Multi-stage filtration and stripping desulfurization purification method for storage batteries

Abstract

This invention relates to the field of resource recycling and environmental protection technology, and discloses a method for multi-stage sulfuric acid filtration and air-stripping desulfurization purification of storage batteries. The method includes: constructing a digital twin containing geometric, physical, behavioral, and rule models; collecting data in real time through a multi-source sensor network and driving the dynamic evolution of the twin; simulating the multi-stage filtration flow field and air-stripping desulfurization reaction process within the twin to generate a spatiotemporal evolution field of key parameters; identifying abnormal signs based on adaptive thresholds and tracing the root cause of faults using a Bayesian network; and generating a collaborative control instruction set through multi-objective optimization to achieve closed-loop control of the physical system's actuators. This invention achieves trial-and-error-free process optimization and early fault warning through virtual-real fusion, reducing commissioning costs, extending filter material life, and minimizing unplanned downtime.

Description

Technical Field

[0001] This invention belongs to the field of resource recycling and environmental protection technology, specifically relating to a multi-stage sulfuric acid filtration and gas stripping desulfurization purification method for storage batteries. Background Technology

[0002] With the continuous growth in demand for the use of acid batteries in the industrial sector, sulfuric acid purification, as a core component, directly affects the efficiency and environmental compliance of the entire regeneration system due to its process stability and operational reliability. Traditional sulfuric acid purification systems generally adopt a multi-stage treatment process combining physical filtration and air stripping desulfurization, relying on fixed-cycle maintenance and empirical parameter settings for operation and management.

[0003] However, traditional sulfuric acid purification systems have the following problems in actual operation: the commissioning of new processes or equipment modifications is highly dependent on on-site physical tests, requiring repeated adjustments to key parameters such as filter membrane pore size, air volume of the stripping tower, and liquid-gas ratio, resulting in long commissioning cycles, large material losses, and high labor costs; the system lacks the ability to perceive and predict the evolution of key components such as the degree of filter membrane blockage and corrosion and aging of stripping tower packing in real time, and faults are often only discovered after a sudden shutdown, causing production interruptions, secondary pollution, or even safety accidents.

[0004] Digital twin-based industrial process optimization technology has demonstrated significant potential in fields such as chemical engineering and power generation in recent years. This technology constructs a high-fidelity virtual mapping of the physical system, integrating historical operating data and multi-source sensor information to achieve dynamic simulation and intelligent extrapolation of equipment status, process performance, and potential risks. In the sulfuric acid purification scenario, establishing a digital twin model encompassing fluid dynamics, mass transfer reaction mechanisms, and material degradation laws will provide a novel approach for the virtual verification of process parameters and the forward-looking assessment of the lifespan of key components.

[0005] While existing technologies have attempted to introduce online monitoring and simple threshold alarm mechanisms, they generally suffer from problems such as static models, data silos, and weak predictive capabilities. Most systems can only reflect the current operating conditions and cannot simulate long-term evolution trends under different operating conditions; for progressive failures such as filter membrane flux decline and packing surface fouling, there is a lack of lifetime prediction methods based on mechanism and data fusion; and it is impossible to safely and cost-effectively conduct process optimization experiments with multiple parameter combinations in virtual space.

[0006] Therefore, in highly corrosive and high-risk sulfuric acid purification environments, there is an urgent need for an adaptive purification method that deeply integrates digital twin technology to achieve integrated intelligent operation and maintenance that includes virtual process commissioning, component life prediction, and proactive fault warning. This would significantly reduce commissioning costs, mitigate the risk of unplanned downtime, and improve the overall resilience and intelligence of the system. Summary of the Invention

[0007] This invention provides a multi-stage filtration and desulfurization purification method for sulfuric acid in storage batteries, aiming to solve the technical problems of high costs due to the heavy reliance on physical testing for equipment commissioning and the high risk of downtime caused by weak fault early warning capabilities in existing technologies. This method constructs a high-fidelity digital twin to achieve virtual mapping, dynamic simulation, and intelligent control of the entire sulfuric acid electrolyte purification process. It performs full-element modeling, real-time state simulation, and abnormal behavior prediction of filtration and desulfurization operations in the physical space within the information space. This allows for process parameter optimization without repeated trial and error with physical equipment, and early identification of potential failure modes during operation, ensuring continuous and stable system operation.

[0008] This invention provides a method for multi-stage filtration and gas stripping desulfurization purification of sulfuric acid in storage batteries, comprising:

[0009] Construct a digital twin that corresponds one-to-one with the physical purification system. The digital twin includes a geometric model, a physical model, a behavioral model, and a rule model.

