An optical-emission-based model predictive control method for discharge-state stabilization in biomass straw plasma electrolytic liquefaction

Abstract

Plasma electrolytic liquefaction of biomass straw is highly sensitive to discharge-state fluctuation, which directly affects reaction continuity and process stability. However, due to the strong nonlinearity and time-varying characteristics of the plasma–electrolyte system, conventional fixed-parameter operation is often insufficient to maintain a stable discharge regime throughout the liquefaction process. To address this issue, this paper proposes an optical-emission-based model predictive control method for discharge-state stabilization in biomass straw plasma electrolytic liquefaction. In the proposed method, optical emission signals from the reaction zone are collected online, and representative spectral features are extracted to construct a discharge-state stability index. Then, a control-oriented dynamic prediction model is established to describe the evolution of the discharge state under varying power input. Based on the predicted deviation between the future discharge state and the target operating condition, a model predictive controller is designed to compute the optimal power adjustment in a rolling optimization manner, thereby suppressing discharge oscillation and improving process continuity. Experimental results under different initial operating conditions show that the proposed method can effectively reduce discharge-state fluctuation, shorten disturbance recovery time, and maintain a more stable liquefaction process than conventional fixed-parameter and rule-based strategies. These results demonstrate that optical emission spectroscopy can be effectively integrated with model predictive control for real-time discharge-state stabilization in biomass straw plasma electrolytic liquefaction.