Fixed bed reaction process coupling regulation method and system based on industrial cloud computing

Abstract

The present application relates to the technical field of industrial control, and especially relates to a fixed bed reaction process coupling regulation method and system based on industrial cloud computing, and the present application proposes the following scheme, the method is solved by multi-field coupling through porous medium, component transfer, heat transfer and reaction kinetics model, obtains bed layer reaction data and identifies the reaction uneven area; according to the effective reaction specific surface area, the local reaction rate and the metal cobalt consumption, the demand of replenishment is judged, and the contact state of the new reaction area and the continuous covering area of the reaction liquid after replenishment is predicted, so as to update the replenishment scheme and generate the flow field regulation scheme, and improve the reaction uniformity.

Description

Technical Field

[0001] This invention relates to the field of industrial control technology, and in particular to a fixed-bed reaction process coupling control method and system based on industrial cloud computing. Background Technology

[0002] Fixed-bed reactors are commonly used in liquid-solid reaction processes such as metal salt preparation due to their relatively simple structure, large solid-liquid contact area, and ease of continuous operation. In the process of preparing cobalt acetate from cobalt metal, the cobalt raw material is usually packed into the reactor in the form of blocks, particles, flakes, or irregular lumps to form a fixed bed. The reaction liquid flows within the bed and reacts with the cobalt metal surface. However, the morphology, particle size, bulk density, and exposed specific surface area of ​​the cobalt raw material vary greatly, which can easily lead to uneven porosity, localized flow deviation, stagnation, insufficient mass transfer, uneven temperature distribution, and localized differences in reaction rates within the bed. Existing fixed-bed operation control relies heavily on empirical judgment based on external macroscopic parameters such as inlet and outlet pressure, temperature, flow rate, or product concentration. Even when numerical simulation methods are used, they usually focus on the fluid distribution analysis under a single flow field or steady-state conditions, making it difficult to simultaneously reflect the flow resistance of the reaction liquid in the cobalt metal bed, the mass transfer process between cobalt metal and the reaction liquid, the heat transfer process of the reaction, and the localized reaction process of cobalt metal to cobalt acetate. Since the location, extent, and causes of uneven reaction regions within the bed are difficult to accurately identify, existing methods struggle to generate targeted control schemes that match the flow field, mass transfer, heat transfer, and local reaction state. This, in turn, affects the utilization rate of metallic cobalt, the stability of cobalt acetate formation, and the reliability of continuous operation of the fixed-bed reaction process.

[0003] For example, patent application CN117672390A discloses a multi-scale simulation method for catalytic reaction processes. It mainly analyzes the internal flow state of the reactor by establishing a fluid simulation model to obtain flow field distribution information. The above-mentioned prior art usually analyzes the overall flow characteristics of the reactor, the distribution of a single physical field, or macroscopic operating parameters. It is difficult to combine the actual filling state of the cobalt raw material in the fixed bed, and comprehensively consider the coupling effect between the bed porosity change, the effective reaction specific surface area distribution, the flow resistance, the mass transfer process, the heat transfer process, and the local reaction kinetics. It cannot accurately characterize the formation cause and spatial distribution of the uneven reaction region inside the cobalt bed.

[0004] To address the above issues, this application presents a method and system for coupled control of fixed-bed reaction processes based on industrial cloud computing. Summary of the Invention

[0005] The technical problem to be solved by this invention is to address the shortcomings of existing technologies by providing a fixed-bed reaction process coupling control method and system based on industrial cloud computing. The method uses a porous medium, component transfer, heat transfer and reaction kinetics model to perform multi-field coupling solution to obtain bed reaction data and identify reaction inhomogeneity regions. Based on the effective reaction specific surface area, local reaction rate and cobalt consumption, the feeding requirement is determined, and the contact state between the newly added reaction area and the continuous coverage area of ​​the reaction liquid after feeding is predicted to update the feeding scheme and generate a flow field control scheme to improve reaction uniformity.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for coupled control of a fixed-bed reaction process based on industrial cloud computing, the method comprising:

[0008] The physical properties and operating parameters of the reaction liquid in the fixed-bed reactor are obtained, and the reaction data in the fixed-bed reactor are obtained by multi-field coupling solution through a preset numerical simulation model. The numerical simulation model includes a porous media model for characterizing the flow resistance of the reaction liquid in the cobalt bed, a component transfer model for characterizing the mass transfer process between cobalt and the reaction liquid, a heat transfer model for characterizing the exothermic or endothermic process of the reaction, and a reaction kinetic model for characterizing the process of cobalt forming cobalt acetate.

[0009] Based on the reaction data, identify regions of uneven reaction within the cobalt metal bed;

[0010] Based on the location and extent of the uneven reaction region and its corresponding flow field parameters, mass transfer parameters, heat transfer parameters, and local reaction rate parameters, a fixed-bed reaction flow field control scheme is generated.

[0011] The numerical simulation model is constructed in the following ways:

[0012] The basic parameters of the fixed-bed reactor used for the reaction of metallic cobalt to prepare cobalt acetate are obtained, including reactor structural parameters and metallic cobalt bed packing parameters;

[0013] Based on the aforementioned cobalt bed loading parameters, the morphological distribution, particle size distribution, bulk density distribution, porosity distribution, and effective reaction specific surface area distribution of the cobalt raw material in the fixed bed are determined.

[0014] A numerical simulation model of the cobalt fixed-bed reaction zone is constructed based on the reactor structural parameters, the porosity distribution, and the effective reaction surface area distribution.

[0015] The method for constructing the numerical simulation model also includes:

[0016] The cobalt fixed-bed reaction zone is divided into multiple computational units;

[0017] For each calculation unit, the permeability parameters and inertial drag parameters of the calculation unit are determined according to the porosity, cobalt particle size, cobalt morphology parameters and packing density of the calculation unit, and the porous medium model is established based on the permeability parameters and inertial drag parameters.

[0018] For each calculation unit, the local mass transfer coefficient between the cobalt metal surface and the reaction liquid is determined based on the effective reaction surface area, reaction liquid flow rate and reactant concentration corresponding to the calculation unit, and the component transfer model is established based on the local mass transfer coefficient.

[0019] For each calculation unit, the local heat source term of the calculation unit is determined according to the effective reaction specific surface area, local reaction rate and reaction heat parameters corresponding to the calculation unit, and the heat transfer model is established based on the local heat source term.

[0020] For each calculation unit, the local reaction rate of cobalt metal to cobalt acetate is determined based on the effective reaction surface area, reactant concentration and temperature corresponding to the calculation unit, and the reaction kinetic model is established based on the local reaction rate.

[0021] The multi-field coupling solution includes:

[0022] The physical properties of the reaction liquid are used as boundary conditions, and the reaction operating parameters are used as initial conditions input into the numerical simulation model.

[0023] Based on the coupling relationship between the porous medium model, the component transport model, the heat transfer model and the reaction kinetics model, the reaction process of reaction liquid flow, reactant mass transfer, heat transfer and the formation of cobalt acetate from metal cobalt is solved synchronously and iteratively in the metal cobalt fixed bed reaction zone.

[0024] During the synchronous iterative solution process, the local mass transfer coefficient and local reaction rate are updated based on the local flow rate, local temperature and local reactant concentration obtained in the current iteration step. The reactant consumption, cobalt acetate production and local heat source terms are corrected based on the updated local reaction rate until the changes in flow rate, temperature, concentration and local reaction rate between adjacent iteration steps meet the preset convergence conditions.

[0025] The bed spatial distribution result obtained when the preset convergence condition is met is used as the reaction data.

[0026] The methods for identifying uneven reaction regions include:

[0027] The reaction data is spatially gridded to obtain the flow rate, pressure, temperature, reactant concentration, cobalt acetate concentration, and local reaction rate values ​​corresponding to each calculation unit.

[0028] Based on the flow rate, reactant concentration, temperature, and local reaction rate values ​​corresponding to each calculation unit, the flow field deviation index, mass transfer deviation index, temperature deviation index, and reaction rate deviation index are calculated respectively.

[0029] The flow field deviation index, the mass transfer deviation index, the temperature deviation index, and the reaction rate deviation index are weighted to obtain the reaction uniformity evaluation value of each calculation unit.

[0030] The reaction uniformity evaluation value is compared with a preset uniformity threshold, and the calculation unit whose reaction uniformity evaluation value is greater than the preset uniformity threshold is identified as an abnormal calculation unit.

[0031] By merging spatially adjacent abnormal computational units into connected regions, the non-uniform reaction region within the cobalt metal bed is obtained.

[0032] The fixed-bed reaction flow field control scheme includes:

[0033] Based on the effective specific surface area of ​​the reaction, local reaction rate, reactant concentration, cobalt acetate formation concentration and cobalt metal consumption corresponding to the uneven reaction region, it is determined whether there is a need for cobalt metal replenishment in the uneven reaction region.

[0034] When the effective reaction surface area in the uneven reaction region is lower than a preset specific surface area threshold, and the local reaction rate is lower than a preset reaction rate threshold, the corresponding uneven reaction region is identified as a region with insufficient cobalt metal reaction capacity, and cobalt metal replenishment demand information is generated. The cobalt metal replenishment demand information includes at least one of the following: replenishment area, replenishment amount demand, replenishment timing demand, and target effective reaction surface area after replenishment.

[0035] The fixed-bed reaction flow field control scheme further includes:

[0036] Based on the cobalt metal feeding requirement information, candidate feeding schemes are generated. The candidate feeding schemes include at least one of the following: feeding location, feeding amount, particle size range, morphology type, packing density, and packing layers of cobalt metal.

[0037] Based on the candidate feeding scheme, the porosity distribution and effective reaction specific surface area distribution in the cobalt fixed bed reaction zone are updated to obtain the predicted bed structure after feeding.

[0038] Input the predicted bed structure after feeding into the numerical simulation model to calculate the predicted reaction data after feeding corresponding to the candidate feeding scheme;

[0039] The degree of feed disturbance corresponding to the candidate feed scheme is determined based on the predicted reaction data after feed feeding, and the fixed bed reaction flow field control scheme is determined based on the degree of feed disturbance.

[0040] The methods for determining the degree of material feeding disturbance include:

[0041] Based on the candidate feeding scheme, determine the newly added effective reaction area region of metallic cobalt in the predicted bed structure after feeding;

[0042] Based on the predicted reaction data after feeding, determine the area within the cobalt fixed-bed reaction zone that can be continuously covered by the reaction liquid through which the reaction liquid continuously flows.

