CN-122006438-A - Membrane method denitration automatic control system and control method based on AI prediction

CN122006438ACN 122006438 ACN122006438 ACN 122006438ACN-122006438-A

Abstract

The invention provides an automatic membrane method denitration control system and method based on AI prediction, wherein the automatic membrane method denitration control system comprises a sensor unit, an execution mechanism unit, a control and monitoring unit and an AI prediction hardware unit, the control and monitoring unit comprises a PLC (programmable logic controller), a man-machine touch screen and an upper computer, the AI prediction hardware unit comprises an AI edge computing node, a data storage module and a communication module, the sensor unit is connected with the PLC, the execution mechanism unit is connected with the PLC, the PLC is respectively communicated with the man-machine touch screen and the upper computer, the AI edge computing node is in bidirectional connection with the PLC through the communication module, real-time operation data are acquired from the PLC by the AI edge computing node and stored in the data storage module, a prediction result and a control instruction are generated through model reasoning, and the prediction result and the control instruction are sent to the PLC for execution. Solves the problems of complex operation, low parameter control precision, safety interlocking deficiency and film cleaning hysteresis of the existing film method denitration system.

Inventors

ZHANG SHOULEI
ZHANG TINGYU
WANG GUANJUN
SUN YANLI

Assignees

山东泉益环保科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260129

Claims (10)

1. The membrane method denitration automatic control system based on AI prediction is characterized by comprising a sensor unit, an execution mechanism unit, a control and monitoring unit and an AI prediction hardware unit, wherein the sensor unit comprises a basic sensor and an AI prediction supplementary sensor, the execution mechanism unit comprises a basic execution mechanism and an execution mechanism with an AI control interface, the control and monitoring unit comprises a PLC (programmable logic controller), a man-machine touch screen and an upper computer, and the AI prediction hardware unit comprises an AI edge calculation node, a data storage module and a communication module; The sensor unit is connected with the PLC, the actuating mechanism unit is connected with the PLC, the PLC is communicated with the man-machine touch screen and the upper computer, the AI edge computing node is connected with the PLC in a two-way mode through the communication module, the AI edge computing node acquires real-time operation data from the PLC, stores the real-time operation data in the data storage module, generates a prediction result and a control instruction through model reasoning, and sends the prediction result and the control instruction to the PLC for execution.
2. The membrane method denitration automatic control system based on AI prediction as claimed in claim 1 is characterized in that the basic sensor comprises a pressure sensor, a flow sensor, a water quality sensor and a liquid level sensor, wherein the AI prediction supplementary sensor comprises a water inlet turbidity sensor and a water inlet temperature sensor, the pressure sensor is arranged at an inlet and an outlet of a monitoring membrane assembly and an inlet and an outlet of a security filter, the flow sensor is arranged on a raw water inlet pipeline, a membrane system water production pipeline and a concentrated water pipeline, the water quality sensor comprises a pH sensor arranged on a pretreatment water outlet pipeline, a conductivity sensor arranged on a water production pipeline and an ORP sensor arranged at an inlet of the security filter, the liquid level sensor is arranged in a raw water tank, a water production tank and a medical tank, the water inlet turbidity sensor is arranged on the raw water inlet pipeline, and the water inlet temperature sensor is arranged on the pretreatment water inlet pipeline.
3. The membrane method denitration automatic control system based on AI prediction as claimed in claim 1, wherein the basic execution mechanism comprises a variable frequency pump set, an electric regulating valve and an electromagnetic valve, and AI control signal interfaces are arranged in a cleaning pump in the variable frequency pump set and a medicament quantity regulating valve in the electric regulating valve and used for receiving cleaning flow and medicament concentration instructions.
4. The membrane method denitration automatic control system based on AI prediction as claimed in claim 1 is characterized in that an AI prediction result receiving module is arranged in the PLC controller and used for analyzing and executing AI instructions in a circulation interrupt organization block, an AI prediction interface is arranged in the man-machine touch screen and used for displaying future transmembrane pressure difference prediction curves, cleaning suggestions and risk levels, and an AI prediction model management module is arranged in the upper computer and used for supporting model training triggering, precision evaluation and data drift detection.
5. The automatic control system for membrane denitration based on AI prediction as claimed in claim 1 is characterized in that ARM architecture processors and NPUs are built in AI edge computing nodes to support TensorFlow Lite model reasoning, the data storage module is used for storing historical operation data, and the communication module is a module supporting Profinet IRT and 5G double-link communication.
6. The control method of the membrane method denitration automatic control system based on AI prediction according to any one of claims 1 to 5, comprising the steps of: the method comprises the steps that firstly, a system sequentially starts a pretreatment unit, performs membrane system low-pressure flushing and boosting operation and links a post-treatment unit according to a preset sequence through a one-key starting instruction until a normal operation condition is entered; the AI prediction hardware unit performs model reasoning based on data collected in real time, predicts future transmembrane pressure difference change and overpressure risk, and provides a cleaning strategy through the control and monitoring unit or issues an adjusting instruction in advance when the overpressure risk is predicted; Step three, through the stop triggered by a stop command or a safety interlock, the system executes the steps of membrane system depressurization, flushing, emptying and stop according to a preset sequence, and optimizes flushing parameters according to AI prediction results, and if the stop is caused by pollution, automatically triggers a cleaning mode, and executes a cleaning program according to a cleaning strategy generated by AI.
7. The control method of the membrane denitration automatic control system based on AI prediction as set forth in claim 6, wherein, The cleaning strategy is generated in a differentiated mode according to the pollution level: when the pollution level is low, performing delayed cleaning in a system load valley period, and adopting a first cleaning duration and a first medicament concentration; When the pollution level is medium, triggering a standard cleaning flow, including low-pressure flushing, first concentration reagent circulation and pure water rinsing, and adopting a second cleaning time period; when the pollution level is high, triggering the intensified cleaning process, improving the frequency of the cleaning pump and the concentration of the second medicament, prolonging the circulating time of the medicament, and adopting the third cleaning time.
8. The method for controlling the automatic membrane denitration control system based on AI prediction as claimed in claim 6, wherein the method is characterized by collecting historical operation data covering different working conditions, performing cleaning, labeling and feature engineering to construct a training dataset comprising time sequence features and target labels, training a GRU time sequence prediction model and a XGBoost risk classification model by using the training dataset, and performing precision verification and model light weight treatment to obtain a final combined AI prediction model.
9. The method according to claim 6, further comprising an AI prediction overpressure interlock step, wherein the PLC controls the pressure release valve to be opened in advance and reduces the frequency of the high-pressure pump to prevent overpressure when the AI prediction model outputs a risk that the future 10 minutes of the pre-membrane pressure will exceed 10% of the rated value.
10. The method for controlling an automatic control system for membrane denitration based on AI prediction as claimed in claim 6, further comprising a model operation and maintenance optimization step of performing incremental training and calibration on the AI prediction model by periodically using newly added real-time data, and monitoring hardware state and data drift to ensure prediction accuracy and control stability.

