Research on the stability assessment method of bubbling fluidized bed sludge incineration system based on multi-source features

Abstract

Stable operation of incineration systems is crucial for ensuring safe production, controlling pollutant emissions, and improving energy efficiency. However, existing research primarily focuses on flame stability identification while neglecting other aspects of incineration systems, which makes it difficult to provide specific guidance for operational regulation. System stability encompasses not only combustion flame stability but also process parameter consistency across fuel feeding, air supply, temperature control, and emission management. This study investigates the stability of a bubbling fluidized bed sludge incinerator by integrating multi-source data from Distributed Control System (DCS), Continuous Emission Monitoring Systems (CEMS), and flame image features. First, temporal lag relationships between features are determined through mutual information analysis, enabling data reorganization. Second, the composite instability indices for each feature are defined, including three typical types: abrupt, trend, and boundary-exceeding instability indices. Then, the weights of each feature's composite instability index are determined through Principal Component Analysis (PCA), after which a weighted sum of these indices is calculated to obtain the weighted instability index at the current moment. Finally, the weighted instability index is scaled to the range of 0–1 based on Sigmoid function, resulting in the stability index of the incineration system for overall operational performance assessment. Through the comparative analysis of typical operating conditions, the effectiveness of this method is verified in dynamic response, anomaly identification and trend monitoring, which provides support for the stable operation and intelligent regulation of complex incineration processes.