[0043] Determine the overlap and connectivity between the newly added effective reaction area of ​​cobalt metal and the continuously covered area of ​​the reaction liquid;

[0044] Based on the overlapping state and the connectivity state, the effective contact degree of the newly added reaction area corresponding to the candidate feeding scheme is determined, and the effective contact degree of the newly added reaction area is used as the feeding disturbance degree corresponding to the candidate feeding scheme.

[0045] The fixed-bed reaction flow field control scheme further includes:

[0046] The degree of material feeding disturbance is compared with the preset disturbance allowable conditions;

[0047] When the degree of feeding disturbance does not meet the preset disturbance allowable condition, at least one of the following in the candidate feeding scheme is updated according to the overlap and connection state between the newly added effective reaction area of ​​metallic cobalt and the continuous coverage area of ​​the reaction liquid: feeding position, feeding amount, particle size range, morphology type, filling density and filling layer.

[0048] Based on the updated candidate feeding schemes, the predicted bed structure after feeding is redefined, and the predicted reaction data after feeding corresponding to the updated candidate feeding schemes are recalculated.

[0049] Based on the recalculated predicted reaction data after feeding, the feeding disturbance level corresponding to the updated candidate feeding scheme is determined again until the feeding disturbance level meets the preset disturbance allowance condition.

[0050] The candidate feeding scheme that meets the preset disturbance allowance condition is determined as the fixed bed reaction flow field control scheme.

[0051] A fixed-bed reaction process coupling control system based on industrial cloud computing, the system comprising:

[0052] The numerical simulation module is used to acquire the physical properties and operating parameters of the reaction liquid in the fixed-bed reactor, and to obtain the reaction data in the fixed-bed reactor by multi-field coupling solution through a preset numerical simulation model. The numerical simulation model includes a porous media model for characterizing the flow resistance of the reaction liquid in the cobalt bed, a component transfer model for characterizing the mass transfer process between cobalt and the reaction liquid, a heat transfer model for characterizing the exothermic or endothermic process of the reaction, and a reaction kinetic model for characterizing the process of cobalt forming cobalt acetate.

[0053] The reaction state identification module is used to identify the reaction uneven region in the cobalt metal bed based on the reaction data, and to determine whether there is a need for cobalt metal replenishment in the reaction uneven region based on the effective reaction specific surface area, local reaction rate, reactant concentration value, cobalt acetate generation concentration value and cobalt metal consumption corresponding to the reaction uneven region.

[0054] The control scheme generation module is used to generate candidate feeding schemes when there is a demand for cobalt metal feeding, update the porosity distribution and effective reaction specific surface area distribution in the cobalt metal fixed bed reaction zone based on the candidate feeding schemes, obtain the predicted bed structure after feeding, and input the predicted bed structure after feeding into the numerical simulation model to obtain predicted reaction data after feeding; and determine the feeding disturbance degree corresponding to the candidate feeding scheme based on the predicted reaction data after feeding, and update the candidate feeding scheme based on the feeding disturbance degree until the updated candidate feeding scheme meets the preset disturbance allowable condition, and determine the candidate feeding scheme that meets the preset disturbance allowable condition as the fixed bed reaction flow field control scheme.

[0055] Compared with the prior art, the beneficial effects of the present invention are:

[0056] This application obtains the internal flow, mass transfer, heat transfer and reaction rate distribution of a cobalt fixed bed through multi-field coupled numerical simulation, which can identify regions with uneven reaction and determine the cobalt feeding requirements; further, it updates the feeding scheme based on the contact relationship between the newly added effective reaction area after feeding and the continuous coverage area of ​​the reaction liquid, thereby improving the effective utilization rate of cobalt and the stability of cobalt acetate formation. Attached Figure Description

[0057] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0058] Figure 1 An exemplary application scenario diagram provided for an embodiment of this application;

[0059] Figure 2 A schematic flowchart of the fixed-bed reaction process coupling control method based on industrial cloud computing provided in the embodiments of this application;

[0060] Figure 3 This is a schematic diagram illustrating the generation principle of the numerical simulation model provided in the embodiments of this application.

[0061] Figure label:

[0062] 100. Fixed-bed reactor; 101. Reactor shell; 102. Reaction liquid inlet; 103. Reaction liquid outlet; 104. Cobalt metal fixed bed; 105. Feed distributor; 106. Bed support; 107. Circulation pipeline; 108. Feed inlet; 109. Temperature sensor; 110. Pressure sensor; 111. Cobalt metal raw material; 112. Arrow. Detailed Implementation

[0063] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0064] The term "embodiment" as used herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0065] This application applies to reaction processes that use metallic cobalt as a solid reactant to form a fixed bed, and allow a reaction liquid containing acetic acid to flow and contact within the fixed bed to prepare cobalt acetate. This type of process differs from conventional fixed-bed operations that use inert packing or fixed catalysts. In this case, metallic cobalt itself constitutes both the flow framework of the bed and is continuously consumed as a reactant phase during operation. Therefore, the flow state, mass transfer state, and reaction state within the fixed bed are not solely determined by the reactor shell, inlet / outlet structure, or external circulation parameters, but are closely related to the morphology of the metallic cobalt feedstock—whether it is blocky, granular, flaky, or irregularly shaped—and the resulting local porosity, exposed surface area, and packing contact relationship.

[0066] In actual production, cobalt raw materials are usually not ideally regular particles; their size, shape, surface roughness, and contact / obstruction status after loading all vary. Larger bulk cobalt may form supporting voids within the bed, allowing the reaction liquid to preferentially pass through low-resistance channels. Smaller particles or cobalt with more broken edges and corners may create larger fresh exposed areas locally, causing the reaction to occur in concentrated areas. Flake-shaped or irregularly shaped clumps of cobalt may also cause bridging, obstruction, or local compaction after accumulation, preventing continuous and effective contact between some surfaces, although geometrically belonging to the cobalt surface, and the reaction liquid. Therefore, within the same fixed-bed reactor, there may simultaneously exist areas where the reaction liquid passes through rapidly and does not stay sufficiently, areas where the reaction liquid is slowly renewed, areas with insufficient effective reaction area, and areas with excessively strong local reactions. These areas often manifest as combined changes in pressure drop, outlet concentration, or temperature in external process parameters, but their true source lies within the bed and is difficult to observe directly using conventional detection methods.

[0067] Current production control methods typically rely on empirical judgment based on macroscopic parameters such as feed flow rate, circulation flow rate, reaction temperature, outlet cobalt acetate concentration, or bed pressure drop. For example, when the outlet concentration decreases, operators may increase the circulation rate, adjust the reaction temperature, or add a certain amount of metallic cobalt; when the pressure drop is abnormal, it may be assumed that the bed is blocked or locally overly dense. However, this control method is difficult to distinguish whether the problem stems from uneven distribution of the reaction liquid, insufficient effective reaction surface area of ​​metallic cobalt, limited local mass transfer, or local reaction rate imbalance. Especially for fixed beds used in the preparation of cobalt acetate from metallic cobalt, the bed structure is not static. The metallic cobalt is gradually consumed as the reaction proceeds, and its effective reaction area, local porosity, and flow path all change. Relying solely on initial loading experience or empirical corrections under a single steady-state condition is insufficient to continuously ensure uniform contact between the reaction liquid and metallic cobalt.

[0068] The concept of this application is based on the aforementioned process characteristics. For cobalt-consuming fixed-bed reaction processes, simply describing the velocity field or pressure drop distribution within the reactor is insufficient to reflect the true state. It is necessary to treat the cobalt bed as a spatially distributed object possessing both "flow resistance properties" and "reaction area properties." In other words, whether a certain region in the bed can effectively participate in cobalt acetate formation depends not only on the presence of cobalt at that location, but also on whether the effective reactive surface area of ​​the cobalt at that location can be continuously covered by the reaction liquid, whether the reactants can continuously reach the cobalt surface, and whether the heat generated or consumed by the reaction can be transferred under the corresponding flow conditions.

[0069] In its implementation, this application does not simply equate the cobalt bed to a uniform porous medium. Instead, it combines the morphological distribution, particle size distribution, bulk density distribution, porosity distribution, and effective reaction surface area distribution of the cobalt raw material to establish a numerical simulation model that can simultaneously characterize flow resistance, component transfer, heat transfer, and the reaction process of cobalt to cobalt acetate.

[0070] It should be noted that the fixed-bed reaction flow field control method disclosed in this application is not limited to a fixed-bed reactor of a specific size or internal structure. It is applicable as long as the reaction zone contains a bed formed by the accumulation of cobalt raw materials, and the reaction liquid contacts the cobalt bed to generate cobalt acetate. Application scenarios can include vertical fixed-bed reactors, fixed-bed reactors with circulating liquid distribution structures, fixed-bed reactors with top-in / bottom-out or bottom-in / top-out flow patterns, and reactors equipped with guiding components, distribution components, or bed support components. A common feature of these application scenarios is that the actual flow path of the reaction liquid is affected by the cobalt accumulation structure, the actual reactivity of the cobalt is affected by its effective exposure area and liquid-phase contact state, and the differences in reaction within the bed cannot be directly determined solely by external macroscopic parameters.

[0071] It should also be noted that the non-uniform reaction region in this application does not refer only to the traditional low-velocity region or flow deviation region, but rather to the region in the bed spatial distribution results obtained by numerical simulation where the reaction state of metallic cobalt deviates from the target state due to imbalances in the flow field, mass transfer, heat transfer, or local reaction rate distribution. This region can manifest as a high flow rate of the reaction liquid but insufficient residence time, a low flow rate with insufficient reactant renewal, insufficient effective reaction surface area leading to a decrease in cobalt acetate formation capacity, or excessive concentration of fresh surface leading to excessively high local reaction intensity. By incorporating the above phenomena into the numerical simulation model for multi-field coupled solution, this application can transform the bed internal state, which originally relied on empirical judgment, into calculable, locatable, and controllable spatial information.

[0072] Building upon this, this application further considers the unique operating characteristics of a cobalt fixed bed. Since cobalt is continuously consumed as a reactant, the need to replenish cobalt during production is unavoidable. Conventionally, replenishment is typically understood as restoring the existing cobalt stock in the bed; that is, adding a certain amount of cobalt when it is insufficient.

[0073] It is important to emphasize that the effectiveness of feedstock replenishment does not depend on the total amount of cobalt added, but rather on whether the newly added cobalt can be continuously and effectively contacted by the reaction liquid. If the added cobalt causes localized compaction or obstruction, even with sufficient feedstock, new ineffective reaction areas may be created. If the particle size, shape, or packing layer of the added cobalt is unreasonable, it may alter the original flow path, causing the reaction liquid to bypass the feedstock area or only react locally within the feedstock area, thus preventing the feedstock operation from truly improving the overall reaction state of the fixed bed.

