A method and system for dynamically balancing control of ethylene glycol deep regeneration

Abstract

The application provides a kind of ethylene glycol deep regeneration dynamic balance control method and system, the system includes: labyrinth pretreatment unit, micro-interface atomization unit, entropy drive rectification unit and control module;The labyrinth pretreatment unit is used to carry out adsorption to metal cation and mechanical impurities in ethylene glycol rich solution by filter material, and the adsorption efficiency of the filter material is adjusted;The micro-interface atomization unit is used to atomize the ethylene glycol rich solution into ultrafine droplets after processing, and the atomization efficiency of the ultrafine droplets is adjusted;The entropy drive rectification unit is used to optimize the vacuum gradient and reflux ratio of each stage entropy drive gradient vacuum rectification column in real time, so that the total entropy production rate of the system tends to a minimum value;The control module is used to adjust the operating parameters of the system in real time, and to determine the thickness change rate of the equipment surface fouling in real time to monitor the equipment wall fouling state;Through the above technical scheme, the efficient and low-consumption deep regeneration of ethylene glycol can be realized.

Description

Technical Field

[0001] This application relates to the field of chemical process intensification technology, and in particular to a dynamic equilibrium control method and system for deep regeneration of ethylene glycol. Background Technology

[0002] As natural gas extraction extends into deep-sea and complex geological environments, ethylene glycol (MEG) is widely used as a major hydrate inhibitor. However, existing MEG regeneration units (MRUs) face extremely severe thermodynamic-mass transfer-reaction coupling challenges, with the primary issue being multiphase complex scaling: the presence of divalent metal cations (such as...) in the ethylene glycol-rich solution... , During the high-temperature regeneration process, mechanical impurities can easily form stubborn scale on the equipment surface, leading to pipeline blockage and seriously affecting the long-term operation of the system.

[0003] Secondly, there are issues of high energy consumption and thermal degradation: Traditional MRU processes rely on boiling evaporation to regenerate and separate ethylene glycol-rich solutions. This process requires a continuous input of large amounts of heat energy to overcome the enormous latent heat of vaporization required for gas-liquid phase transition, resulting in persistently high overall energy consumption. At the same time, the microscopic heat distribution at the gas-liquid interface is difficult to control precisely, and local overheating can easily lead to the breakage of ethylene glycol molecular chains and changes in chemical properties. This not only increases the loss rate of the inhibitor's effective components and raises the cost of reagent replenishment, but also generates byproducts such as aldehydes and carboxylic acids. These byproducts further react with impurities in the system, exacerbating the risk of equipment corrosion and scaling, forming a vicious cycle of "increased energy consumption - accelerated degradation - system deterioration," which seriously restricts the operational stability and economy of MRU.

[0004] Finally, there are challenges related to control lag and nonlinearity: the MRU system is a typical multivariable strongly coupled nonlinear system (temperature, pressure, vacuum, concentration). Traditional PID control or a single neural network model is difficult to adapt to the drastic fluctuations of the offshore platform and is prone to getting trapped in local optima, resulting in fluctuations in separation efficiency.

[0005] It is evident that the relevant technologies lack a deep mapping mechanism between physical fields (fluid viscoelasticity, thermodynamic entropy change) and control algorithm parameters, making it impossible to achieve adaptive dynamic equilibrium of the MRU system. Summary of the Invention

[0006] This application provides a dynamic balance control method and system for deep regeneration of ethylene glycol, which solves the problems of severe scaling and clogging, high energy consumption and control lag in the prior art, and realizes efficient and low-consumption deep regeneration of ethylene glycol.

[0007] To achieve the above objectives, the technical solution of this application embodiment is as follows: In a first aspect, embodiments of this application provide a dynamic equilibrium control system for deep regeneration of ethylene glycol. The system includes: a labyrinth pretreatment unit, a micro-interface atomization unit, an entropy-driven distillation unit, and a control module. The labyrinth pretreatment unit is connected to the entropy-driven distillation unit through the micro-interface atomization unit. The control module is connected to the labyrinth pretreatment unit, the micro-interface atomization unit, and the entropy-driven distillation unit, respectively. The labyrinth-type pretreatment unit is used to adsorb metal cations and mechanical impurities in ethylene glycol-rich solution through a labyrinth filter media modified with composite molecular sieves. Based on a pre-established bifunctional fractal topological adsorption kinetic model, the effective adsorption capacity of the filter media is adjusted in real time. In order to adjust the adsorption efficiency of the filter media; The micro-interface atomization unit is used to atomize the pretreated ethylene glycol-rich liquid into ultra-micro droplets with an average diameter smaller than a preset diameter threshold through a pneumatic ultra-micro nozzle, and to adjust the atomization efficiency of the ultra-micro droplets in real time using a pre-established transcritical micro-interface aero-thermodynamic Weber-O'Nezog coupled model. This improves the atomization efficiency. Always greater than or equal to the preset efficiency threshold; The entropy-driven distillation unit is used to further separate and purify the ultrafine droplets through a multi-stage entropy-driven gradient vacuum distillation column. Based on the non-equilibrium thermodynamic entropy production minimization transport equation, it optimizes the vacuum gradient and reflux ratio of each stage of the entropy-driven gradient vacuum distillation column in real time to maximize the overall entropy yield of the system. It tends to a minimum value; The control module is used to implement the objective function of the supervisoelastic wavelet kernel stochastic dynamic equilibrium control method. By combining physical field constraints and regularization terms, the system's operating parameters are adjusted in real time to cope with fluctuations in the operating conditions of offshore platforms; and the rate of change of scale thickness on the equipment surface is determined in real time through an anisotropic scale growth tensor observer. To monitor the scaling condition on the equipment walls.