[0010] The physicochemical state data of sulfuric acid, equipment operating parameters and environmental conditions are collected in real time by a multi-source sensor network deployed in the physical purification system, and the data is synchronized to the digital twin to drive its dynamic evolution.

[0011] The fluid dynamics simulation of the multi-stage filtration process and the gas stripping desulfurization reaction kinetics simulation are performed in the digital twin to generate the spatiotemporal evolution field of sulfuric acid suspended particulate matter concentration distribution, crystal particle size spectrum, dissolved sulfide concentration gradient and gas mass transfer efficiency.

[0012] Based on the aforementioned spatiotemporal evolution field, an adaptive threshold determination mechanism is used to identify filter media clogging trends, desulfurization efficiency decay inflection points, and signs of abnormal system operation.

[0013] When abnormal signs are detected, the fault root cause tracing algorithm is activated, and by combining the historical operation database with real-time simulation results, the key components or process parameter deviations that cause performance degradation are located.

[0014] Based on the positioning results, multi-objective optimization is performed in the digital twin to generate a set of coordinated control instructions, including filter differential pressure setpoint, air lift carrier gas flow rate, circulating pump speed and backwash cycle.

[0015] The coordinated control instruction set is sent to the actuator of the physical purification system to achieve closed-loop control of the multi-stage filtration unit and the gas stripping desulfurization unit.

[0016] Preferably, constructing a digital twin that corresponds one-to-one with the physical purification system specifically includes:

[0017] A three-dimensional geometric topology of a three-stage filtration module consisting of a coarse filtration unit, a fine filtration unit, and an ultrafiltration unit is established. The coarse filtration unit uses a sintered metal filter element, the fine filtration unit uses a polytetrafluoroethylene pleated filter membrane, and the ultrafiltration unit uses a hollow fiber ultrafiltration membrane.

[0018] An internal component model of the gas stripping desulfurization tower was established, including a gas distributor, packing layer, demister, and liquid seal device. The packing layer was filled with polypropylene stepped rings with a specific surface area.

[0019] The Darcy-Fochheimer flow equation for sulfuric acid in each stage of the filtration unit and the two-membrane mass transfer model in the stripping tower were established. The two-membrane mass transfer model is coupled with Henry's law and the first-order chemical reaction kinetic equation to describe the migration rate of hydrogen sulfide from the liquid phase to the gas phase.

[0020] A degradation rule model is established for equipment wear, filter media aging, and scale accumulation. The degradation rule model takes the cumulative throughput, runtime, and differential pressure change rate as input variables and outputs the filter media permeability decay coefficient and the packing effective specific surface area reduction rate.

[0021] Preferably, the multi-source sensor network includes:

[0022] Pressure transmitters installed at the inlet and outlet of the coarse filter unit are used to measure the filtration pressure difference.

[0023] A laser particle size analyzer installed at the outlet of the fine filtration unit is used for online monitoring of the particle size distribution of suspended particulate matter.

[0024] A turbidity meter installed on the permeate side of the ultrafiltration unit is used to detect the turbidity of the permeate.

[0025] The pH meter and redox potential sensor installed at the liquid inlet of the stripping tower are used to characterize the acidity, alkalinity and oxidation state of the electrolyte.

[0026] An ultraviolet fluorescence hydrogen sulfide analyzer installed at the exhaust port of the gas stripping tower is used to quantitatively detect the concentration of hydrogen sulfide in the exhaust gas.

[0027] The current transformer and vibration acceleration sensor installed on the circulating pump motor are used to monitor the pump load and mechanical condition.

[0028] All sensors have a unified sampling frequency, and the data is transmitted in real time to the data access layer of the digital twin via the industrial Ethernet protocol.

[0029] Preferably, the fluid dynamics simulation of performing a multi-stage filtration process in a digital twin specifically includes:

[0030] Based on real-time collected inlet flow rate, temperature and viscosity data, the boundary conditions of the Navier-Stokes equations are initialized.

[0031] The flow field inside the filter cavity is discretized and solved using the finite volume method to obtain the velocity vector field and pressure field;

[0032] The filter cake formation rate and local clogging area are calculated by combining the filter media permeability attenuation coefficient;

[0033] The particle tracking algorithm is used to simulate the motion trajectory of particles of different sizes in the flow field and predict their deposition density on the surface of filter media at each level.

[0034] When the deposition density in a certain area exceeds a preset threshold, the area is marked as a high-blockage risk zone, and a local backwashing strategy simulation is triggered.