[0074] Therefore, when generating a fixed-bed reaction flow field control scheme, this application does not treat feeding as a simple follow-up action, but incorporates feeding location, feeding amount, particle size range, morphology type, packing density, and packing layers into the bed structure prediction process. By predicting and updating the porosity distribution and effective reaction surface area distribution after feeding, and then using a numerical simulation model to calculate the predicted reaction data after feeding, the contact relationship between the newly added effective reaction area after feeding and the area continuously covered by the reaction liquid can be determined. This determines whether the candidate feeding scheme can truly form a usable reaction area, and updates the candidate feeding scheme when necessary, so that the final fixed-bed reaction flow field control scheme can not only supplement the area with insufficient cobalt reaction capacity, but also avoid the introduction of new flow imbalances or ineffective packing by feeding.

[0075] refer to Figure 1 , Figure 1 This is an exemplary application scenario diagram provided for an embodiment of this application.

[0076] like Figure 1 As shown, the application scenarios include a fixed-bed reactor 100, a reactor shell 101, a reaction liquid inlet 102, a reaction liquid outlet 103, a fixed cobalt metal bed 104, a feed distributor 105, a bed support 106, a circulation pipeline 107, a feed inlet 108, a temperature detection device 109, a pressure detection device 110, and a cobalt metal raw material 111.

[0077] The fixed-bed reactor 100 is used to contact the reaction liquid with cobalt raw material 111 to prepare cobalt acetate. The reactor shell 101 forms a reaction chamber for the flow of the reaction liquid and the loading of cobalt raw material 111. The cobalt raw material 111 is stacked inside the reactor shell 101 to form a fixed cobalt bed 104. The cobalt raw material 111 may include at least one of blocky, granular, flake, or irregularly shaped cobalt. After the cobalt raw materials 111 of different shapes and sizes are stacked inside the reactor shell 101, a fixed bed with a non-uniform porosity structure is formed.

[0078] The reaction liquid inlet 102 is located at the upper part of the reactor shell 101, and the reaction liquid outlet 103 is located at the lower part of the reactor shell 101. A feed distributor 105 is located between the reaction liquid inlet 102 and the cobalt metal fixed bed 104, used to distribute the reaction liquid entering through the reaction liquid inlet 102 to the upper region of the cobalt metal fixed bed 104. A bed support 106 is located below the cobalt metal fixed bed 104, used to support the cobalt metal fixed bed 104 and allow the reaction liquid to pass through. As shown by arrow 112, the reaction liquid enters the fixed bed reactor 100 through the reaction liquid inlet 102, flows through the cobalt metal fixed bed 104 after being distributed by the feed distributor 105, and reacts during contact with the cobalt metal raw material 111, subsequently being discharged through the reaction liquid outlet 103.

[0079] A circulation pipeline 107 is connected between the reaction liquid outlet 103 and the reaction liquid inlet 102, used to circulate at least a portion of the reaction liquid back to the fixed-bed reactor 100 to adjust the contact and residence states of the reaction liquid within the cobalt metal fixed bed 104. A feed inlet 108 is located at the top or upper region of the reactor shell 101 and communicates with the interior of the reactor shell 101, used to replenish the cobalt metal raw material 111 to the cobalt metal fixed bed 104 after the cobalt metal raw material 111 is consumed during the reaction. Temperature sensors 109 and pressure sensors 110 are located on the reactor shell 101 or on the pipeline connected to it, used to acquire temperature and pressure information during the operation of the fixed-bed reactor 100.

[0080] It should be noted that, Figure 1 This application only illustrates one exemplary application scenario of the present application and does not constitute a limitation on the specific structural form of the fixed-bed reactor 100. In other embodiments, the positions of the reaction liquid inlet 102 and the reaction liquid outlet 103 can be adjusted according to the actual flow pattern, such as adopting a bottom-in-top-out, top-in-bottom-out, or side-in-out structure; the specific shape and arrangement of the feed distributor 105, bed support 106, circulation pipeline 107, and feed port 108 can also be set according to the size, filling method, and process requirements of the fixed-bed reactor 100. As long as the reactor is filled with cobalt raw material 111 to form a cobalt fixed bed 104, and the reaction liquid flows through or contacts the cobalt fixed bed 104 to generate cobalt acetate, it can be considered as the applicable object of the method and system of this application.

[0081] It should be noted that the method described in this application can specifically be a program instruction, which is stored in an industrial cloud computing platform and executed by the computing resources in the industrial cloud computing platform. Specifically, the industrial cloud computing platform may include a data access layer, a model calculation layer, and a control and analysis layer. The data access layer is used to receive industrial production data such as reaction liquid physical properties, reaction operation parameters, and bed structure parameters collected during the operation of the fixed-bed reactor. The model calculation layer is used to call a pre-built numerical simulation model and use cloud computing resources to perform parallel calculations and multi-field coupling solutions on the porous media model, component transfer model, heat transfer model, and reaction kinetic model to obtain reaction state data such as flow rate, pressure, temperature, concentration, and local reaction rate at different spatial locations inside the fixed-bed reactor. The control and analysis layer is used to identify reaction inhomogeneity regions based on the reaction state data and further predict the impact of different control schemes on the bed structure and reaction state through cloud simulation calculations, thereby generating a fixed-bed reaction flow field control scheme that meets preset control conditions. By deploying the above numerical simulation, state analysis, and control optimization processes on an industrial cloud computing platform, the elastic computing capabilities of the cloud can be used to realize large-scale calculations of complex multiphysics models. This avoids the problems of low model solving efficiency and difficulty in real-time optimization caused by the limitation of computing resources on traditional local computing devices. At the same time, it can realize model reuse, data sharing, and remote intelligent control for different fixed-bed reaction equipment.

[0082] Next, with reference to the accompanying drawings, the fixed-bed reaction process coupling control method based on industrial cloud computing provided in the embodiments of this application will be further described. Figure 2 The methods shown include:

[0083] S1: Obtain the physical properties and operating parameters of the reaction liquid in the fixed-bed reactor, and solve the reaction data in the fixed-bed reactor by multi-field coupling through a preset numerical simulation model;

[0084] The numerical simulation models include a porous media model for characterizing the flow resistance of the reaction liquid in the cobalt bed, a component transfer model for characterizing the mass transfer process between cobalt and the reaction liquid, a heat transfer model for characterizing the exothermic or endothermic process of the reaction, and a reaction kinetic model for characterizing the process of cobalt forming cobalt acetate.

[0085] In this embodiment, the fixed-bed reactor can be Figure 1The vertical fixed-bed reactor 100 shown has cobalt raw material 111 packed inside the reactor shell 101 to form a cobalt fixed bed 104. The reaction liquid enters through the reaction liquid inlet 102 and is distributed by the feed distributor 105. As it flows through the cobalt fixed bed 104, it contacts and reacts with the cobalt raw material 111 to generate cobalt acetate. The physical properties of the reaction liquid may include at least one of the following: reaction liquid density, viscosity, specific heat capacity, thermal conductivity, diffusion coefficient, and initial concentration of reactants. The reaction operating parameters may include at least one of the following: feed flow rate, circulation flow rate, feed temperature, operating pressure, bed pressure drop, reaction time, and outlet cobalt acetate concentration.

[0086] Those skilled in the art will understand that the above parameters can be collected according to the detection conditions of the actual production site, such as by temperature detection device 109, pressure detection device 110, flow meter, concentration detection device or production records. As long as they can characterize the current operating status of the fixed bed reactor, this application does not impose any further restrictions on the specific collection method.

[0087] S2: Identify the regions of uneven reaction within the cobalt metal bed based on the reaction data;

[0088] In this embodiment, the non-uniform reaction region is not limited to the traditional flow deviation region or low-speed region, but refers to the region within the cobalt metal fixed bed where the local reaction state deviates from the target reaction state due to differences in reaction liquid distribution, reactant supply, heat migration, or effective reaction specific surface area of ​​cobalt metal.

[0089] Specifically, the reaction data obtained from multi-field coupling can be mapped to the computational units corresponding to the cobalt fixed bed. By combining the flow rate, pressure, temperature, reactant concentration, cobalt acetate formation concentration, and local reaction rate values ​​of each computational unit, it can be determined whether local areas exhibit conditions such as rapid reaction liquid flow with insufficient residence, slow reaction liquid renewal, insufficient reactant supply, abnormal local temperature, excessively strong local reaction, or insufficient local reaction. Because cobalt raw materials typically exist in various morphologies such as lumps, granules, flakes, or irregular clumps, the porosity and effective reaction surface area vary across different regions. Therefore, even with identical external feed conditions, spatial regions with significantly different reactivity may still form within the bed.

[0090] S3: Based on the location and range of the non-uniform reaction region and its corresponding flow field parameters, mass transfer parameters, heat transfer parameters and local reaction rate parameters, generate a fixed-bed reaction flow field control scheme;

[0091] In this embodiment, the fixed-bed reaction flow field control scheme is not simply adjusted by regulating the feed flow rate or circulation flow rate, but is specifically determined based on the causes of the uneven reaction regions. When the uneven reaction regions are mainly characterized by excessively high local flow velocities and insufficient residence time, the control scheme may include adjusting the feed distribution method, circulation flow rate, or flow guiding structure; when the uneven reaction regions are mainly characterized by insufficient reactant concentration or insufficient mass transfer, the control scheme may include adjusting the feed concentration, feed flow rate, circulation method, or local bed packing state.

[0092] It should be noted that when the uneven reaction region manifests as insufficient effective reaction surface area accompanied by a local decrease in reaction rate, the control scheme can further include a cobalt metal feeding scheme. This includes determining the feeding area, the required feeding amount, the timing of feeding, and the target effective reaction surface area after feeding. Since cobalt metal is both a component of the fixed bed and a continuously consumed solid reactant in the preparation of cobalt acetate, feeding should not be simply understood as adding the same mass of cobalt metal. Instead, it should be determined by considering the change in bed porosity after feeding and whether the newly added effective reaction area can be continuously contacted by the reaction liquid.

[0093] Next, we will further elaborate on the technical content of the numerical simulation model in this application.