[0008] In one possible implementation, the pre-established bifunctional fractal topological adsorption kinetic model is expressed as: ; in, for t The effective adsorption amount at any given time, the integral term achieves the effect for different pore sizes r The summation calculation of adsorption contributions, molecular part It belongs to the Langmuir type adsorption kinetics term. It is the Langmuir constant. for tThe concentration of matter at time t, the denominator The influence of fractal pore structure on adsorption sites was then incorporated. For the fractal dimension of the filter media pores, the exponential term is... A quantitative description of the diffusion process within the pores is provided. Pore ​​diffusion coefficient, It is the tortuosity factor.

[0009] In one possible implementation, the pre-established transcritical micro-interface aero-thermodynamic Weber-Ounezoglu coupling model is expressed as: ; in, For atomization efficiency, the integral term is the product of the atomization interface. The summation of contributions from each region is calculated. Characterizing the jet kinetic energy properties of the mixed fluid, For the density of the mixed fluid, For jet velocity, Nozzle diameter, The surface tension of a fluid under specific pressure and temperature conditions; For the viscosity of the mixed fluid, This is a composite term of fluid properties and nozzle parameters; For fluid enthalpy of evaporation, This refers to the mole fraction of carbon dioxide. The enthalpy of dissolution of carbon dioxide. R The gas constant is T It is the thermodynamic temperature.

[0010] In one possible implementation, the non-equilibrium thermodynamic entropy production minimization transport equation is expressed as: ; Among them, entropy production rate The smaller the value, the lower the energy loss and irreversibility of the system. The heat flux density vector, The gradient is the reciprocal of the given value. For the first k Mass flux vector of each component For the first k The gradient of the chemical potential of a component as a function of temperature; For viscous stress tensor, Let T be the velocity gradient tensor, and let its dot product represent the viscous dissipation function, where T is the thermodynamic temperature. This represents the area ratio of scale buildup on the wall surface. The enthalpy change of the scaling reaction, This refers to the wall temperature.

[0011] In one possible implementation, the objective function is: ; in, To use network weight W Bias b The objective function for optimizing variables; For the first k The predicted value for each sample, It is the output layer mapping function. It is the hidden layer activated by the function The output after integration; physical constraints are introduced through integration. For viscoelastic constraints, H It is a generalized function related to physical processes. Let the gradient of the weights be the generalized function. These are the weighting coefficients of the constraint terms; It is the time decay factor. Through chaotic terms Enhance the exploratory nature of the model. For particle velocity, It is the regularization weight. It is a small constant to prevent overfitting.

[0012] In one possible implementation, the rate of change of scale thickness on the device surface is: ; Among them, the rate of change of scale thickness This indicates the change in the thickness of the scale layer on the equipment surface per unit time. A positive value represents scale growth, and a negative value represents scale peeling. The nucleation rate constant represents the efficiency of scale nucleation. The degree of supersaturation of the solution. As a supersaturation driving term, n The reaction order is [number]. The activation energy for the scaling reaction. R The gas constant is T Thermodynamic temperature The Arrhenius factor characterizes the effect of temperature on the scaling reaction rate. This represents the shear stress exerted by the fluid on the wall. The viscosity of the scale layer, This refers to the adhesion factor between the scale layer and the equipment wall.

[0013] In one possible implementation, the control module is further configured to dynamically update the inertia weights using a weight evolution formula based on Sigmoidal cosine modulation. This ensures that the algorithm converges quickly to the global optimal solution and controls the system lag time to be less than a preset time threshold.

[0014] In one possible implementation, the weight evolution formula is: ; in, Indicates the first u The particle's inertial weight at the next iteration; the fundamental weight term. This represents the minimum value of the inertia weight; the weight magnitude term. The maximum change in inertia weight determines the adjustable range of the weight. This represents the maximum value of the inertia weight; u This represents the current iteration number. The average number of iterations. The standard deviation of the number of iterations; This represents the maximum number of iterations.

