CN-121990624-A - Wet dust removal circulating water purification cooperative control method and system based on dust removal load feedforward compensation

CN121990624ACN 121990624 ACN121990624 ACN 121990624ACN-121990624-A

Abstract

The application discloses a wet dust removal circulating water purification cooperative control method and system based on dust removal load feedforward compensation, comprising the steps of constructing a multi-source heterogeneous data and dynamic space-time alignment database; the method comprises the steps of establishing a deep learning model, predicting flux characteristics of future pollutants according to working condition data of a dust removal system, establishing a hydrodynamic model, calculating dynamic time lag of pollutant transportation, taking the predicted flux characteristics and the dynamic time lag as feedforward signals, cooperatively controlling medicament adding equipment of a circulating water purification system, and carrying out closed-loop correction based on effluent quality feedback. According to the application, through prospective prediction of the dust removal load and accurate calculation of pollutant conveying time lag, predictive control of the purification system is realized, the problem of response delay of traditional feedback control is solved, the load impact resistance and control precision of the system are improved, the medicament proportion can be optimized, and the medicament consumption is reduced.

Inventors

ZHANG WENMING
MA ERWEI

Assignees

北京博创凯盛机械制造有限公司

Dates

Publication Date: 20260508
Application Date: 20260228

Claims (10)