[0094] like Figure 3 As shown, the process of constructing the numerical simulation model may include: obtaining the basic parameters of the fixed-bed reactor used for the reaction of metallic cobalt to prepare cobalt acetate; determining the morphological distribution, particle size distribution, bulk density distribution, porosity distribution, and effective reaction specific surface area distribution of the metallic cobalt raw material in the fixed-bed according to the metallic cobalt bed packing parameters; and constructing a numerical simulation model of the metallic cobalt fixed-bed reaction zone based on the reactor structural parameters, porosity distribution, and effective reaction specific surface area distribution.

[0095] It should be noted that the cobalt fixed-bed reaction zone can be divided into multiple computational units, each corresponding to a local space within the bed. For each computational unit, the permeability and inertial drag parameters can be determined based on the porosity, cobalt particle size, cobalt morphology parameters, and packing density within that unit. This allows the porous media model to reflect the differences in resistance to the flow of the reaction liquid in different local areas. Compared to using uniform bed drag parameters, this approach can express the flow differences caused by varying cobalt particle sizes, local packing density, and irregular morphologies, thus making the resulting velocity and pressure fields closer to the actual operating state of the cobalt fixed bed.

[0096] It is understood that the numerical simulation model in this application includes multiple sub-models, and the specific implementation of each sub-model can refer to the multi-field coupling model based on the physical control equations, rather than the neural network model that is mainly based on black box prediction.

[0097] In one example, the porous media model can adopt the Darcy-Fochheimer resistance form, using the permeability and inertial resistance parameters corresponding to each computational unit in the cobalt fixed bed as resistance source terms in the momentum equation. This allows the reaction liquid to exhibit different flow resistances in regions with different porosities, different packing densities, and different cobalt metal morphologies. This can better express the local flow differences formed after the irregular packing of cobalt raw materials, avoiding the equivalent of the entire bed as a uniform resistance region. The permeability and inertial resistance parameters can be calculated based on the physical laws corresponding to the material of the computational unit, or determined by a large number of experiments conducted by those skilled in the art, which will not be elaborated here.

[0098] In another example, the component transport model can employ a convection-diffusion-reaction form of component transport equations. The migration of reactants in the liquid phase is determined by local flow rate, diffusion coefficient, and concentration gradient. The reactant consumption resulting from the reaction on the cobalt surface is presented as a consumption term in the component equation, while the cobalt acetate formation is presented as a formation term in the product component equation. For each computational unit, the consumption and formation terms can be correlated with the effective reaction surface area, local mass transfer coefficient, and local reaction rate of that unit. This allows for the differentiation between two different scenarios: "the reaction liquid can flow through but does not stay sufficiently" and "the cobalt surface is not adequately fed."

[0099] In another example, the heat transfer model can employ an energy equation that includes convective heat transfer, conduction, and reaction heat source terms. The reaction liquid carries heat as it flows within the bed, and solid-liquid heat exchange occurs between the cobalt bed and the reaction liquid. The reaction heat during the formation of cobalt acetate from cobalt forms a local heat source term based on the local reaction rate and effective reaction surface area. This local heat source term varies with the reaction state of the computational unit, ensuring that the temperature field is no longer solely determined by the inlet temperature or external heat transfer conditions, but rather reflects the differences in heat distribution caused by the intensity of local reactions within the bed.

[0100] In another example, the reaction kinetic model can adopt an apparent kinetic form, where the local reaction rate is set as a function determined by the effective reaction surface area, reactant concentration, and temperature. Specifically, within the same computational unit, the effective reaction surface area characterizes the number of cobalt metal reaction interfaces that can actually contact the reaction liquid; as the effective reaction surface area increases, more cobalt metal surfaces can participate in the reaction, correspondingly increasing the local reaction capacity. The reactant concentration characterizes the level of reactant supply available for cobalt metal surfaces in the liquid phase; when the reactant concentration is high, the ability of the cobalt metal surface to obtain reactants is enhanced, and the local reaction rate increases accordingly; when the reactant concentration decreases to a certain range, even if there is a large exposed area of ​​cobalt metal in that region, the local reaction rate will be limited due to insufficient reactant supply. Temperature characterizes the reaction activation conditions; within the allowable process temperature range, an increase in temperature usually increases the apparent reaction rate constant, making the reaction of cobalt metal to cobalt acetate more likely to occur. However, when the temperature exceeds the preset safety or process stability range, a temperature correction term can be used to limit the local reaction rate to avoid misjudging abnormal local temperature increases as effective reaction enhancement.

[0101] In this embodiment, the aforementioned multiple sub-models can be discretely solved using the finite volume method, finite element method, or finite difference method. During the solution process, the local flow velocity and pressure distribution within each computational unit are first updated based on the porous medium model. Then, the local flow velocity influences the component transport process, and the component concentration and temperature jointly influence the local reaction rate. Finally, the local reaction rate updates the reactant consumption, cobalt acetate production, and local heat source term. This process is repeated iteratively until the changes in velocity, temperature, concentration, and local reaction rate satisfy a preset convergence condition. This coupled solution method allows differences in the packing structure, liquid phase flow, and reaction formation of the cobalt bed to be mutually transmitted within the same computational framework.

[0102] In other optional implementations, the numerical simulation model may also include a neural network surrogate model for accelerating prediction. However, this neural network surrogate model does not replace the aforementioned physical meaning but is used to approximate the multi-field coupling solution results. Specifically, the inputs to the neural network surrogate model may include the porosity, cobalt particle size, cobalt morphology parameters, bulk density, effective reaction surface area, reaction liquid density, viscosity, diffusion coefficient, feed flow rate, feed temperature, operating pressure, and initial reactant concentration of each computational unit; the outputs may include the local flow rate, temperature, reactant concentration, cobalt acetate formation concentration, and local reaction rate corresponding to each computational unit. The neural network surrogate model may employ a multilayer perceptron structure, including an input layer, at least two hidden layers, and an output layer. The hidden layers may use ReLU, tanh, or sigmoid activation functions, and the output layer may use a linear output; the training samples may consist of the physical multi-field coupling solution results, experimental calibration data, or production operation data, and the loss function may include the mean square error term between the predicted flow rate, temperature, concentration, and local reaction rate and the calibration results. Therefore, while ensuring that the model input and output have clear physical meaning, the computational efficiency can be improved when updating the feeding scheme or adjusting the operating parameters.

[0103] In one example, the numerical simulation model is constructed in the following ways:

[0104] The basic parameters of the fixed-bed reactor used for the reaction of metallic cobalt to prepare cobalt acetate are obtained, including reactor structural parameters and metallic cobalt bed packing parameters;

[0105] Based on the aforementioned cobalt bed loading parameters, the morphological distribution, particle size distribution, bulk density distribution, porosity distribution, and effective reaction specific surface area distribution of the cobalt raw material in the fixed bed are determined.

[0106] A numerical simulation model of the cobalt fixed-bed reaction zone is constructed based on the reactor structural parameters, the porosity distribution, and the effective reaction surface area distribution.

[0107] It should be noted that the basic parameters of a fixed-bed reactor are not only used to establish the geometric boundaries of the reactor shell, but also to transform the "reactor spatial structure" and the "actual filling state of the cobalt bed" into a calculable description of the reaction zone.

[0108] In practical implementation, the geometric boundaries of the fixed-bed reaction zone can be determined first based on the reactor inner diameter, bed height, inlet and outlet positions, feed distributor position, bed support position, and feed inlet position. Then, based on the screening results, weighing results, filling records, image measurement results, or historical operation calibration results of the cobalt raw material, the particle size range, morphology, packing density, and unit mass surface area of ​​cobalt in different regions can be obtained. For blocky, granular, flake, or irregularly shaped cobalt, corresponding equivalent particle size, morphology correction coefficient, and surface area correction coefficient can be set respectively, so that the influence of different forms of cobalt on the flow space and reaction contact area in the bed can be distinguished and expressed. Those skilled in the art will understand that the above-mentioned filling parameters do not need to achieve precise reconstruction at the single-particle level; it is sufficient to express the differences in porosity, flow resistance, and effective reaction area in different regions of the bed to a minimum to meet the requirements for constructing a numerical simulation model.

[0109] The equivalent particle size, morphology correction factor, and surface area correction factor can be determined by referring to the conventional equivalent treatment methods in fixed bed packing, granular bed, or porous media modeling. Specifically, the equivalent particle size can be selected based on the sieved particle size, median particle size, average particle size, or external dimensions obtained from image measurement of the cobalt raw material. When multiple particle sizes exist in the same region, a representative particle size can be determined according to the mass or volume percentage of each particle size as the equivalent particle size for that region.

[0110] The shape correction factor can be set according to the degree of deviation of the cobalt raw material from the shape of regular particles. For example, for cobalt that is close to regular granular shape, the shape correction factor can be set to close to 1; for blocky, flake, or irregularly shaped cobalt, the shape correction factor can be appropriately increased according to its major and minor axis ratio, angularity, and tendency to bridging or shading after stacking. For example, the shape correction factor for granular cobalt can be 1.0 to 1.1, for blocky cobalt it can be 1.1 to 1.3, for flake cobalt it can be 1.3 to 1.6, and for irregularly shaped cobalt it can be 1.2 to 1.5.

[0111] The surface area correction factor can be set according to the surface roughness of cobalt metal, the degree of mutual contact and obstruction, and the contactability of the reaction liquid. For areas with relatively regular surfaces, less obstruction, and continuous flow of the reaction liquid, a higher surface area correction factor can be used; for areas with local compaction, significant obstruction from adjacent cobalt metal, or insufficient reaction liquid renewal, a lower surface area correction factor can be used. For example, the surface area correction factor can be 0.5 to 1.2, where areas with strong contact and obstruction can be 0.5 to 0.8, areas with good flow conditions can be 0.8 to 1.0, and areas with rough surfaces and sufficient liquid phase contact can be 1.0 to 1.2.

[0112] Furthermore, the porosity distribution can be determined by the cobalt packing volume, local bulk density, equivalent particle size, and morphology correction factor; the effective reaction surface area distribution can be further modified based on the theoretical surface area of ​​cobalt. This modification is not a simple subtraction of a fixed proportion, but rather takes into account factors such as contact obstruction between cobalt particles, local compaction, liquid phase accessibility, and the flow state of the reaction liquid. For example, surfaces located at the point of contact between adjacent cobalt particles, areas obstructed by large cobalt particles, or in regions where the reaction liquid is difficult to continuously renew can be considered as insufficiently effective surfaces; surfaces located near the main flow path and capable of continuous contact with the reaction liquid can be considered as effective reaction surfaces. The resulting effective reaction surface area distribution reflects the difference between the presence of a cobalt particle surface and the surface's actual participation in the cobalt acetate formation reaction.