[0015] Secondly, embodiments of this application provide a method for dynamic equilibrium control of deep regeneration of ethylene glycol, the method being applied to a dynamic equilibrium control system for deep regeneration of ethylene glycol according to the first aspect, the method comprising: Metal cations and mechanical impurities in ethylene glycol-rich solutions are adsorbed using a labyrinth filter media modified with composite molecular sieves. Based on a pre-established bifunctional fractal topological adsorption kinetic model, the effective adsorption capacity of the filter media is adjusted in real time. In order to adjust the adsorption efficiency of the filter media; The pretreated ethylene glycol-rich solution is atomized into ultrafine droplets with an average diameter smaller than a preset diameter threshold using a pneumatic ultrafine nozzle. The atomization efficiency of the ultrafine droplets is adjusted in real time using a pre-established transcritical micro-interface aero-thermodynamic Weber-O'Nezog coupled model. This improves the atomization efficiency. Always greater than or equal to the preset efficiency threshold; The ultrafine droplets are further separated and purified using a multi-stage entropy-driven gradient vacuum distillation column. Based on the non-equilibrium thermodynamic entropy production minimization transport equation, the vacuum gradient and reflux ratio of each stage of the entropy-driven gradient vacuum distillation column are optimized in real time to maximize the overall entropy yield of the system. It tends to a minimum value; The objective function of the supervisoelastic wavelet kernel stochastic dynamic equilibrium control method is... By combining physical field constraints and regularization terms, the system's operating parameters are adjusted in real time to cope with fluctuations in the operating conditions of offshore platforms; and the rate of change of scale thickness on the equipment surface is determined in real time through an anisotropic scale growth tensor observer. To monitor the scaling condition on the equipment walls.

[0016] In one possible implementation, the method further includes: The inertia weights are dynamically updated using a weight evolution formula based on Sigmoidal cosine modulation. This ensures that the algorithm converges quickly to the global optimal solution and controls the system lag time to be less than a preset time threshold.

[0017] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: (1) It improves the regeneration efficiency and quality of ethylene glycol, making the efficiency much higher than that of traditional regeneration methods, and meets the requirements for hydrate inhibition in deep-water natural gas mining; (2) It reduces the overall energy consumption of the system and the thermal degradation loss rate of ethylene glycol, thereby reducing the amount of ethylene glycol replenishment and the cost of reagents; (3) Improved the scale control effect, reduced the number of shutdowns for scale removal and the resulting economic losses, thereby increasing the natural gas extraction cycle of the platform and enabling the platform to extract natural gas stably and continuously; (4) It can effectively cope with the drastic fluctuations in operating conditions such as temperature and pressure, and the system operating parameters can be adjusted and responded quickly to meet the regeneration needs in deep water complex geological environments. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments of this application or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A block diagram of a dynamic equilibrium control system for deep regeneration of ethylene glycol provided in an embodiment of this application; Figure 2 A block diagram of another ethylene glycol deep regeneration dynamic balance control system provided in the embodiments of this application; Figure 3 A flowchart of a dynamic equilibrium control method for deep regeneration of ethylene glycol provided in this application embodiment. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0021] In the description of the embodiments of this application, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the embodiments of this application and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. The terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; a mechanical connection or an electrical connection; a direct connection or an indirect connection through an intermediate medium; or a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of this application according to the specific circumstances.

[0022] Figure 1 This is a block diagram of a dynamic equilibrium control system 100 for deep regeneration of ethylene glycol provided in an embodiment of this application. Figure 1 As shown, the system 100 may include: a labyrinth pretreatment unit 110, a micro-interface atomization unit 120, an entropy-driven distillation unit 130, and a control module 140; the labyrinth pretreatment unit 110 is connected to the entropy-driven distillation unit 130 through the micro-interface atomization unit 120; the control module 140 is connected to the labyrinth pretreatment unit 110, the micro-interface atomization unit 120, and the entropy-driven distillation unit 130 respectively.

[0023] The labyrinth pretreatment unit 110 is used to adsorb metal cations and mechanical impurities in ethylene glycol-rich solution through a labyrinth filter media modified with composite molecular sieves. Based on a pre-established bifunctional fractal topological adsorption kinetic model, the effective adsorption capacity of the filter media is adjusted in real time. This is to adjust the adsorption efficiency of the filter media.

[0024] The micro-interface atomization unit 120 is used to atomize the pretreated ethylene glycol-rich liquid into ultra-micro droplets with an average diameter smaller than a preset diameter threshold through a pneumatic ultra-micro nozzle, and to adjust the atomization efficiency of the ultra-micro droplets in real time through a pre-established transcritical micro-interface aero-thermodynamic Weber-Onnezog coupling model. This improves the atomization efficiency. It is always greater than or equal to the preset efficiency threshold.

[0025] For example, the preset diameter threshold can be between 3μm and 8μm, such as 3μm, 5μm or 8μm; the preset efficiency threshold can be between 90% and 99%, such as 90%, 92%, 95% or 995, etc., and there is no limitation here.