1. The wet dust removal circulating water purification cooperative control method based on dust removal load feedforward compensation is characterized by comprising the following steps of: The method comprises the steps of firstly, constructing a multi-source heterogeneous data acquisition and dynamic space-time alignment database, wherein the first step comprises the steps of acquiring dust removal side characteristic data representing the running state of a dust removal system, a production rhythm signal representing the production rule of a pollution source, transmission side characteristic data representing the transportation process of pollutants and purification side characteristic data representing the running effect of a purification system, calculating the actual physical transmission lag time of a purification side water quality data sample in the purification side characteristic data in a pipe network by utilizing a dynamic sample alignment technology based on reverse virtual tracing, and combining the purification side water quality data sample, the dust removal side characteristic data of the retrospective source moment and the production rhythm signal into a space-time accurate aligned sample pair to construct a training data set; Training a dedusting load characteristic prediction model based on deep learning by using the training data set, wherein the model is used for predicting pollutant flux characteristics of a dosing point in a circulating water purification system to be reached in the future according to real-time dedusting side characteristic data and production rhythm signals; Step three, constructing a pollutant conveying time lag calculation model based on hydrodynamics, which is used for calculating the dynamic conveying time of pollutants from the dust remover to a dosing point in the circulating water purification system in real time; Step four, based on the pollutant flux characteristics predicted by the dust removal load characteristic prediction model and the dynamic transmission time calculated by the pollutant conveying time lag calculation model, feedforward compensation control is implemented on the circulating water purification system; And fifthly, executing feedback correction on the dust removal load characteristic prediction model according to the deviation of the water quality index of the outlet water of the circulating water purification system and the control target value.
2. The method according to claim 1, wherein the specific step of constructing the training data set by using the dynamic sample alignment technique based on reverse virtual tracing in the first step is: For each purification side water quality data sample in a historical database, a pipe network hydraulic model is utilized to input a historical pipe network instantaneous flow rate data sequence before the sampling time of the purification side water quality data sample, and the actual physical transmission lag time of a water group corresponding to the purification side water quality data sample in the pipe network is calculated back through reverse integral calculation; and combining the purification side water quality data sample, the dedusting side characteristic data at the retrospective source moment and the production rhythm signal into a sample pair with accurate time-space alignment by utilizing the actual physical transmission lag time.
3. The method of claim 1, wherein the dust load characteristic prediction model is a hybrid neural network that uses a long-term memory network in combination with an attention mechanism.
4. A method according to any one of claims 1 to 3, wherein the contaminant flux profile comprises a total mass concentration of suspended solids indicative of a load, and an equivalent particle size distribution index indicative of a load, wherein the equivalent particle size distribution index is a comprehensive index for quantifying settling properties of suspended particulate matter in wastewater, and is obtained based on a mapping relationship between source production operating parameters and physical settling properties of the wastewater.
5. The method according to claim 1, wherein the fluid mechanics based pollutant transportation time lag calculation model in the third step is implemented by a virtual trace particle method.
6. The method of claim 5, wherein the virtual trace particle method is implemented by: When the change rate of the pollutant flux characteristics predicted by the dust removal load characteristic prediction model exceeds a preset threshold value, generating a virtual trace particle carrying a time stamp and pollutant attributes at the position of a waste water outlet of the dust remover; The instantaneous flow velocity of the backwater pipe network is collected in real time, the displacement of the virtual trace particles is subjected to real-time integral deduction, and the accumulated displacement is obtained by carrying out time integral calculation on the instantaneous flow velocity v (tau) from the generation time T0 to the current time T; and when the accumulated displacement S (t) of the virtual tracer particles is equal to the physical length of the pipe network from the outlet of the dust remover to the dosing point, determining that the virtual tracer particles reach the dosing point, thereby obtaining the dynamic transmission time.
7. The method according to claim 1, wherein the feedforward compensation control implemented in the fourth step includes a feedforward dosing correction control, specifically: Adjusting the output of a dosing metering pump at a preset advance moment before the virtual tracer particles are expected to reach the dosing point; and implementing a dynamic proportioning strategy, when the predicted equivalent particle size distribution index is higher than a first preset threshold value, the predicted equivalent particle size distribution index indicates that coarse particles which are easy to settle are taken as main particles in the wastewater, the adding amount of the coagulant is increased, and when the predicted equivalent particle size distribution index is lower than a second preset threshold value, the predicted equivalent particle size distribution index indicates that fine particles are taken as main particles in the wastewater, the adding ratio of the coagulant is increased, and meanwhile, the coagulant adding ratio is increased.
8. The method according to claim 1 or 7, wherein the feedforward compensation control performed in the fourth step further includes a load leveling and sludge discharging linkage control, and the load leveling and sludge discharging linkage control includes at least one of the following control modes: when the predicted peak value of the total mass concentration of suspended matters exceeds the maximum treatment capacity under the current working condition of the circulating water purification system, a switching valve in a pipe network is controlled in advance, and high-turbidity wastewater is temporarily split into an accident emergency pool; And according to the total mass concentration of suspended matters in the pollutant flux characteristics, integrating and calculating the sludge quantity which is increased in a preset time period in the future by the sedimentation tank, and starting a sludge discharge pump or opening a sludge discharge valve in advance to perform preventive sludge discharge before the calculated sludge level reaches a high alarm threshold.
9. The method according to claim 1, wherein the specific way to perform feedback correction in the fifth step is: When the deviation duration time of the real-time monitoring sedimentation tank effluent turbidity and the control target value exceeds the preset duration, triggering an online learning or offline updating mechanism, and performing fine adjustment or retraining on the weight parameters of the dust removal load characteristic prediction model by utilizing the data which are acquired recently and subjected to space-time alignment processing.
10. The wet dust removal circulating water purification cooperative control system based on dust removal load feedforward compensation is characterized by comprising: The data construction module is configured to construct a multi-source heterogeneous data acquisition and dynamic space-time alignment database and comprises acquisition of dust removal side characteristic data representing the running state of a dust removal system, production rhythm signals representing the production rule of a pollution source, transmission side characteristic data representing the transportation process of pollutants and purification side characteristic data representing the running effect of a purification system, calculation of the actual physical transmission lag time of a purification side water quality data sample in the purification side characteristic data in a pipe network by using a dynamic sample alignment technology based on reverse virtual tracing, and combination of the purification side water quality data sample, the dust removal side characteristic data of the retrospective source moment and the production rhythm signals into a space-time accurate aligned sample pair to construct a training data set; The model training module is configured to train a deep learning-based dedusting load characteristic prediction model by utilizing the training data set and is used for predicting pollutant flux characteristics of a dosing point in a circulating water purification system in the future according to real-time dedusting side characteristic data and production rhythm signals; the time lag calculation module is configured to construct a pollutant conveying time lag calculation model based on hydrodynamics and is used for calculating the dynamic transmission time of pollutants from the dust remover to a dosing point in the circulating water purification system in real time; a cooperative control module configured to implement feedforward compensation control on the circulating water purification system based on the pollutant flux characteristics predicted by the dust removal load characteristic prediction model and the dynamic transmission time calculated by the pollutant transportation time lag calculation model; And the feedback correction module is configured to execute feedback correction on the dust removal load characteristic prediction model according to the deviation of the water quality index of the water outlet of the circulating water purification system and the control target value.

