Dioxin Emission Detection Based on Dynamic Pruning Online Ensemble Binary Tree Algorithm

As one of the by-products of the municipal solid waste incineration (MSWI) process, dioxin (DXN) is not only difficult to detect but also potential harm to humans and the environment. The article proposes a method for detecting DXN emissions. It addresses the challenge of poor generalization performance in detection models due to the dynamic nature of the MSWI process. Firstly, the method constructs a historical soft sensor model and a drift detection model based on historical samples. Secondly, it assesses online samples for drift detection. When drift is detected, it calculates a distance threshold to prune the ensemble model. Subsequently, it reconstructs new ensemble sub-models and integrates them with the historical model to form an preliminary online ensemble model. Finally, it conducts local pruning and reconstruction on each sub-model, refining the final online ensemble model based on weighted posterior information. The efficacy of this approach is validated using synthetic, benchmark, and real DXN datasets from an MSWI plant in Beijing. Note to Practitioners-As an effective method for solid waste treatment, the MSWI process efficiently converts chemical energy into thermal and electrical energy. However, it also produces DXN as a by-product, posing significant risks that lead to the "Not-In-My-Back-Yard" effect for MSWI facilities. Currently, the mechanisms behind DXN generation, adsorption and emission remain unclear, and challenges such as high detection costs and significant lag hinder real-time detection efforts. To reduce DXN emission concentrations, operators in actual MSWI plants often resort to injecting large quantities of activated carbon for DXN adsorption. However, while effective, this method holds great importance for optimizing MSWI process in terms of pollution reduction control. Our objective is to achieve online soft measurement of DXN emissions, considering the dynamic operational conditions (conceptual drift). This approach can be applied by practitioners in similar industrial processes to enhance control and monitoring capabilities.