[0113] Furthermore, after obtaining the porosity distribution and effective reaction surface area distribution, the cobalt fixed-bed reaction zone can be discretized into multiple computational units, and each computational unit can be assigned initial values ​​for local porosity, permeability, inertial drag coefficient, effective reaction surface area, mass transfer coefficient, and reaction rate parameters. Thus, the porous media model no longer uses uniform resistance parameters across the entire bed, but reflects the flow resistance of the reaction liquid in different regions based on local porosity and packing conditions; the component transfer model no longer only describes liquid phase concentration diffusion, but is related to the local effective reaction area and liquid phase accessibility; the reaction kinetic model no longer uses the total amount of cobalt added as the basis for reaction capacity, but uses the area of ​​cobalt within the computational unit that can be effectively utilized by the reaction liquid as the basis for local reaction capacity.

[0114] Understandably, once the porosity distribution and effective reactive surface area distribution are obtained, the specific method for assigning values ​​to the computational units can refer to existing techniques such as mesh attribute mapping, region averaging, interpolation, or element parameter initialization methods in the finite volume method. For example, the reaction zone can be divided into several computational units based on the axial height, radial position, and structural changes near the feed distributor, bed supports, and feed inlet of the fixed bed reaction zone. For computational units that fall entirely within the same filling area, the porosity, effective reactive surface area, and packing state parameters corresponding to that filling area can be directly assigned to the computational unit. For computational units that span two or more filling areas, averaging can be performed based on the proportion of each filling area within the computational unit to obtain a representative local parameter for that computational unit.

[0115] For example, calculation units located near the reactor inlet or feed distributor can appropriately increase the initial value of the mass transfer coefficient based on the inlet flow distribution; calculation units located near the reactor wall, bed supports, or locally compacted areas can assign relatively strong initial resistance values ​​to the permeability and inertial drag coefficient based on wall effects, support obstruction, or packing density; calculation units located in areas where cobalt particles are fine and their surfaces are sufficiently exposed can be assigned a high initial value for the effective reaction specific surface area; and calculation units located in areas where large cobalt particles are obstructed, areas where sheet-like cobalt particles are bridged, or areas where liquid phase renewal is insufficient can be assigned a low initial value for the effective reaction specific surface area.

[0116] The above assignment process can be achieved through material property definition, region parameter mapping, user-defined field functions, or mesh initialization files in existing numerical simulation software. The initial values ​​of permeability, inertial drag coefficient, mass transfer coefficient, and reaction rate parameters can be determined based on existing fixed-bed reactor design manuals, empirical parameters of similar granular beds, pre-experiment calibration results, or historical stable operation data, and will be updated according to local flow velocity, concentration, temperature, and reaction state during subsequent multi-field coupled solutions.

[0117] In one optional implementation, the numerical simulation model is further constructed by:

[0118] The cobalt fixed-bed reaction zone is divided into multiple computational units;

[0119] For each calculation unit, the permeability parameters and inertial drag parameters of the calculation unit are determined according to the porosity, cobalt particle size, cobalt morphology parameters and packing density of the calculation unit, and the porous medium model is established based on the permeability parameters and inertial drag parameters.

[0120] For each calculation unit, the local mass transfer coefficient between the cobalt metal surface and the reaction liquid is determined based on the effective reaction surface area, reaction liquid flow rate and reactant concentration corresponding to the calculation unit, and the component transfer model is established based on the local mass transfer coefficient.

[0121] For each calculation unit, the local heat source term of the calculation unit is determined according to the effective reaction specific surface area, local reaction rate and reaction heat parameters corresponding to the calculation unit, and the heat transfer model is established based on the local heat source term.

[0122] For each calculation unit, the local reaction rate of cobalt metal to cobalt acetate is determined based on the effective reaction surface area, reactant concentration and temperature corresponding to the calculation unit, and the reaction kinetic model is established based on the local reaction rate.

[0123] It should be noted that during the construction of the sub-model, the local loading parameters, local flow parameters, local mass transfer parameters, local heat transfer parameters, and local reaction parameters corresponding to each calculation unit need to be established as a unified set of unit parameters.

[0124] In other words, the porosity, cobalt particle size, cobalt morphology parameters, bulk density, and effective reaction surface area within the same computational unit are not isolated parameters that can be independently accessed by different sub-models. Instead, they collectively describe the local properties of the cobalt packing state and the reachable state of the reaction liquid within that computational unit. In practice, the cobalt fixed-bed reaction zone can be divided into multiple computational units along the axial and radial directions. Alternatively, a non-uniform division can be used based on differences in feed distributors, bed supports, areas near feed inlets, or local bed packing. For each computational unit, the cobalt packing state is approximated as an equivalent porous reaction region within a local area. This allows the local area to provide both flow resistance to the reaction liquid and a reaction interface between the cobalt and the reaction liquid. This approach avoids simply averaging the entire cobalt bed and is particularly suitable for fixed-bed reaction scenarios where the cobalt raw material has significant differences in particle size, irregular morphology, and varying degrees of local packing density.

[0125] Next, we will further elaborate on the technical aspects of the multi-field coupling solution method in this application.

[0126] In one example, the multi-field coupling solution includes:

[0127] The physical properties of the reaction liquid are used as boundary conditions, and the reaction operating parameters are used as initial conditions input into the numerical simulation model.

[0128] Based on the coupling relationship between the porous medium model, the component transport model, the heat transfer model and the reaction kinetics model, the reaction process of reaction liquid flow, reactant mass transfer, heat transfer and the formation of cobalt acetate from metal cobalt is solved synchronously and iteratively in the metal cobalt fixed bed reaction zone.

[0129] During the synchronous iterative solution process, the local mass transfer coefficient and local reaction rate are updated based on the local flow rate, local temperature and local reactant concentration obtained in the current iteration step. The reactant consumption, cobalt acetate production and local heat source terms are corrected based on the updated local reaction rate until the changes in flow rate, temperature, concentration and local reaction rate between adjacent iteration steps meet the preset convergence conditions.

[0130] The bed spatial distribution result obtained when the preset convergence condition is met is used as the reaction data.

[0131] In this embodiment, when the physical properties of the reaction liquid are input into the numerical simulation model, they are not required to be fixed values ​​on the geometric boundaries. Parameters such as density, viscosity, specific heat capacity, thermal conductivity, and diffusion coefficient can be written into the computational domain as medium properties required for solving the flow equation, component transport equation, and energy equation. Parameters such as initial reactant concentration, inlet reactant concentration, and inlet temperature can be assigned as inlet boundary or initial field values. Reaction operating parameters can be converted into corresponding solution conditions based on actual operating conditions. For example, the feed flow rate can be converted into the average inlet velocity or mass flow rate boundary of the reaction liquid inlet section, the operating pressure can be set as the outlet pressure boundary or reference pressure, the circulating flow rate can be converted into an additional term of the inlet flow rate, and the reaction time can be set as the computation time for transient solutions. Figure 1 Taking the fixed-bed reactor 100 as an example, if the reaction liquid enters through the reaction liquid inlet 102, flows through the cobalt metal fixed bed 104 after passing through the feed distributor 105, then the inlet velocity, inlet temperature, and inlet reactant concentration can be set at the reaction liquid inlet 102, the outlet pressure or free flow condition can be set at the reaction liquid outlet 103, and a non-slip wall condition can be set on the inner wall of the reactor shell 101. Furthermore, an adiabatic wall condition or a wall heat transfer condition with a given heat transfer coefficient can be set according to the actual insulation or heat transfer state. If the temperature sensor 109 and pressure sensor 110 can provide on-site data, the detected bed inlet temperature, outlet pressure, or bed pressure drop can be used as model correction parameters to ensure that the initial state of the computational domain corresponds to the current operating state.

[0132] In one specific implementation, the finite volume method can be used to discretize the cobalt fixed-bed reaction zone. Each discretized computational unit stores local porosity, permeability parameters, inertial drag parameters, effective reaction surface area, local temperature, local reactant concentration, and local cobalt acetate formation concentration. At the start of the solution process, initial velocity, pressure, temperature, and concentration fields can be assigned based on the reaction operating parameters. The initial velocity field can be estimated from the inlet flow rate and bed cross-sectional area; the initial pressure field can be set according to a linear distribution of outlet pressure and empirical bed pressure drop; the initial temperature field can be the inlet temperature or the bed detection temperature; and the initial concentration field can be the inlet concentration of the reaction solution or the calculation result from the previous operating time. Subsequently, within the same iteration step, the velocity and pressure solutions, component transport solutions, energy equation solutions, and local reaction rate updates are completed. The porous media model provides drag terms related to local permeability and inertial drag in the momentum equation. The component transport model calculates the supply of reactants to the effective surface area of ​​cobalt metal based on the current local flow rate and concentration gradient. The reaction kinetics model updates the local reaction rate of cobalt metal to cobalt acetate based on the current effective reaction surface area, reactant concentration, and temperature. The heat transfer model then corrects the heat source term based on the updated local reaction rate. The solution is not a simple superposition of the calculations of the four sub-models, but rather the flow rate, concentration, temperature, and reaction rate are mutually transferred in each iteration step until the field variables in the computational domain reach stability or the set reaction time is reached.

[0133] During synchronous iteration, the local mass transfer coefficient can be updated based on the local flow rate, reaction liquid diffusion coefficient, and equivalent cobalt particle size of the current iteration step. For example, when the local flow rate increases within a computational unit, the liquid phase renewal capability is enhanced, and the local mass transfer coefficient increases accordingly. When the local flow rate is low and the reactant concentration decreases, although the computational unit may have a large effective reaction surface area, insufficient reactant supply will limit the local reaction rate. After updating the local reaction rate, the reactant consumption, cobalt acetate generation, and local heat source term can be corrected synchronously, and the corrected concentration field and temperature field are returned to the next iteration step for continued calculation. For steady-state conditions, a pseudo-transient approach can be used to iterate until convergence; for conditions requiring observation of changes in the operating process, an actual time step can be set for transient advancement. As an example, the time step can be 1 to 30 seconds, and the total calculation time can be 30 minutes to 4 hours; for rapid bed changes or prediction calculations after feeding, a smaller time step, such as 1 to 5 seconds, can be used; for routine simulations approaching stable operation, 10 to 30 seconds can be used. The values ​​above are merely examples for ease of implementation; specific values ​​can be adjusted based on reactor volume, bed height, circulation flow rate, and required calculation accuracy.