Description

Wet dust removal circulating water purification cooperative control method and system based on dust removal load feedforward compensation Technical Field The application relates to the technical field of industrial automation control and environmental protection equipment, in particular to a wet dust removal circulating water purification cooperative control method and system based on dust removal load feedforward compensation. Background In the heavy industrial production processes of metallurgy, electric power, chemical industry and the like, a wet dust removal system is a key environment-friendly device for treating dust-containing waste gas. The system captures particulate matters in the exhaust gas by a circulating water spray washing mode, and the generated dust-containing wastewater (turbid circulating water) is conveyed to a circulating water purification system. The circulating water purifying system is subjected to a series of physical and chemical treatments such as precipitation, coagulation, filtration and the like, so that the water quality is recovered to the technological requirements, and then the circulating water is recycled and supplied to the dust removing system for use. In the prior art, a wet dust removal system and a circulating water purification system are generally designed as two independent control units. The dust removal system is focused on guaranteeing the dust removal efficiency and safe operation, and the circulating water purification system is focused on standard treatment of the tail end water quality. This split control scheme exposes several technical problems in dealing with the dynamic changes of modern industrial production. First, there is a close physical coupling relationship and significant signal transmission hysteresis between the two systems. Industrial production (e.g., converting cycle of converter steelmaking) has strong periodic and intermittent characteristics, resulting in drastic fluctuation of dust removal load in a short time. The high-concentration dust-containing wastewater passes through a long-distance pipe network and a treatment tank with huge volume, and the water quality characteristics of the high-concentration dust-containing wastewater can be transmitted to key control nodes of a water treatment system, such as a dosing point, in tens of minutes or even hours. Second, conventional control strategies are difficult to effectively cope with such large hysteresis and high impact load changes. Currently, automatic dosing control of circulating water purification systems is mostly dependent on feedback control loops based on influent or effluent turbidity detection, such as proportional-integral-derivative (PID) control. An inherent disadvantage of this control scheme is the hysteresis of its response. When the turbidity sensor detects that the water quality is deteriorated, the high-concentration polluted water mass enters or passes through the processing unit, the dosage of the medicament is readjusted at the moment, and the high-concentration polluted water mass cannot be effectively processed due to the response hysteresis, so that the water quality of the discharged water exceeds the standard or the medicament is excessively added. Furthermore, the prior art lacks means for prospective sensing and fine tuning of contaminant physicochemical properties. The change of the dust removal load is not only reflected by the change of the concentration of suspended matters, but also accompanied by the dynamic change of the physical and chemical properties such as the particle size distribution, the density, the surface charge and the like of the particulate matters. These properties directly affect the effect of subsequent coagulating sedimentation and the type and formulation of the agent required. The traditional control system cannot span gas, liquid and solid multiphase media, and the fine characteristics of the terminal water quality can be prejudged from the production working condition data of the source, so that the optimal adding of the water treatment agent and the cooperative scheduling of the equipment cannot be realized. Disclosure of Invention The invention aims to provide a wet dust removal circulating water purification cooperative control method and system based on dust removal load feedforward compensation, which are used for solving the technical problems of response lag, insufficient load impact resistance, low control precision and the like caused by separation control of a wet dust removal system and a circulating water purification system in the background art. In one aspect of the invention, a wet dedusting circulating water purification cooperative control method based on dedusting load feedforward compensation is provided, and the method comprises the following steps: The method comprises the steps of firstly, constructing a multi-source heterogeneous data acquisition and dynamic space-time alignment database, wherein the f