[0035] Preferably, the gas stripping desulfurization reaction kinetic simulation specifically includes:

[0036] The liquid phase sulfide concentration field in the stripping tower is initialized using the real-time measured influent sulfide concentration, pH value and temperature as initial conditions.

[0037] Set the carrier gas to air, the flow range, and the orifice ratio of the gas distributor;

[0038] Solve the continuity equation and momentum equation for gas-liquid two-phase flow to obtain the bubble diameter distribution and rising velocity;

[0039] Based on the two-film theory, the hydrogen sulfide partial pressure difference at the gas-liquid interface is calculated, and combined with the first-order reaction rate constant, the decay curve of sulfide concentration in the bulk liquid phase with the height of the tower is solved.

[0040] When the sulfide concentration in the effluent at the top of the tower exceeds 10 mg / L, the desulfurization efficiency is deemed insufficient, and a parameter scanning simulation of the carrier gas flow rate and liquid-gas ratio is initiated.

[0041] Preferably, the adaptive threshold determination mechanism specifically includes:

[0042] Set a dynamic upper limit threshold for the filter differential pressure;

[0043] Set a fixed upper limit threshold for the turbidity of ultrafiltration permeate;

[0044] Set a fixed upper limit threshold for the concentration of hydrogen sulfide in exhaust gas;

[0045] When any monitoring parameter exceeds the corresponding threshold for three consecutive sampling periods, the abnormal symptom identification module is activated.

[0046] Preferably, the root cause tracing algorithm adopts a Bayesian network structure, and the nodes include filter media blockage, pump performance degradation, gas distributor blockage, packing fouling, and sensor drift. The edge weights are determined by conditional probability statistics in the historical fault case library.

[0047] The algorithm takes the current combination of parameters that exceed the limit as input and outputs the posterior probability of each fault node.

[0048] When the posterior probability of a filter media blockage node is greater than 80%, it is determined to be the main cause, and the degradation rule model corresponding to the filter media is called to predict the remaining life.

[0049] Preferably, the multi-objective optimization solution takes minimizing energy consumption, maximizing desulfurization efficiency, and extending filter material life as optimization objectives, and the constraints include the maximum allowable pressure difference, the minimum circulation flow rate, and the safe operating temperature range.

[0050] A non-dominated sorting genetic algorithm was used to iteratively optimize the decision variables, which included the coarse filter backwashing cycle, the fine filter bypass valve opening, the ultrafiltration crossflow velocity, the airlift carrier gas flow rate, and the circulation pump frequency.

[0051] The optimization results generate a set of Pareto optimal solutions, from which the solution with the highest comprehensive score is selected as the coordinated control instruction set.

[0052] Preferably, the actuator of the physical purification system includes:

[0053] The pneumatic backwash valve installed at the bottom of the coarse filter unit is controlled by backwash cycle commands issued by the digital twin.

[0054] The opening degree of the electric regulating valve installed in the bypass pipeline of the fine filtration unit is set by the bypass valve opening command.

[0055] The frequency converter installed in the ultrafiltration circulation loop has its output frequency calculated from the cross-flow velocity command.

[0056] The speed of the Roots blower installed at the bottom of the air lift tower is controlled by the carrier gas flow command after conversion through a flow-speed mapping table.

[0057] The motion feedback signals of all actuators are transmitted back to the digital twin in real time to correct the execution deviations of the simulation model.

[0058] Preferably, the mathematical expression of the Darcy-Fochheimer flow equation is:

[0059] ;

[0060] For pressure, For pressure gradient, For dynamic viscosity, For filter media permeability, For flow velocity vectors, For fluid density, This is the inertial drag coefficient.

[0061] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0062] 1. This invention transforms the sulfuric acid purification process of batteries from a physical trial-and-error mode to a virtual simulation-driven mode by constructing a high-fidelity digital twin, eliminating the reliance on a large number of physical materials and manual intervention in the traditional debugging process and reducing equipment debugging costs.

[0063] 2. Digital twins can simulate system responses under different operating conditions in advance based on multi-physics coupling simulation, thereby achieving global optimization of process parameters and avoiding efficiency losses caused by repeated on-site adjustments.

[0064] 3. During the operation phase, by integrating real-time sensor data and dynamic simulation results, this invention achieves early warning and accurate root cause location for key failure modes such as filter blockage and desulfurization failure, with the warning time window being earlier than that of traditional threshold alarm methods.