[0134] The preset convergence condition can be determined based on the magnitude of the field variable changes and the conservation error. As one possible implementation, the maximum relative changes in velocity, temperature, reactant concentration, and local reaction rate within all computational units can be calculated between adjacent iteration steps. The preset convergence condition is considered met when the maximum relative change in velocity is no greater than 0.5%, the maximum change in temperature is no greater than 0.2°C, the maximum relative change in reactant concentration is no greater than 0.5%, and the maximum relative change in local reaction rate is no greater than 1%, and these conditions are met for five consecutive iteration steps. For implementations with high numerical stability requirements, the mass conservation error between the computational domain inlet and outlet can also be required to be no greater than 0.5%, the energy conservation error no greater than 1%, and the component conservation error no greater than 1%. The aforementioned thresholds can be obtained through empty bed hydraulic testing, stable operation batch data, or laboratory pilot-scale data. For example, a continuous operating range of a fixed-bed reactor with outlet cobalt acetate concentration fluctuation of less than ±1%, bed pressure drop fluctuation of less than ±2%, and bed temperature fluctuation of less than ±0.5℃ can be selected as a stable sample, and the variable fluctuation range of the model iteration results within this range can be used as a reference range for the convergence threshold.

[0135] When the preset convergence conditions are met, the spatial distribution results of velocity, pressure, temperature, reactant concentration, cobalt acetate formation concentration, residence time, and local reaction rate corresponding to each computational unit can be used as reaction data. This reaction data retains the spatial position of each computational unit in the fixed-bed reaction zone, thus not only obtaining macroscopic results such as outlet concentration or overall pressure drop, but also locating the reaction state of different regions within the bed. For continuous operation scenarios, the spatial distribution results of the bed that met the convergence conditions at the previous moment can also be used as the initial field for the next moment; for conditions with significant cobalt consumption, the amount of cobalt consumption can be accumulated based on the local reaction rate, and the effective reaction surface area or porosity of the corresponding computational unit can be updated in the next round of calculation. The resulting reaction data can be used for subsequent identification of uneven reaction regions, determination of cobalt feeding requirements, and generation of fixed-bed reaction flow field control schemes.

[0136] Next, we will further elaborate on the technical content of the method of this application regarding the identification of regions with uneven reactions.

[0137] In one example, the method for identifying the uneven reaction region includes:

[0138] S2.1: Spatial gridding is performed on the reaction data to obtain the flow rate, pressure, temperature, reactant concentration, cobalt acetate concentration and local reaction rate values ​​corresponding to each calculation unit;

[0139] Specifically, the reaction data obtained from numerical simulations are typically represented as continuous spatial fields or discrete grid fields. Identifying regions of inhomogeneity requires mapping these field quantities to specific spatial locations within a cobalt-fixed bed. In a cobalt-fixed bed reaction zone, within the same height section, there may simultaneously exist low-resistivity channels formed by large chunks of cobalt, high-resistivity regions formed by small particle accumulations, and localized low-contact regions formed by sheet-like cobalt obstruction. If only average bed velocity, average temperature, or outlet concentration is used for judgment, it is difficult to determine whether the abnormal state is actually located in the upper, middle, or lower part of the bed, or near the wall. Therefore, before identification, it is necessary to map the velocity field, pressure field, temperature field, reactant concentration field, cobalt acetate formation concentration field, and local reaction rate field to the same spatial grid, enabling comparison of different physical quantities at the same computational unit scale.

[0140] In this embodiment, spatial gridding can be used to divide the cobalt fixed-bed reaction zone along the axial and radial directions of the fixed-bed reactor, or along the main flow direction of the reaction liquid and the cross-sectional direction of the bed. For a vertical cylindrical fixed-bed reactor, the bed height direction can be divided into several axial layers, and the bed radius direction can be divided into several radial rings. Then, each axial layer and radial ring intersect to form multiple calculation units. For reactors with complex local structures, a denser grid can be used below the feed distributor, in the feed inlet area, near the bed support, or near the wall, while a sparser grid can be used in the relatively stable flow area in the middle of the bed.

[0141] As one possible implementation method, the gridding process can refer to the aforementioned computational unit format, so that the computational units formed in the model building stage are consistent with the computational units used in the reaction unevenness identification stage.

[0142] S2.2: Based on the flow rate, reactant concentration, temperature and local reaction rate values ​​corresponding to each calculation unit, calculate the flow field deviation index, mass transfer deviation index, temperature deviation index and reaction rate deviation index respectively;

[0143] Specifically, the differences in reactions within a cobalt fixed bed are often not caused by a single variable. A high flow rate in a certain region may indicate that the reaction liquid passes through too quickly and does not stay sufficiently; a low reactant concentration in a certain region may indicate that the reactants are being consumed rapidly in a localized area or that the liquid phase supply is insufficient; a temperature deviation from the target temperature in a certain region may reflect concentrated local heat release or insufficient heat transfer in the liquid phase; and an excessively low or high local reaction rate in a certain region corresponds to insufficient effective utilization of cobalt or concentrated local reactions. Converting these variables into deviation indicators can avoid the distortion of judgments caused by different dimensions and ranges of variation when directly comparing original physical quantities.

[0144] In this embodiment, the flow field deviation index can be determined based on the degree of deviation of the flow velocity value within the calculation unit from the average flow velocity of the same bed cross section, the flow velocity of adjacent calculation units, or the target residence velocity range. For the transition region near the inlet distributor or outlet support in a fixed-bed reactor, a separate reference flow velocity can be set to avoid the inlet jet or outlet contraction affecting the judgment of the normal bed region.

[0145] As an optional implementation, when the flow velocity of a certain calculation unit is more than 1.5 times the median flow velocity of the same axial section, the calculation unit can be considered to have a tendency to deviate from the flow velocity; when the flow velocity is less than 0.4 times the median flow velocity of the same axial section, it can be considered to have a tendency to stagnate. The flow field deviation index can be normalized according to the degree of deviation from the target flow velocity range, so that the value is smaller when the flow velocity is closer to the reasonable range, and larger when the flow velocity deviates more from the reasonable range.

[0146] Furthermore, the mass transfer deviation index can be determined based on reactant concentration, reactant concentration gradient, and reactant renewal between adjacent computational units. For the cobalt acetate preparation process from metallic cobalt, a low reactant concentration does not necessarily indicate an anomaly, as the normal reaction itself consumes reactants; it is necessary to consider whether the location still has a high effective reaction surface area and the reaction liquid renewal capacity for judgment. As an implementation method, the reactant concentration at the same axial position or in the adjacent upstream region can be used as a reference. When the reactant concentration of a certain computational unit is lower than 70% to 85% of the reference concentration, and a high effective reaction surface area still exists near that computational unit, its mass transfer deviation can be increased. If the decrease in reactant concentration is accompanied by an increase in the cobalt acetate formation concentration and the local flow rate is within a reasonable range, the mass transfer deviation can be reduced to avoid misjudging normal reaction consumption as insufficient mass transfer. Temperature deviation index can be determined based on the deviation of the calculated unit temperature from the target reaction temperature range. For example, when the target temperature is 70℃ to 85℃, any temperature below 70℃ or above 85℃ can be counted as temperature deviation. For areas with large temperature gradients, the temperature difference between adjacent units can also be used as an auxiliary criterion. For example, if the temperature difference between adjacent units exceeds 2℃ to 5℃, it indicates uneven heat transfer in that area, and the temperature deviation level should be increased. Reaction rate deviation index can be determined based on the local reaction rate relative to the target reaction rate range, the median reaction rate of the bed, and the expected reaction rate corresponding to the effective reaction surface area. When the local reaction rate is significantly lower than the expected level, it usually corresponds to insufficient cobalt metallic reactivity or insufficient liquid phase contact. When the local reaction rate is significantly higher than the expected level, it usually corresponds to concentrated exposure of fresh surface or excessive concentration of the reaction liquid in a local area.

[0147] It should be noted that the aforementioned numerical examples can be determined by referring to historical data of the fixed-bed reactor under stable operating conditions, cold-state flow test data, tracer residence time test data, small-scale reaction calibration data, and the allowable fluctuation range of the process, and should not be construed as the sole limitation of this application. For the cobalt fixed-bed reaction zone, the bed itself exhibits normal spatial fluctuations caused by the particle size of the packing material, local porosity, and inlet distribution structure. Therefore, the threshold setting should not be based solely on absolute flow rate, absolute concentration, or absolute temperature, but rather on statistical reference values ​​within the same axial cross-section or adjacent local areas. Using the median flow rate within the same axial cross-section as a reference value can reduce the bias of the average value caused by the inlet region, outlet region, and local high-speed channels, making the identification of flow deviation and stagnation closer to the true local distribution state of the bed.

[0148] S2.3: The flow field deviation index, the mass transfer deviation index, the temperature deviation index, and the reaction rate deviation index are weighted to obtain the reaction uniformity evaluation value of each calculation unit;

[0149] Specifically, relying solely on a single type of deviation indicator can easily lead to biased judgments. For example, when identifying deviations based solely on flow rate, high-speed regions may not necessarily have insufficient reaction, as they could still contribute significantly to the reaction if the effective reaction surface area is large and the reactant concentration is sufficient. Similarly, when identifying deviations based solely on local reaction rates, low-reaction-rate regions may not necessarily require refeeding; the reactants in that region may simply have been consumed normally by the upstream region. Unifying deviations in flow field, mass transfer, temperature, and reaction rate into a single reaction uniformity evaluation value allows each calculation unit to simultaneously consider whether the reaction solution can reach the target area, whether reactants can be supplied, whether heat can migrate stably, and whether cobalt metal effectively participates in the reaction.

[0150] In this embodiment, before weighting, various deviation indicators can be normalized to fall within the same numerical range, such as 0 to 1 or 0 to 100. During normalization, lower values ​​can be assigned to cases with smaller deviations from the target range or within the allowable fluctuation range, while higher values ​​can be assigned to cases that significantly exceed the allowable range or simultaneously exhibit extreme values ​​and large gradients. The weighting coefficients can be determined based on the sensitivity of the cobalt acetate preparation process to different risks. As an example, during the normal stable production stage, the weights for flow field deviation, mass transfer deviation, temperature deviation, and reaction rate deviation can be set to 0.25, 0.30, 0.20, and 0.25, respectively. Under conditions where the reaction solution concentration is low or insufficient feed is likely to occur in the later stages of the bed, the weight for mass transfer deviation can be increased to 0.35 to 0.40. Under conditions where temperature control is more sensitive or local temperature rise in the bed is more likely to affect product stability, the weight for temperature deviation can be increased to approximately 0.30. The aforementioned weights can be adjusted based on historical batch data. For example, multiple stable operating batches and multiple abnormal operating batches can be selected, and the correspondence between the abnormal area identification results and the actual pressure drop fluctuations, outlet cobalt acetate concentration fluctuations, temperature alarm records, or shutdown inspection results under different weight combinations can be compared to determine the weight combination that is more suitable for the current reactor.