[0026] The entropy-driven distillation unit 130 is used to further separate and purify the ultrafine droplets through a multi-stage entropy-driven gradient vacuum distillation column. Based on the non-equilibrium thermodynamic entropy production minimization transport equation, it optimizes the vacuum gradient and reflux ratio of each stage of the entropy-driven gradient vacuum distillation column in real time, maximizing the overall entropy yield of the system. It tends to a minimum value.

[0027] The control module 140 is used to implement the objective function of the hyperviscoelastic wavelet kernel stochastic dynamic equilibrium control method. By combining physical field constraints and regularization terms, the system's operating parameters are adjusted in real time to cope with fluctuations in the operating conditions of offshore platforms; and the rate of change of scale thickness on the equipment surface is determined in real time through an anisotropic scale growth tensor observer. To monitor the scaling condition on the equipment walls.

[0028] For example, the rate of change of surface fouling thickness can characterize the amount of change in the thickness of the fouling layer on the equipment surface per unit time; when the rate of change of surface fouling thickness is greater than or equal to a preset rate of change threshold, the system can automatically adjust the fluid shear stress. With operating temperature T This system inhibits scale growth and outputs a warning message to remind the user to pay attention to the scaling situation. The preset change rate threshold can be between 0.005-0.02 mm / h, for example, 0.005 mm / h, 0.01 mm / h, 0.02 mm / h, etc., without limitation here. This allows for dynamic monitoring of scale growth even when the scale thickness change rate is too large, promptly reminding the user to perform appropriate preventative measures, thereby reducing downtime for scaling removal and minimizing economic losses.

[0029] The above technical solutions can improve the regeneration efficiency and quality of ethylene glycol, making the efficiency far higher than that of traditional regeneration methods, thus meeting the requirements for hydrate suppression in deep-water natural gas extraction. Furthermore, they can improve scale control, reduce downtime for scale removal and associated economic losses, thereby extending the platform's natural gas extraction cycle and enabling stable and continuous natural gas extraction. Additionally, they can effectively cope with drastic fluctuations in operating conditions such as temperature and pressure, with rapid response to system parameter adjustments, adapting to the regeneration needs of complex geological environments in deep water.

[0030] In one possible implementation, the pre-established bifunctional fractal topological adsorption kinetic model is expressed as follows: ;in, for tThe effective adsorption amount at any given time, the integral term achieves the effect for different pore sizes r The summation calculation of adsorption contributions, molecular part It belongs to the Langmuir type adsorption kinetics term. It is the Langmuir constant. for t The concentration of matter at time t, the denominator The influence of fractal pore structure on adsorption sites was then incorporated. For the fractal dimension of the filter media pores, the exponential term is... A quantitative description of the diffusion process within the pores is provided. Pore ​​diffusion coefficient, It is the tortuosity factor.

[0031] In one possible implementation, the pre-established transcritical micro-interface aero-thermodynamic Weber-O'Nezog coupling model is expressed as: ;in, For atomization efficiency, the integral term is the product of the atomization interface. The summation of contributions from each region is calculated. Characterizing the jet kinetic energy properties of the mixed fluid, For the density of the mixed fluid, For jet velocity, Nozzle diameter, For the surface tension of a fluid under specific pressure and temperature conditions, The overall effect of the competition between jet kinetic energy and surface tension on atomization is demonstrated. For the viscosity of the mixed fluid, It is a composite term of fluid properties and nozzle parameters, and the effect of fluid viscosity on atomization efficiency is quantified by exponential power. For fluid enthalpy of evaporation, This refers to the mole fraction of carbon dioxide. The enthalpy of dissolution of carbon dioxide. R The gas constant is T Thermodynamic temperature The thermodynamic effects of phase change evaporation and gas dissolution processes are comprehensively reflected in their constraint on atomization.

[0032] In one possible implementation, the transport equation for minimizing nonequilibrium thermodynamic entropy production is expressed as: The equation represents the optimization objective as minimizing the entropy yield of the system. Entropy production rate The smaller the value, the lower the energy loss and irreversibility of the system. The heat flux density vector, The gradient is the reciprocal. Entropy is a characterization of the irreversible heat transfer process caused by the temperature gradient within the system. For the first k Mass flux vector of each component For the first k The gradient of the chemical potential of a component as a function of temperature; For viscous stress tensor, Let T be the velocity gradient tensor, and let its dot product represent the viscous dissipation function, where T is the thermodynamic temperature. This represents the area ratio of scale buildup on the wall surface. The enthalpy change of the scaling reaction, This refers to the wall temperature.

[0033] In one possible implementation, the objective function is: ; in, To use network weight W Bias b The objective function (loss function) for optimizing variables; For the first k The predicted value for each sample, It is the output layer mapping function. It is the hidden layer activated by the function The output after integration; physical constraints are introduced through integration. For viscoelastic constraints, H It is a generalized function related to physical processes. Let the gradient of the weights be the generalized function. These are the weight coefficients of the constraint terms, ensuring that the model learning process conforms to physical laws; It is the time decay factor. Through chaotic terms Enhance the exploratory nature of the model. For particle velocity, It is the regularization weight. It is a small constant that prevents overfitting and is used to balance the complexity and generalization ability of the model.