[0065] 4. The generated collaborative control instruction set comprehensively considers energy consumption, efficiency and equipment lifespan, so that the system maintains the optimal operating state throughout its entire life cycle, extends the average service life of filter media, and reduces the number of unplanned downtimes.

[0066] 5. The online calibration mechanism of the digital twin ensures the long-term effectiveness of the model, overcomes the defect of traditional static models becoming inaccurate as equipment ages, and provides reliable technical support for green and intelligent manufacturing in the field of battery recycling. Attached Figure Description

[0067] Figure 1 This is a schematic diagram of the overall technical solution architecture of the present invention;

[0068] Figure 2 This is a schematic diagram of the core principle framework of the digital twin-driven multiphysics coupling simulation and adaptive control in this invention;

[0069] Figure 3 This is a logical flow diagram of the multi-stage filtration and air stripping desulfurization physical purification subsystem in this invention.

[0070] Figure 4 This is a logical flowchart of the digital twin construction and dynamic evolution mechanism in this invention;

[0071] Figure 5 This is a closed-loop decision-making logic framework diagram of anomaly identification, root cause tracing, and collaborative optimization control in this invention;

[0072] Figure 6 This is a schematic diagram of the multi-level interaction relationship and data flow between the physical purification system and the digital twin in this invention. Detailed Implementation

[0073] refer to Figures 1 to 6This invention provides a method for multi-stage sulfuric acid filtration and gas stripping desulfurization purification of storage batteries. This method constructs a high-fidelity digital twin to achieve full-element mapping, dynamic simulation, and closed-loop control of the physical purification system. The method strictly follows the following steps:

[0074] S1, constructing a digital twin that corresponds one-to-one with the physical purification system;

[0075] S2 collects physical and chemical state data, equipment operating parameters and environmental conditions in real time through a multi-source sensor network and synchronizes them to the digital twin.

[0076] S3 performs fluid dynamics simulation of multi-stage filtration process and gas stripping desulfurization reaction dynamics simulation in a digital twin, generating a spatiotemporal evolution field;

[0077] S4. Based on the spatiotemporal evolution field, an adaptive threshold determination mechanism is used to identify abnormal signs.

[0078] S5, When abnormal signs are detected, the fault root cause tracing algorithm is activated to locate the root cause of performance degradation;

[0079] S6, based on the positioning results, performs multi-objective optimization in the digital twin to generate a set of coordinated control instructions;

[0080] S7, the coordinated control instruction set is sent to the actuator of the physical purification system to achieve closed-loop control.

[0081] In step S1, a digital twin corresponding one-to-one with the physical purification system is constructed. The digital twin includes a geometric model, a physical model, a behavioral model, and a rule model. The geometric model accurately reproduces the spatial topology of the physical purification subsystem, including the three-dimensional solid form and internal component layout of the coarse filtration unit, fine filtration unit, ultrafiltration unit, and air stripping desulfurization tower.

[0082] The coarse filtration unit is filled with a sintered metal filter element with a pore size of 50 micrometers. Its geometric features include filter element length, outer diameter, porosity, and supporting skeleton structure. The fine filtration unit uses a polytetrafluoroethylene pleated filter membrane with a pore size of 5 micrometers. Its geometric model needs to describe the pleat density, effective filtration area, and sealing end cap structure. The ultrafiltration unit is composed of a hollow fiber ultrafiltration membrane with a molecular weight cutoff of 100,000 Daltons. Its geometric model includes the fiber bundle arrangement, inner diameter, wall thickness, and distribution of the encapsulating adhesive layer.

[0083] The geometric model of the gas stripping desulfurization tower includes the tower height, diameter, gas distributor opening layout, packing layer filling height, demister blade inclination angle, and liquid seal device level control structure. The packing layer is filled with polypropylene stepped rings with a specific surface area of ​​200 square meters per cubic meter, and its individual geometric dimensions and packing void ratio are parametrically modeled.

[0084] The physical model describes the physical laws governing mass transport and energy conversion within the system. For the multi-stage filtration process, the Darcy-Fochheimer flow equation is established to characterize the flow properties of sulfuric acid electrolyte in a heterogeneous porous medium. Its mathematical expression is:

[0085] ;

[0086] For pressure, For pressure gradient, For dynamic viscosity, For filter media permeability, For flow velocity vectors, For fluid density, This represents the inertial drag coefficient. The equation is coupled with the Navier-Stokes equations to solve for the velocity and pressure field distributions within the filter cavity.