[0151] Furthermore, the reaction uniformity evaluation value can be understood as the overall degree to which the computational unit deviates from the target reaction state. For the case of normalization to 0 to 100, the closer the evaluation value is to 0, the closer the flow, mass transfer, temperature and reaction rate of the computational unit are to the target uniform state; the higher the evaluation value, the more likely the computational unit is to have reaction non-uniformity.

[0152] S2.4: Compare the reaction uniformity evaluation value with a preset uniformity threshold, and determine the calculation unit whose reaction uniformity evaluation value is greater than the preset uniformity threshold as an abnormal calculation unit;

[0153] Specifically, a preset uniformity threshold is used to distinguish between normal spatial fluctuations and reaction inhomogeneities requiring attention. In actual operation, a cobalt fixed-bed reactor cannot be completely uniform; variations in the field near the inlet, outlet, wall, or large localized areas of cobalt are common phenomena in fixed-bed reactions. If the threshold is set too low, many normal fluctuations will be flagged as abnormal, leading to over-adjustment of subsequent control schemes. If the threshold is set too high, areas with insufficient effective reaction area, limited mass transfer, or excessively strong reactions may be missed. Therefore, the preset uniformity threshold should not be arbitrarily given based on experience but should be determined in conjunction with stable operating samples, product quality requirements, and safe operating boundaries.

[0154] In this embodiment, the preset uniformity threshold can be determined using historical stable batch data. Specifically, several operating intervals of the fixed-bed reactor with outlet cobalt acetate concentration fluctuations not exceeding ±1.5%, bed pressure drop fluctuations not exceeding ±3%, and representative bed temperature fluctuations not exceeding ±0.5℃ can be selected as stable samples. The reaction data corresponding to these stable samples are substituted into the above identification process to obtain the distribution of reaction uniformity evaluation values ​​for each calculation unit. The upper limit of the evaluation values ​​of 95% of the calculation units in the stable samples can be used as the base value of the threshold, and then a margin of 5% to 15% can be added according to the process safety margin. As an example, when the reaction uniformity evaluation value adopts a scale of 0 to 100, if the evaluation value of 95% of the calculation units in the stable samples does not exceed 52, the preset uniformity threshold can be set to 55 to 60; for operating conditions with higher requirements for product concentration stability, 55 can be used; for large-size reactors with large natural differences in bed packing, 60 to 65 can be used. If a scale of 0 to 1 is adopted, the above threshold can be set to 0.55 to 0.65.

[0155] S2.5: Merge the connected regions of spatially adjacent abnormal calculation units to obtain the reaction non-uniform region within the cobalt metal bed;

[0156] Specifically, anomalies in fixed-bed reactors typically manifest as spatially continuous regions rather than isolated individual computational units. Excessive localized packing of cobalt metal can create low-flow regions, and insufficient effective reaction area for cobalt metal is usually distributed along local heights or radial ranges within the bed. Localized obstructions or short-circuit channels formed after feed also affect the continuous space. Therefore, directly using anomaly computational units as the final result leads to overly fragmented identification, making it difficult to determine control locations, feed areas, or flow adjustment targets. By merging connected regions, spatially adjacent computational units with the same or similar anomaly characteristics can be organized into reaction heterogeneity regions with defined locations, boundaries, and extents.

[0157] In this embodiment, spatial adjacency can be determined based on the shared boundary relationships of computational units. For regular meshes, computational units that are coplanar with the target anomalous computational unit in the axial or radial direction can be considered directly adjacent; for unstructured meshes, computational units that share a face or whose center distance is less than a preset adjacency distance can be considered adjacent. The preset adjacency distance can be 1.2 to 1.5 times the average mesh size of the corresponding region. For three-dimensional computational domains, the six-neighbor connectivity relationship of shared faces can be preferentially adopted; when the bed mesh is coarse or the anomalous region extends significantly along the oblique direction, computational units that share edges or corners can also be allowed to participate in merging, but this needs to be confirmed by considering whether the dominant deviation types of adjacent computational units are consistent.

[0158] As an example, when two adjacent anomalous calculation units mainly exhibit mass transfer deviation and reaction rate deviation, and both of their evaluation values ​​are greater than the preset uniformity threshold, they can be merged into the same reaction non-uniformity region. If one calculation unit mainly exhibits high-speed flow deviation and the other calculation unit mainly exhibits low-temperature boundary fluctuation, they can be temporarily not merged to avoid mixing anomalies of different origins into the same region.

[0159] Next, we will further elaborate on the technical content of the fixed-bed reaction flow field control scheme of this application.

[0160] It should be noted that, as mentioned above, the fixed-bed reaction flow field control scheme can include a cobalt feeding scheme, or it can include a conventional operation control scheme unrelated to cobalt feeding. Specifically, when the reaction non-uniformity region is mainly characterized by uneven flow rate distribution, insufficient local residence time, uneven reaction liquid distribution, or temperature deviation, while the effective reaction surface area, local reaction rate, and cobalt consumption in this region are still within the preset allowable range, it can be considered that the reaction non-uniformity is not caused by insufficient cobalt reactivity. In this case, there is no need to trigger the cobalt feeding process. Conventional control schemes such as feed flow rate adjustment, circulation flow rate adjustment, reaction temperature adjustment, feed distribution adjustment, or flow guide structure adjustment can be generated based on the flow field parameters, mass transfer parameters, or heat transfer parameters corresponding to the reaction non-uniformity region.

[0161] In this embodiment, the aforementioned situation where no feeding is required can be understood as the fixed-bed reactor still having sufficient effective reaction area for cobalt. Uneven reaction primarily stems from deviations in the reaction solution organization or operating conditions. For example, when the flow rate in a certain area is too high but the effective reaction surface area is not lower than a preset specific surface area threshold, the residence state can be improved by reducing the inlet local flux, adjusting the circulation flow rate, or changing the feed distribution method. When the temperature in a certain area deviates from the target range but the local reaction rate is not lower than a preset reaction rate threshold, it can be corrected by adjusting the feed temperature, cooling intensity, or circulation liquid ratio. When mass transfer in a certain area is insufficient but the cobalt consumption does not meet the feeding criterion, it can be addressed by increasing the circulation renewal capacity or adjusting the reaction solution concentration. The above-mentioned adjustments are routine operating parameter adjustments in the fixed-bed reaction process, and their specific implementation methods can be determined according to the reactor's on-site control conditions. This application will not further elaborate on this type of routine adjustment process where no feeding is required.

[0162] Furthermore, subsequent embodiments focus on how to determine the need for cobalt replenishment when the reaction heterogeneous region simultaneously exhibits insufficient effective reaction surface area, reduced local reaction rate, decreased cobalt acetate production capacity, or cobalt metal consumption reaching preset conditions, and how to predict the bed structure and reaction liquid contact state after replenishment. In other words, the following replenishment-related technical content does not exclude the existence of other conventional adjustment measures in the fixed-bed reaction flow field control scheme, but rather elaborates on the replenishment judgment and replenishment scheme update process unique to cobalt metal as a consumable solid reactant. This distinction avoids attributing all reaction heterogeneity to insufficient cobalt metal and ensures that replenishment is only triggered when the cobalt metal reactivity is indeed insufficient.

[0163] In one example, the scheme for generating a fixed-bed reaction flow field control includes:

[0164] S3.1: Based on the effective reaction surface area, local reaction rate, reactant concentration, cobalt acetate formation concentration, and cobalt metal consumption corresponding to the uneven reaction region, determine whether there is a need for cobalt metal replenishment in the uneven reaction region;

[0165] Specifically, in the preparation of cobalt acetate using a fixed-bed cobalt metal bed, uneven reaction zones do not necessarily correspond to insufficient cobalt metal. A localized decrease in reaction rate in a certain area could be due to insufficient effective reaction area resulting from the consumption of cobalt metal in that region, insufficient flow of the reaction liquid through that area, consumption of reactant concentration by upstream regions, local temperature deviation from the optimal reaction range, or even the presence of cobalt metal but its surface being obscured, preventing effective liquid-phase contact. Therefore, when determining whether to add feed, it is necessary to correlate the effective reaction surface area, local reaction rate, reactant concentration, cobalt acetate formation concentration, and cobalt metal consumption within the same spatial region, rather than relying solely on the total bed packing, outlet cobalt acetate concentration, or a single local reaction rate.

[0166] In this embodiment, the region with uneven reaction can be considered as the target area, and regional statistics can be performed on the effective reaction surface area, local reaction rate, reactant concentration, cobalt acetate concentration, and cobalt metal consumption of each calculation unit within that region. Regional statistics can be performed using either volume weighting or effective reaction surface area weighting. For cases where the cobalt metal reactivity is deemed insufficient, the effective reaction surface area is used as one of the weights, ensuring that the actual cobalt metal surface that can be contacted by the reaction liquid contributes more to the judgment result. The cobalt metal consumption can be calculated based on the cumulative result of the local reaction rate over the operating time, or it can be corrected by combining feed records, feed consumption, the increase in outlet cobalt acetate concentration, and changes in bed pressure drop.

[0167] For example, the current effective specific surface area of ​​the uneven reaction zone can be compared with the effective specific surface area of ​​the zone under initial loading conditions, or it can be compared with the effective specific surface area of ​​the same spatial zone in historically stable operating batches. When the effective specific surface area of ​​the zone decreases to 60% to 80% of the reference value, and the cumulative consumption of cobalt metal reaches 15% to 35% of the initially estimated amount of cobalt metal in the zone, it can be considered that the zone has entered a state requiring key assessment of replenishment needs. The above values ​​can be adjusted according to the form of cobalt metal raw material, bed height, reaction solution concentration, and production load. Those skilled in the art can calibrate these values ​​based on stable operating data and actual bed consumption during maintenance.

[0168] S3.2: When the effective reaction surface area in the uneven reaction region is lower than the preset surface area threshold and the local reaction rate is lower than the preset reaction rate threshold, the corresponding uneven reaction region is determined as a region with insufficient cobalt metal reaction capacity, and cobalt metal replenishment demand information is generated. The cobalt metal replenishment demand information includes at least one of the following: replenishment area, replenishment amount demand, replenishment timing demand, and target effective reaction surface area after replenishment.