[0034] In one possible implementation, the rate of change of scale thickness on the device surface is: Among them, the rate of change of scale thickness This indicates the change in the thickness of the scale layer on the equipment surface per unit time. A positive value represents scale growth, and a negative value represents scale peeling. The nucleation rate constant represents the efficiency of scale nucleation. The degree of supersaturation of the solution. As a supersaturation driving term, n The higher the supersaturation, the faster the scaling rate. The activation energy for the scaling reaction. RThe gas constant is T Thermodynamic temperature The Arrhenius factor characterizes the effect of temperature on the scaling reaction rate. This refers to the shear stress exerted by the fluid on the wall surface. The greater the shear stress, the stronger the scouring effect on the scale layer. The viscosity of the scale layer characterizes the density of the scale layer. This refers to the adhesion factor between the scale layer and the equipment wall. The smaller the adhesion factor, the easier it is for the scale to be peeled off.

[0035] In one possible implementation, the control module is further configured to dynamically update the inertia weights using a weight evolution formula based on Sigmoidal cosine modulation. This ensures that the algorithm converges quickly to the global optimum, controlling the system lag time to be less than a preset time threshold. For example, this preset time threshold can be 0.2-0.5s, such as 0.2s, 0.3s, or 0.5s, etc., and is not limited here.

[0036] In one possible implementation, the weight evolution formula is: ; in, Indicates the first u The inertial weight of the particle in the next iteration determines the degree to which the particle inherits the historical velocity; the basic weight term. The minimum value of the inertia weight ensures that the algorithm still possesses a certain degree of global search capability in the later stages of iteration; the weight magnitude term The maximum change in inertia weight determines the adjustable range of the weight. This represents the maximum value of the inertia weight; u This represents the current iteration number. The average number of iterations. The standard deviation of the number of iterations. For the Sigmoid function; The maximum number of iterations, Using a cosine function, the weights exhibit periodic small fluctuations during the iteration process, enhancing the algorithm's ability to escape local optima.

[0037] Figure 2 A block diagram of another ethylene glycol deep regeneration dynamic equilibrium control system provided in the embodiments of this application; the following is an example. Figure 2 This example illustrates the workflow of this application. This embodiment can be applied to an offshore platform in a deep-water natural gas field in the South China Sea. The platform's natural gas extraction depth reaches 1500 meters, with complex geological conditions, and the produced natural gas is accompanied by high-salinity formation water (containing...). , The total concentration of divalent metal cations is approximately 800 mg / L, and the mining conditions fluctuate greatly (temperature range 40-120℃, pressure fluctuation range). The platform originally used a traditional ethylene glycol regeneration unit (MRU), which frequently experienced problems such as pipeline scaling and blockage (requiring shutdown for scaling removal on average every 3 months), ethylene glycol thermal degradation loss rate as high as 15%, and regeneration energy consumption accounting for 28% of the platform's total energy consumption, seriously affecting the continuity and economics of natural gas extraction. To solve the above problems, the platform uses the technical solution of this application to achieve deep regeneration and efficient recycling of ethylene glycol.

[0038] I. Configuration of core system components.

[0039] 1. Labyrinth-style pretreatment unit: Employs composite molecular sieve-modified labyrinth filter media designed based on a bifunctional fractal topological adsorption kinetics model, with fractal dimensions of the filter media pores. The Langmuir constant should be controlled between 2.3 and 2.6. Optimized to 0.85 L / mg, pore diffusion coefficient ≥1.2×10 -9 m² / s, tortuosity factor ≤1.8.

[0040] 2. Micro-interface atomization unit: Equipped with a pneumatic ultra-micro nozzle, nozzle diameter... =0.8mm, density of mixed fluid =1030kg / m³, jet velocity The adjustment range is 15-25 m / s, the system operating pressure is controlled at 2.5-3.0 MPa, and the temperature is maintained at 80-90℃.

[0041] 3. Entropy-driven distillation unit: Utilizing a three-stage vacuum distillation column structure, the vacuum gradients within each stage are designed to be 0.03 MPa, 0.05 MPa, and 0.08 MPa, respectively. The reflux ratio can be dynamically adjusted between 1.2 and 2.0. The column wall material is made of corrosion-resistant alloy, and the scaling area ratio is [not specified]. The preset threshold is ≤0.05.

[0042] 4. Control Module: Utilizing a supervisoelastic wavelet kernel stochastic dynamic equilibrium control method, the activation function of the hidden layer in the neural network... The Sigmoid function is selected, and the weighting coefficients of the physics constraint terms are... =0.02, regularization weight =0.005, Inertia weight range =0.4、 =0.9, maximum number of iterations =1000.