[0087] A two-film mass transfer model was established for the gas stripping desulfurization process. Based on the two-film theory, this model assumes the existence of stagnant films on both sides of the gas-liquid interface. Hydrogen sulfide diffuses from the bulk liquid phase to the interface, then crosses the gas film and enters the bulk gas phase. The mass transfer rate is determined by both liquid film resistance and gas film resistance, coupled with first-order chemical reaction kinetics equations. The concentration of sulfide in the liquid phase... Over time With tower height The changes satisfy:

[0088] ;

[0089] The liquid phase flow rate is... Let be the liquid phase diffusion coefficient. This is the first-order reaction rate constant. The partial pressure of hydrogen sulfide in the gas phase. The relationship between liquid phase concentration and Henry's Law is as follows: , is the Henry's constant.

[0090] The behavioral model defines the equipment's response logic under different operating commands, including start / stop sequences, backwash trigger conditions, bypass switching logic, and sewage discharge cycle control. For example, when the differential pressure of the coarse filter unit exceeds the threshold, the behavioral model activates the backwash subroutine, sequentially executing actions such as closing the inlet valve, opening the backwash valve, starting the compressed air pulse, and emptying the filter cake.

[0091] Rule-based models, on the other hand, characterize the inherent patterns of system performance degradation over time, using cumulative processing volume. runtime and pressure difference change rate Input variable, output filter media permeability attenuation coefficient Compared with the effective specific surface area reduction rate of the filler The penetration rate decay model is expressed as:

[0092] ;

[0093] and The material aging coefficient is determined by fitting historical operating data.

[0094] In step S2, data is collected in real time through a multi-source sensor network deployed in the physical purification system. The sensor network includes pressure transmitters installed at the inlet and outlet of the coarse filtration unit, with a sampling frequency of 10 Hz and a measurement accuracy of 0.5% of full scale, used to calculate the filtration pressure difference. ;

[0095] The laser particle size analyzer installed at the outlet of the fine filtration unit uses the forward scattering principle to monitor the particle size distribution of suspended particulate matter ranging from 0.1 to 100 micrometers online, and outputs the volume-weighted median particle size. The proportion of particles larger than 10 micrometers ;

[0096] The turbidity meter installed on the permeate side of the ultrafiltration unit detects the turbidity of the permeate based on the intensity of 90-degree scattered light. The range is 0 to 10 NTU, and the resolution is 0.01 NTU.

[0097] The pH meter and redox potential sensor installed at the liquid inlet of the stripping tower have a pH measurement range of 0 to 2 with an accuracy of ±0.05 and a redox potential measurement range of -500 to +1000 mV with an accuracy of ±5 mV.

[0098] The ultraviolet fluorescence hydrogen sulfide analyzer installed at the exhaust port of the stripping tower has a detection limit of 0.1 mg / m³ and a range of 0 to 100 mg / m³.

[0099] The current transformer and vibration acceleration sensor installed on the circulating pump motor have a current measurement accuracy of 0.5%, a vibration acceleration measurement range of 0 to 50g, and a frequency response range of 10 Hz to 10000 Hz. All sensors transmit timestamped data packets to the data access layer of the digital twin via the industrial Ethernet protocol, with a time synchronization error of less than 10 milliseconds.

[0100] In step S3, fluid dynamics simulation of the multi-stage filtration process and gas stripping desulfurization reaction kinetics simulation are performed in the digital twin. The fluid dynamics simulation uses real-time acquired inlet flow rate data. ,temperature and viscosity The Navier-Stokes equations are initialized as boundary conditions. The filter cavity is meshed using the finite volume method with an unstructured mesh size of no less than 500,000. Prismatic layer meshes are used in the near-wall region to analyze the boundary layer effect.

[0101] The solver employs the pressure-velocity coupled SIMPLE algorithm, with the iterative convergence residual set to 10. -5 Combined with the filter media permeability attenuation coefficient The local Darcy velocity was calculated, and particle motion was simulated using a particle tracking algorithm. The particle tracking model considered Stokes drag, gravity, Brownian motion, and fluid turbulence fluctuations, tracking no fewer than 100,000 representative particles and calculating their deposition density on the surfaces of various filter media. When a certain area Exceeding the preset threshold When the number of particles is 5,000 per square centimeter (e.g., 5,000 particles per square centimeter), it is marked as a high-risk area for clogging.

[0102] Kinetic simulation of air stripping desulfurization reaction based on measured influent sulfide concentration pH value and temperature As an initial condition, the carrier gas is set to air, and the flow rate is... Adjustable from 50 to 300 standard cubic meters per hour, the gas distributor has an orifice ratio of 15% and an orifice diameter of 3 mm.

