential of SMC as a practical and efficient solution for plant-wide control in complex chemical processes.