[0043] 5. Feed conditions: The feed flow rate of the ethylene glycol-rich solution is 5 m³ / h, wherein the ethylene glycol mass fraction is 70%, and the salt content (as expressed in Ca²⁺) is... + Mg² + (Calculated) 800mg / L, mechanical impurity content ≤50mg / L, feed temperature 60℃, feed pressure 1.2MPa.

[0044] II. Work Process.

[0045] 1. Pretreatment stage: After the labyrinth-type pretreatment unit is started, the effective adsorption capacity is determined based on the bifunctional fractal topological adsorption kinetic model. By integrating different apertures r The cumulative effect of adsorption allows for precise control of the filter media's adsorption capacity in ethylene glycol-rich solutions. , The adsorption efficiency for mechanical impurities was also measured. After one hour of operation, the filter media's adsorption capacity for divalent metal cations was found to be 120 mg / g, indicating an effective adsorption capacity. The concentration was stabilized at 105 mg / g, and the salt content in the pretreated ethylene glycol-rich solution was reduced to below 50 mg / L. The mechanical impurity removal rate was ≥98%, laying a clean feed foundation for the subsequent regeneration process.

[0046] 2. Atomization Stage: The pretreated ethylene glycol-rich solution enters the micro-interface atomization unit, and the atomization efficiency is calculated based on the transcritical micro-interface aerodynamic-thermodynamic Weber-Onnezog coupling model. By balancing the competition between jet kinetic energy and surface tension, quantifying the regulation of fluid viscosity, and considering the thermodynamic effects of phase change evaporation and gas dissolution, the atomization efficiency is stabilized at over 92%, forming ultra-micro droplets with an average particle size of ≤5μm.

[0047] 3. Entropy-driven distillation stage: Ultrafine droplets enter the entropy-driven gradient vacuum distillation column. Based on the non-equilibrium thermodynamic entropy production minimization transport equation, the vacuum gradient and reflux ratio of each stage of the column are optimized in real time to minimize the system entropy production rate. The heat transfer entropy production, mass transfer entropy production, and viscous dissipation entropy production are all controlled at extremely low levels, and the system's thermodynamic irreversible losses are reduced by 40%.

[0048] 4. Dynamic Control and Scale Early Warning Stage: The control module uses the objective function of the ultraviscous wavelet kernel stochastic dynamic equilibrium control method. Combined with physical fields (fluid viscoelasticity) Constraints and regularization terms are used to adjust system operating parameters in real time to cope with fluctuations in the operating conditions of offshore platforms; the inertial weights are dynamically updated through a Sigmoidal cosine modulation weight evolution formula. This ensures that the algorithm converges quickly to the global optimum, reducing the control lag time to 0.3 seconds.

[0049] 5. Using an anisotropic scale growth tensor observer, the scale condition on the equipment wall is monitored in real time by the rate of change of scale thickness. When the scale thickness growth rate exceeds 0.01 mm / h, the system automatically adjusts the fluid shear stress. With operating temperature T It inhibits scale growth and enables early warning and dynamic control of scale formation.

[0050] III. Application Results

[0051] 1. Regeneration efficiency and quality: The ethylene glycol regeneration purity reaches over 99.5%, and the regeneration recovery rate is increased to 98%, which is far higher than the 92% recovery rate of traditional MRU, meeting the requirements for hydrate suppression in deep-water natural gas extraction.

[0052] 2. Energy consumption and losses: The overall energy consumption of the system is reduced to 850 kWh / m³, which is 35% lower than that of the traditional process; the thermal degradation loss rate of ethylene glycol is reduced to below 3%, reducing the amount of ethylene glycol replenishment by about 12 tons per year and reducing the cost of reagents by about 800,000 yuan.

[0053] 3. Operational stability: The scaling control effect is significant, the continuous operation cycle of the equipment is extended to 18 months, and there is no need to stop the machine for scaling. The continuity of natural gas extraction on the platform is improved, and the economic loss caused by downtime for scaling is reduced by about RMB 2 million per year.

[0054] 4. Adaptability: The control system can effectively cope with drastic fluctuations in operating conditions such as temperature and pressure. The system operating parameters can be adjusted and responded to quickly, adapting to the regeneration needs in complex geological environments in deep water.

[0055] Figure 3 This is a flowchart illustrating a dynamic equilibrium control method for deep regeneration of ethylene glycol, provided as an embodiment of this application. The method is applied to a dynamic equilibrium control system 100 for deep regeneration of ethylene glycol and may include the following steps.

[0056] S301. The composite molecular sieve-modified labyrinth filter media in the labyrinth filter adsorbs metal cations and mechanical impurities in the ethylene glycol-rich solution, and the effective adsorption capacity of the filter media is adjusted in real time based on a pre-established bifunctional fractal topological adsorption kinetic model. This is to adjust the adsorption efficiency of the filter media.