[0103] The gas-liquid two-phase flow model was solved using the Euler-Euler method, with both the gas and liquid phases considered as continuous media. The bubble diameter distribution was calculated using an algebraic slip model. A two-film mass transfer model was used to calculate the interfacial hydrogen sulfide partial pressure difference. Mass transfer flux , The mass transfer coefficient in the liquid phase is... The equilibrium concentration at the interface is obtained by integrating along the column height. ,when A value of milligrams per liter indicates insufficient desulfurization efficiency.

[0104] In step S4, an adaptive threshold determination mechanism is used to identify abnormal signs. The dynamic upper limit threshold for the filtered differential pressure is used. The calculation formula is:

[0105] ;

[0106] This represents the initial pressure differential of the new filter media, in kilopascals. The unit is cubic meters. The fixed upper limit threshold for ultrafiltration permeate turbidity is 0.5 NTU, and the fixed upper limit threshold for hydrogen sulfide concentration in exhaust gas is 20 mg / m³. The judgment logic is as follows: if any monitoring parameter exceeds its corresponding threshold for three consecutive sampling cycles (i.e., 300 milliseconds), the abnormal symptom identification module is activated, and the vector of parameters exceeding the limit is recorded. . The turbidity of the ultrafiltrate. This refers to the concentration of hydrogen sulfide in the exhaust gas.

[0107] In step S5, the root cause analysis algorithm is initiated. This algorithm employs a Bayesian network structure, with the node set including filter media clogging, pump performance degradation, gas distributor clogging, packing fouling, and sensor drift. Edge weights are determined by conditional probabilities from a historical fault case database. Statistical determination. Algorithm reception. As evidence, the posterior probability of each fault node is calculated using the belief propagation algorithm. . Filter media clogging The issue stems from excessive coarse filtration pressure differential. When the posterior probability of a filter media blockage node exceeds 80%, it is identified as the primary cause, and the corresponding degradation rule model for that filter media is used to predict its remaining service life. , For rated processing capacity, This represents the average daily processing volume.

[0108] In step S6, a multi-objective optimization solution is performed. The optimization objective function is... for:

[0109] ;

[0110] The total energy consumption of the system. For desulfurization efficiency, For the remaining life of the filter media, , , These are normalized weighting coefficients. Constraints include: maximum allowable pressure difference. kPa, minimum circulation flow rate cubic meters per hour, safe operating temperature Celsius.

[0111] Decision variable vector :

[0112] ;

[0113] These represent the coarse filter backwashing cycle, fine filter bypass valve opening, ultrafiltration crossflow velocity, airlift carrier gas flow rate, and circulation pump frequency, respectively. A non-dominated sorting genetic algorithm is used, with a population size of 100, a crossover probability of 0.9, a mutation probability of 0.1, a maximum iteration of 200 generations, and outputs a Pareto front. The solution with the highest comprehensive score is selected as the collaborative control instruction set.

[0114] In step S7, the coordinated control command set is sent to the actuators. The pneumatic backwash valve at the bottom of the coarse filtration unit receives the backwash cycle command and opens according to the set sequence; the electric regulating valve of the bypass pipeline of the fine filtration unit adjusts the opening percentage according to the bypass valve opening command; the frequency converter of the ultrafiltration circulation loop converts the cross-flow velocity command into a motor frequency output; the Roots blower at the bottom of the air lift tower converts the carrier gas flow command into a speed control signal through a flow-speed mapping table. The action completion signals of all actuators and the actual operating parameters are transmitted back to the digital twin in real time for calibration of the simulation model.

[0115] The battery sulfuric acid multi-stage filtration and air-stripping desulfurization purification system based on digital twins includes a physical purification subsystem, a multi-source sensing subsystem, a digital twin construction and simulation subsystem, an anomaly identification and root cause tracing subsystem, a collaborative optimization and control command generation subsystem, a command execution and feedback subsystem, and a data synchronization and model correction subsystem.

[0116] The physical purification subsystem consists of a coarse filter tank, a fine filter tank, an ultrafiltration module, a stripping desulfurization tower, and an acid storage tank connected in sequence. The units are connected by acid-resistant stainless steel pipes, and the circulation pump is an acid-resistant centrifugal pump made of high-silicon cast iron.

[0117] The digital twin construction and simulation subsystem is deployed on an industrial server cluster and adopts a containerized architecture. The geometric modeling layer uses the OpenCASCADE engine, the physics engine layer integrates the OpenFOAM solver and the Cantera chemical reaction library, the behavioral simulation layer is based on a state machine model, and the rule reasoning layer uses the Drools rule engine to load a degenerate rule library.