Description

Membrane method denitration automatic control system and control method based on AI prediction Technical Field The invention belongs to the technical field of monitoring systems, and particularly relates to an automatic membrane method denitration control system and method based on AI prediction. Background The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art. The membrane method denitration technology is used as a key technology of industrial wastewater denitration and brine refining in chlor-alkali industry, however, high dependence on manual operation is common at present. The system is started and stopped, the operations such as pretreatment valve opening, pump frequency adjustment, membrane system starting and the like are completed step by manual work, and if the operation sequence is in error, hydraulic impact and physical damage are very easy to cause to the membrane element. Meanwhile, the traditional system has insufficient precision in the aspect of control of operation parameters, mainly depends on manual experience to adjust the opening degree of a valve and the power of a pump, and is difficult to realize accurate and stable control on key process indexes such as pressure before a membrane, recovery rate and the like, thereby further causing concentration polarization aggravation and membrane pollution acceleration. The water production efficiency of the system is reduced, the replacement period of the membrane element is shortened, and the economical efficiency and the treatment efficiency of the system are severely restricted. The existing system lacks perfect safety interlocking mechanisms such as overpressure, water quality exceeding standard and the like, and when the pressure in front of the membrane exceeds the rated limit value or the concentration of produced water nitrate nitrogen is abnormal, emergency operations such as pressure relief, backflow or shutdown and the like cannot be automatically executed, so that the double risks of equipment failure and out-of-standard discharge are buried. In addition, the membrane cleaning process mostly adopts a passive strategy based on fixed time or fixed pressure difference, and cannot be dynamically adjusted in combination with actual working conditions such as water quality fluctuation, operation load change and the like, so that chemical agent waste and irreversible attenuation of membrane flux are caused by excessive cleaning or untimely cleaning. Disclosure of Invention Aiming at the problems, the invention provides an AI prediction-based automatic control system and a control method for membrane denitration, which solve the problems of complicated manual operation, low parameter control precision, safety interlock loss and membrane cleaning hysteresis of the existing membrane denitration system. In order to achieve the above object, the present invention is realized by the following technical scheme: The invention provides an automatic membrane denitration control system based on AI prediction, which comprises a sensor unit, an execution mechanism unit, a control and monitoring unit and an AI prediction hardware unit, wherein the sensor unit comprises a basic sensor and an AI prediction supplementary sensor, the execution mechanism unit comprises a basic execution mechanism and an execution mechanism with an AI control interface, the control and monitoring unit comprises a PLC (programmable logic controller), a man-machine touch screen and an upper computer, and the AI prediction hardware unit comprises an AI edge calculation node, a data storage module and a communication module; The sensor unit is connected with the PLC, the actuating mechanism unit is connected with the PLC, the PLC is communicated with the man-machine touch screen and the upper computer, the AI edge computing node is connected with the PLC in a two-way mode through the communication module, the AI edge computing node acquires real-time operation data from the PLC, stores the real-time operation data in the data storage module, generates a prediction result and a control instruction through model reasoning, and sends the prediction result and the control instruction to the PLC for execution. As a further implementation mode, the basic sensor comprises a pressure sensor, a flow sensor, a water quality sensor and a liquid level sensor, the AI prediction supplementing sensor comprises a water inlet turbidity sensor and a water inlet temperature sensor, the pressure sensor is arranged at an inlet and an outlet of the monitoring membrane component and an inlet and an outlet of the security filter, the flow sensor is arranged on a raw water inlet pipeline, a membrane system water production pipeline and a concentrated water pipeline, the water quality sensor comprises a pH sensor arranged on a pretreatment water outlet pipeline, a conductivity sensor arranged on a water production