[0057] S302. The pretreated ethylene glycol-rich liquid is atomized into ultra-micro droplets with an average diameter smaller than a preset diameter threshold using a pneumatic ultra-micro nozzle. The atomization efficiency of the ultra-micro droplets is adjusted in real time using a pre-established transcritical micro-interface aero-thermodynamic Weber-Onnezog coupling model. This improves the atomization efficiency. It is always greater than or equal to the preset efficiency threshold.

[0058] S303. The ultrafine droplets are further separated and purified using a multi-stage entropy-driven gradient vacuum distillation column. Based on the non-equilibrium thermodynamic entropy production minimization transport equation, the vacuum gradient and reflux ratio of each stage of the entropy-driven gradient vacuum distillation column are optimized in real time to maximize the overall entropy production rate of the system. It tends to a minimum value.

[0059] S304, Objective function of the supervisoelastic wavelet kernel stochastic dynamic equilibrium control method By combining physical field constraints and regularization terms, the system's operating parameters are adjusted in real time to cope with fluctuations in the operating conditions of offshore platforms; and the rate of change of scale thickness on the equipment surface is determined in real time through an anisotropic scale growth tensor observer. To monitor the scaling condition on the equipment walls.

[0060] The above technical solutions can improve the regeneration efficiency and quality of ethylene glycol, making the efficiency far higher than that of traditional regeneration methods, thus meeting the requirements for hydrate suppression in deep-water natural gas extraction. Furthermore, they can improve scale control, reduce downtime for scale removal and associated economic losses, thereby extending the platform's natural gas extraction cycle and enabling stable and continuous natural gas extraction. Additionally, they can effectively cope with drastic fluctuations in operating conditions such as temperature and pressure, with rapid response to system parameter adjustments, adapting to the regeneration needs of complex geological environments in deep water.

[0061] The various embodiments in this specification are described in a progressive manner. For the same or similar parts between the various embodiments, please refer to each other. Each embodiment focuses on describing the differences from other embodiments.

[0062] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of this application.

Claims

1. A dynamic balance control system for deep regeneration of ethylene glycol, characterized in that, The system includes: a labyrinth-type pretreatment unit, a micro-interface atomization unit, an entropy-driven distillation unit, and a control module; the labyrinth-type pretreatment unit is connected to the entropy-driven distillation unit through the micro-interface atomization unit; the control module is connected to the labyrinth-type pretreatment unit, the micro-interface atomization unit, and the entropy-driven distillation unit respectively. The labyrinth-type pretreatment unit is used to adsorb metal cations and mechanical impurities in ethylene glycol-rich solution through a labyrinth filter media modified with composite molecular sieves. Based on a pre-established bifunctional fractal topological adsorption kinetic model, the effective adsorption capacity of the filter media is adjusted in real time. In order to adjust the adsorption efficiency of the filter media; The micro-interface atomization unit is used to atomize the pretreated ethylene glycol-rich liquid into ultra-micro droplets with an average diameter smaller than a preset diameter threshold through a pneumatic ultra-micro nozzle, and to adjust the atomization efficiency of the ultra-micro droplets in real time using a pre-established transcritical micro-interface aero-thermodynamic Weber-O'Nezog coupled model. This improves the atomization efficiency. Always greater than or equal to the preset efficiency threshold; The entropy-driven distillation unit is used to further separate and purify the ultrafine droplets through a multi-stage entropy-driven gradient vacuum distillation column. Based on the non-equilibrium thermodynamic entropy production minimization transport equation, it optimizes the vacuum gradient and reflux ratio of each stage of the entropy-driven gradient vacuum distillation column in real time to maximize the overall entropy yield of the system. It tends to a minimum value; The control module is used to implement the objective function of the supervisoelastic wavelet kernel stochastic dynamic equilibrium control method. By combining physical field constraints and regularization terms, the system's operating parameters are adjusted in real time to cope with fluctuations in the operating conditions of offshore platforms; and the rate of change of scale thickness on the equipment surface is determined in real time through an anisotropic scale growth tensor observer. To monitor the scaling condition on the equipment walls.

2. The system according to claim 1, characterized in that, The pre-established bifunctional fractal topological adsorption kinetic model is expressed as follows: ; in, for t The effective adsorption amount at any given time, the integral term achieves the effect for different pore sizes r The summation calculation of adsorption contributions, molecular part It belongs to the Langmuir type adsorption kinetics term. It is the Langmuir constant. for t The concentration of matter at time t, the denominator The influence of fractal pore structure on adsorption sites was then incorporated. For the fractal dimension of the filter media pores, the exponential term is... A quantitative description of the diffusion process within the pores is provided. Pore ​​diffusion coefficient, It is the tortuosity factor.

3. The system according to claim 1, characterized in that, The pre-established transcritical micro-interface aero-thermodynamic Weber-O'Nezog coupling model is expressed as follows: ； in, For atomization efficiency, the integral term is the product of the atomization interface. The summation of contributions from each region is calculated. Characterizing the jet kinetic energy properties of the mixed fluid, For the density of the mixed fluid, For jet velocity, Nozzle diameter, The surface tension of a fluid under specific pressure and temperature conditions; For the viscosity of the mixed fluid, This is a composite term of fluid properties and nozzle parameters; For fluid enthalpy of evaporation, This refers to the mole fraction of carbon dioxide. The enthalpy of dissolution of carbon dioxide. R The gas constant is T It is the thermodynamic temperature.