[0118] The anomaly detection and root cause analysis subsystem incorporates a dynamic threshold manager with an hourly threshold update cycle. Its Bayesian inference engine loads a pre-trained network, achieving inference latency of less than 50 milliseconds. In the collaborative optimization and control instruction generation subsystem, the instruction compiler supports ModbusTCP and PROFINET protocol conversion.

[0119] The data synchronization and model calibration subsystem uses the PTP precision time protocol to achieve timestamp alignment with a synchronization accuracy of better than 100 milliseconds. The extended Kalman filter algorithm uses the execution feedback data as the observation value and updates key parameters such as filter media permeability, packing effective specific surface area and pump efficiency every 10 minutes.

[0120] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0121] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for multi-stage filtration and gas stripping desulfurization purification of sulfuric acid in storage batteries, characterized in that, include: Constructing a digital twin that corresponds one-to-one with the physical purification system, specifically including: A three-dimensional geometric topology of a three-stage filtration module consisting of a coarse filtration unit, a fine filtration unit, and an ultrafiltration unit is established. The coarse filtration unit uses a sintered metal filter element, the fine filtration unit uses a polytetrafluoroethylene pleated filter membrane, and the ultrafiltration unit uses a hollow fiber ultrafiltration membrane. An internal component model of the gas stripping desulfurization tower was established, including a gas distributor, packing layer, demister, and liquid seal device. The packing layer was filled with polypropylene stepped rings with a specific surface area. Darcy-Fochheimer flow equations for sulfuric acid electrolyte in each stage of filtration units and a two-membrane mass transfer model in a stripping tower are established. The two-membrane mass transfer model is coupled with Henry's law and first-order chemical reaction kinetic equations to describe the migration rate of hydrogen sulfide from the liquid phase to the gas phase. A degradation rule model for equipment wear, filter media aging, and scale accumulation is established. The degradation rule model takes the cumulative throughput, runtime, and differential pressure change rate as input variables and outputs the filter media permeability decay coefficient and the packing effective specific surface area reduction rate. The digital twin includes a geometric model, a physical model, a behavioral model, and a rule model; The physicochemical state data of sulfuric acid, equipment operating parameters and environmental conditions are collected in real time by a multi-source sensor network deployed in the physical purification system, and the data is synchronized to the digital twin to drive its dynamic evolution. The digital twin performs fluid dynamics simulation of a multi-stage filtration process and gas stripping desulfurization reaction kinetics simulation, generating a spatiotemporal evolution field of suspended particulate matter concentration distribution, crystal size spectrum, dissolved sulfide concentration gradient, and gas mass transfer efficiency in sulfuric acid. Specifically, this includes: Based on real-time collected inlet flow rate, temperature and viscosity data, the boundary conditions of the Navier-Stokes equations are initialized. The flow field inside the filter cavity is discretized and solved using the finite volume method to obtain the velocity vector field and pressure field; The filter cake formation rate and local clogging area are calculated by combining the filter media permeability attenuation coefficient; The particle tracking algorithm is used to simulate the motion trajectory of particles of different sizes in the flow field and predict their deposition density on the surface of filter media at each level. When the deposition density in a certain area exceeds a preset threshold, the area is marked as a high-blockage risk zone, and a local backwashing strategy simulation is triggered. The liquid phase sulfide concentration field in the stripping tower is initialized using the real-time measured influent sulfide concentration, pH value and temperature as initial conditions. Set the carrier gas to air, the flow range, and the orifice ratio of the gas distributor; Solve the continuity equation and momentum equation for gas-liquid two-phase flow to obtain the bubble diameter distribution and rising velocity; Based on the two-film theory, the partial pressure difference of hydrogen sulfide at the gas-liquid interface is calculated, and combined with the first-order reaction rate constant, the decay curve of sulfide concentration in the bulk liquid phase with the height of the tower is solved. When the sulfide concentration in the effluent at the top of the tower exceeds 10 mg / L, the desulfurization efficiency is deemed insufficient, and a parameter scanning simulation of the carrier gas flow rate and liquid-gas ratio is initiated. Based on the aforementioned spatiotemporal evolution field, an adaptive threshold determination mechanism is used to identify filter media clogging trends, desulfurization efficiency decay inflection points, and signs of abnormal system operation. When abnormal signs are detected, the fault root cause tracing algorithm is activated, and by combining the historical operation database with real-time simulation results, the key components or process parameter deviations that cause performance degradation are located. Based on the positioning results, multi-objective optimization is performed in the digital twin to generate a set of coordinated control instructions, including filter differential pressure setpoint, air lift carrier gas flow rate, circulating pump speed and backwash cycle. The coordinated control instruction set is sent to the actuator of the physical purification system to achieve closed-loop control of the multi-stage filtration unit and the gas stripping desulfurization unit.

