Fenton process intelligent control system and method
By optimizing the dosing of Fenton reaction reagents through neural network models and visualization methods, the problem of low automation in the Fenton reaction process was solved, real-time adjustment of reagents and stable effluent were achieved, and operating costs and risks were reduced.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHONGQING CISDI THERMAL & ENVIRONMENTAL ENG CO LTD
- Filing Date
- 2026-04-07
- Publication Date
- 2026-07-17
AI Technical Summary
The nonlinear, time-delayed, and multivariable characteristics of the Fenton reaction process result in low automation of existing control methods, delayed reagent dosing, unstable COD in effluent, and the risk of exceeding standards.
Neural network models, especially LSTM, Attention-LSTM and CNN-LSTM network models, are used in conjunction with data acquisition, analysis and control modules to monitor the Fenton reaction process in real time, optimize the dosage of the reagent, and provide visual monitoring through the central control display screen and the alum flower image acquisition device.
It enables real-time monitoring of the Fenton reaction process, reduces manual intervention, improves the system's automation level, reduces the risk of reagent waste and excessive effluent, lowers operating costs, and enhances the early warning capability for abnormal operating conditions.
Smart Images
Figure CN122403608A_ABST