4. The system according to claim 1, characterized in that, The transport equation for minimizing nonequilibrium thermodynamic entropy production is expressed as: ; Among them, entropy production rate The smaller the value, the lower the energy loss and irreversibility of the system. The heat flux density vector, The gradient is the reciprocal of the given value. For the first k Mass flux vector of each component For the first k The gradient of the chemical potential of a component as a function of temperature; For viscous stress tensor, Let T be the velocity gradient tensor, and let its dot product represent the viscous dissipation function, where T is the thermodynamic temperature. This represents the area ratio of scale buildup on the wall surface. The enthalpy change of the scaling reaction, This refers to the wall temperature.

5. The system according to claim 1, characterized in that, The objective function is: ； in, To use network weight W Bias b The objective function for optimizing variables; For the first k The predicted value for each sample, It is the output layer mapping function. It is the hidden layer activated by the function The output after integration; physical constraints are introduced through integration. For viscoelastic constraints, H It is a generalized function related to physical processes. Let the gradient of the weights be the generalized function. These are the weighting coefficients of the constraint terms; It is the time decay factor. Through chaotic terms Enhance the exploratory nature of the model. For particle velocity, It is the regularization weight. It is a small constant to prevent overfitting.

6. The system according to claim 1, characterized in that, The rate of change of scale thickness on the surface of the equipment is: ； Among them, the rate of change of scale thickness This indicates the change in the thickness of the scale layer on the equipment surface per unit time. A positive value represents scale growth, and a negative value represents scale peeling. The nucleation rate constant represents the efficiency of scale nucleation. The degree of supersaturation of the solution. As a supersaturation driving term, n The reaction order is [number]. The activation energy for the scaling reaction. R The gas constant is T Thermodynamic temperature The Arrhenius factor characterizes the effect of temperature on the scaling reaction rate. This represents the shear stress exerted by the fluid on the wall. The viscosity of the scale layer, This refers to the adhesion factor between the scale layer and the equipment wall.

7. The system according to claim 1, characterized in that, The control module is also used to dynamically update the inertia weights using the weight evolution formula of Sigmoidal cosine modulation. This ensures that the algorithm converges quickly to the global optimal solution and controls the system lag time to be less than a preset time threshold.

8. The system according to claim 7, characterized in that, The weight evolution formula is as follows: ； in, Indicates the first u The particle's inertial weight at the next iteration; the fundamental weight term. This represents the minimum value of the inertia weight; the weight magnitude term. The maximum change in inertia weight determines the adjustable range of the weight. This represents the maximum value of the inertia weight; u This represents the current iteration number. The average number of iterations. The standard deviation of the number of iterations; This represents the maximum number of iterations.

9. A method for dynamic equilibrium control of deep regeneration of ethylene glycol, characterized in that, The method is applied to a dynamic equilibrium control system for deep regeneration of ethylene glycol according to any one of claims 1 to 8, and the method includes: Metal cations and mechanical impurities in ethylene glycol-rich solutions are adsorbed using a labyrinth filter media modified with composite molecular sieves. Based on a pre-established bifunctional fractal topological adsorption kinetic model, the effective adsorption capacity of the filter media is adjusted in real time. In order to adjust the adsorption efficiency of the filter media; The pretreated ethylene glycol-rich solution is atomized into ultrafine droplets with an average diameter smaller than a preset diameter threshold using a pneumatic ultrafine nozzle. The atomization efficiency of the ultrafine droplets is adjusted in real time using a pre-established transcritical micro-interface aero-thermodynamic Weber-O'Nezog coupled model. This improves the atomization efficiency. Always greater than or equal to the preset efficiency threshold; The ultrafine droplets are further separated and purified using a multi-stage entropy-driven gradient vacuum distillation column. Based on the non-equilibrium thermodynamic entropy production minimization transport equation, the vacuum gradient and reflux ratio of each stage of the entropy-driven gradient vacuum distillation column are optimized in real time to maximize the overall entropy yield of the system. It tends to a minimum value; The objective function of the supervisoelastic wavelet kernel stochastic dynamic equilibrium control method is... By combining physical field constraints and regularization terms, the system's operating parameters are adjusted in real time to cope with fluctuations in the operating conditions of offshore platforms; and the rate of change of scale thickness on the equipment surface is determined in real time through an anisotropic scale growth tensor observer. To monitor the scaling condition on the equipment walls.

10. The method according to claim 9, characterized in that, The method further includes: The inertia weights are dynamically updated using a weight evolution formula based on Sigmoidal cosine modulation. This ensures that the algorithm converges quickly to the global optimal solution and controls the system lag time to be less than a preset time threshold.

Images