2. The method for multi-stage sulfuric acid filtration and gas stripping desulfurization purification of storage batteries according to claim 1, characterized in that, The multi-source sensor network includes: Pressure transmitters installed at the inlet and outlet of the coarse filter unit are used to measure the filtration pressure difference. A laser particle size analyzer installed at the outlet of the fine filtration unit is used for online monitoring of the particle size distribution of suspended particulate matter. A turbidity meter installed on the permeate side of the ultrafiltration unit is used to detect the turbidity of the permeate. The pH meter and redox potential sensor installed at the liquid inlet of the stripping tower are used to characterize the acidity, alkalinity and oxidation state of the electrolyte. An ultraviolet fluorescence hydrogen sulfide analyzer installed at the exhaust port of the gas stripping tower is used to quantitatively detect the concentration of hydrogen sulfide in the exhaust gas. The current transformer and vibration acceleration sensor installed on the circulating pump motor are used to monitor the pump load and mechanical condition. All sensors have a unified sampling frequency, and the data is transmitted in real time to the data access layer of the digital twin via the industrial Ethernet protocol.

3. The method for multi-stage sulfuric acid filtration and gas stripping desulfurization purification of storage batteries according to claim 2, characterized in that, The adaptive threshold determination mechanism specifically includes: A dynamic upper limit threshold is set for the filtration pressure difference, which is equal to the initial clean state pressure difference plus an incremental term based on the cumulative throughput. Set a fixed upper limit threshold for the turbidity of ultrafiltration permeate; Set a fixed upper limit threshold for the concentration of hydrogen sulfide in exhaust gas; When any monitoring parameter exceeds the corresponding threshold for three consecutive sampling periods, the abnormal symptom identification module is activated.

4. The method for multi-stage sulfuric acid filtration and gas stripping desulfurization purification of storage batteries according to claim 3, characterized in that, The fault root cause tracing algorithm adopts a Bayesian network structure, with nodes including filter media blockage, pump performance degradation, gas distributor blockage, packing fouling, and sensor drift. The edge weights are determined by conditional probability statistics from the historical fault case library. The algorithm takes the current combination of parameters that exceed the limit as input and outputs the posterior probability of each fault node. When the posterior probability of a filter media blockage node is greater than 80%, it is determined to be the main cause, and the degradation rule model corresponding to the filter media is called to predict the remaining life.

5. The method for multi-stage sulfuric acid filtration and gas stripping desulfurization purification of storage batteries according to claim 4, characterized in that, The multi-objective optimization solution aims to minimize energy consumption, maximize desulfurization efficiency, and extend filter material life. The constraints include the maximum allowable pressure difference, the minimum circulation flow rate, and the safe operating temperature range. A non-dominated sorting genetic algorithm was used to iteratively optimize the decision variables, which included the coarse filter backwashing cycle, the fine filter bypass valve opening, the ultrafiltration crossflow velocity, the airlift carrier gas flow rate, and the circulation pump frequency. The optimization results generate a set of Pareto optimal solutions, from which the solution with the highest comprehensive score is selected as the coordinated control instruction set.

6. The method for multi-stage sulfuric acid filtration and gas stripping desulfurization purification of storage batteries according to claim 5, characterized in that, The actuators of the physical purification system include: The pneumatic backwash valve installed at the bottom of the coarse filter unit is controlled by backwash cycle commands issued by the digital twin. The opening degree of the electric regulating valve installed in the bypass pipeline of the fine filtration unit is set by the bypass valve opening command. The frequency converter installed in the ultrafiltration circulation loop has its output frequency calculated from the cross-flow velocity command. The speed of the Roots blower installed at the bottom of the air lift tower is controlled by the carrier gas flow command after conversion through a flow-speed mapping table. The motion feedback signals of all actuators are transmitted back to the digital twin in real time to correct the execution deviations of the simulation model.

7. The method for multi-stage sulfuric acid filtration and gas stripping desulfurization purification of storage batteries according to claim 6, characterized in that, The mathematical expression for the Darcy-Fochheimer flow equation is: ； For pressure, For pressure gradient, For dynamic viscosity, For filter media permeability, For flow velocity vectors, For fluid density, This is the inertial drag coefficient.

Images