[0169] Specifically, an effective reaction surface area lower than a preset specific surface area threshold indicates insufficient cobalt surface area in that region that can actually contact the reaction liquid and participate in the reaction; a local reaction rate lower than a preset reaction rate threshold indicates that the region has already shown a decrease in production capacity in the reaction results. Using both conditions together can reduce the bias caused by judging based on a single condition. For example, if only the effective reaction surface area decreases but the local reaction rate is still within the target range, it may indicate that the reaction load in that region is low or the liquid phase supply is good, and immediate replenishment is not necessary; if only the local reaction rate decreases but the effective reaction surface area is still sufficient, the low reaction rate may be due to insufficient reactant concentration, low temperature, or insufficient flow contact, in which case it is more suitable to treat it through feeding, circulation, or temperature adjustment. Only when both insufficient effective reaction interface and decreased local production capacity occur simultaneously can the region be identified as a region with insufficient cobalt reactivity, which is more in line with the actual operating logic of a cobalt-consuming fixed bed.

[0170] In this embodiment, the preset specific surface area threshold can be determined based on the initial filling state, stable operating state, or target production load. As one possible implementation, the effective reaction specific surface area of ​​a certain spatial region after initial filling and wetting can be used as an initial reference value, and the lower limit of the effective reaction specific surface area of ​​that region in a stable operating batch can be used as a process reference value. A replenishment trigger threshold can be selected between the two. For example, if the initial effective reaction specific surface area of ​​a certain reaction region is 100 units, and this region typically maintains above 75 units during historical stable operation, the preset specific surface area threshold can be set to 70 to 80 units. If the cobalt raw material is mainly in bulk form and has a low theoretical surface area, the threshold can be appropriately lowered. If the cobalt raw material is mainly in granular or flake form and the local fresh surface changes rapidly, the threshold can be appropriately increased. The preset reaction rate threshold can be determined based on the stable local reaction rate under conditions of the same temperature, the same reactant concentration, and a similar effective reaction surface area. For example, it can be taken as 50% to 70% of the stable operating reference reaction rate. When the local reaction rate is lower than this range for multiple consecutive calculation times, and the cobalt acetate formation concentration in this region is lower than that in the adjacent normal region or the target formation level, this region can be identified as a region with insufficient cobalt metallic reactivity. The above threshold can be corrected through small-scale calibration, historical batch data, online operating data, and bed inspection results after shutdown. This application does not impose a unique limitation on the specific value.

[0171] In yet another example, the scheme for generating a fixed-bed reaction flow field also includes:

[0172] S3.3: Based on the cobalt metal replenishment requirement information, generate candidate replenishment schemes, wherein the candidate replenishment schemes include at least one of the following: replenishment location, replenishment amount, particle size range, morphology type, filling density, and filling layers of cobalt metal.

[0173] Specifically, information on cobalt feed requirements typically indicates a reaction capacity gap in a region of uneven reaction, but this gap is not directly equivalent to adding the same mass of cobalt. For a cobalt fixed bed, the reaction effect after feeding depends on whether the added cobalt can fall into the target area, whether it can form a surface that can be contacted by the reaction liquid, whether it can maintain the necessary bed voids, and whether it will change the original flow path of the reaction liquid. Therefore, candidate feeding schemes need to convert the feeding area, feed quantity requirement, feeding timing requirement, and target effective reaction surface area after feeding into executable feeding parameters. The feeding location can be determined based on the axial, radial, and circumferential positions of the region with insufficient cobalt reaction capacity in the bed, and modified in conjunction with the feeding port location, feeding channel, cobalt drop path, and the morphology of the upper surface of the bed; the feed quantity can be estimated based on the difference between the target effective reaction surface area and the current effective reaction surface area, combined with the effective reaction area that can be formed per unit mass of the cobalt to be added; the particle size range and morphology are used to adjust the relationship between the increased reaction area and the local flow voids.

[0174] In this embodiment, if the area with insufficient reactivity is located in the upper part of the bed and close to the projection range of the feed inlet, the feed position can be set above or in the adjacent upstream area of ​​this area, so that the added cobalt can cover the target area after gravity accumulation. If the area with insufficient reactivity is located in the middle or lower part of the bed or close to the wall, the feed position can be set according to the reachable path between the feed inlet and the target area, so that the area can be gradually affected by subsequent reaction liquid flushing, local sedimentation, or layered filling. The particle size range of the added cobalt can be determined according to the porosity of the target area and the current flow state. For example, when the porosity of the target area is large and the reaction liquid has a tendency to pass through quickly, medium-sized or different particle size combinations of cobalt can be selected to increase the effective contact area and reduce local large channels. When the porosity of the target area is small or there is already a tendency to compaction, a large number of fine particles can be avoided from being added in a concentrated manner. Instead, blocky or larger cobalt particles can be used in combination with a small number of small particles to maintain the flow space for the reaction liquid to enter the feed area. The morphology can include block, granular, flake and irregular agglomerate. Among them, blocky cobalt is more suitable for maintaining bed pore support, granular cobalt is more suitable for supplementing effective surface area, and flake or irregular agglomerate cobalt can be used in combination with the filling layer to control its shielding and pressing state.

[0175] Furthermore, the filling density and filling level can be determined jointly based on the target effective reaction surface area and target porosity after feeding.

[0176] As an optional implementation method, the theoretical feed amount can be determined first based on the target effective reaction surface area. Then, adjustments can be made based on the potential for compression, obstruction, and unreachable liquid phase surfaces after the feed cobalt is actually deposited. The adjustment amount can be 5% to 20% of the theoretical feed amount. For feed schemes with a high proportion of flake or irregular cobalt, a higher obstruction correction value can be used; for granular feed schemes with more uniform particle size and better surface accessibility, a lower obstruction correction value can be used. The packing layers can be determined based on the reaction liquid flow direction. For example, in a top-in, bottom-out fixed bed, relatively large-diameter cobalt particles can be arranged upstream of the feed area to maintain liquid phase distribution channels, cobalt particles with a high effective surface area can be arranged in the target reaction area to restore reaction capacity, and excessive concentration of fine particles in the downstream area can be avoided to reduce local pressure drop increases. The resulting candidate feed scheme is not simply a list of feed weights, but a combination of feed location, feed shape, feed particle size, local stacking state, and target reaction area.

[0177] S3.4: Based on the candidate feeding scheme, update the porosity distribution and effective reaction specific surface area distribution in the cobalt metal fixed bed reaction zone to obtain the predicted bed structure after feeding;

[0178] Specifically, candidate feeding schemes need to be transformed into changes in bed structure before implementation. After cobalt is added to the fixed bed, it is not uniformly dispersed throughout the entire bed. Instead, it is affected by the location of the feeding port, the particle size and morphology of the added cobalt, the current bed height, local voids, the original cobalt packing state, and the direction of the reaction liquid flow, resulting in the formation of new packing structures in certain areas. These packing structures alter the local solid volume fraction and porosity, and also change the proportion of the newly added cobalt surface that can be contacted by the reaction liquid.

[0179] In this embodiment, the effective range of the added cobalt metal in the bed can be determined based on the feeding location. This effective range can be determined by the projection area of ​​the feeding port, the diffusion radius of the cobalt metal, the slope of the upper surface of the bed, and the feeding amount. For a vertical fixed-bed reactor, the added cobalt metal usually forms a local accumulation zone below the feeding port; when the feeding amount is small and the particle size is large, the effective range can be mainly concentrated in several calculation units below the feeding port; when the feeding amount is large or granular cobalt metal is used, the effective range can extend outwards along the upper surface of the bed. As an example, the feeding effective radius can be determined based on the feeding port diameter, the natural angle of repose of the cobalt metal, and the distance between the upper surface of the bed and the feeding port; for a reactor with a diameter of 0.8 m, when the single feeding amount is small, the effective range can be set to 10% to 25% of the reactor radius; when the feeding amount is large or the feeding position is eccentric, it can be expanded to 25% to 40% of the reactor radius. The above range can be corrected through cold feeding tests, historical pressure drop changes after feeding, or shutdown inspection results.

[0180] S3.5: Input the predicted bed structure after feeding into the numerical simulation model to calculate the predicted reaction data after feeding corresponding to the candidate feeding scheme;

[0181] Specifically, after the predicted bed structure is formed following feeding, it needs to be used as the new bed loading state input into the numerical simulation model. This input is not a newly established initial model unrelated to the current operation, but rather, based on the existing reactor geometry, reaction liquid properties, and current operating parameters, it replaces or modifies the porosity distribution, effective reaction surface area distribution, permeability parameters, inertial drag parameters, and local reaction parameters of the corresponding computational units within the cobalt fixed-bed reaction zone.

[0182] In this embodiment, the predicted reaction data after feeding can be obtained through steady-state prediction or transient prediction. For cases where the feeding amount is small, the feeding area is located in the upper part of the bed, and it is not expected to cause significant bed rearrangement, steady-state prediction can be used. The predicted bed structure after feeding is input into the numerical simulation model and iterated until stable to obtain the velocity field, pressure field, temperature field, reactant concentration field, cobalt acetate formation concentration field, and local reaction rate field after feeding. For cases where the feeding amount is large, the particle size of the fed cobalt metal changes significantly, or the feeding position is close to the original main flow path, transient prediction can be used. The reaction data obtained before feeding convergence is used as the initial state, and the predicted bed structure after feeding is used as the structure update state to calculate the change in reaction data within a preset time window after feeding. The preset time window can be determined based on the bed liquid residence time, circulation pipeline renewal time, and production control response time. For example, it can range from 10 to 120 minutes. For reactors with a large circulation volume and small bed volume, it can range from 10 to 30 minutes; for reactors with a large bed height or slow circulation renewal, it can range from 60 to 120 minutes. The time step can range from 1 to 30 seconds. A smaller time step can be used initially after feeding, and can be appropriately increased after nearing stability.

[0183] S3.6: Determine the degree of feeding disturbance corresponding to the candidate feeding scheme based on the predicted reaction data after feeding, and determine the fixed bed reaction flow field control scheme based on the degree of feeding disturbance;

[0184] Specifically, the degree of feed disturbance described here is not solely determined by changes in flow rate, temperature, or concentration before and after feed, but rather by whether the newly added effective cobalt reaction area created by the candidate feed scheme can be continuously and effectively contacted by the reaction liquid. The core of cobalt feed is not simply increasing the bed solids, but rather increasing the effective reaction interface capable of participating in cobalt acetate formation. If the added cobalt causes localized compaction, obstruction, or flow around the reaction liquid, even if the feed amount meets the requirements, the newly added surface may still be unusable by the reaction liquid. Accordingly, the degree of feed disturbance can be understood as the extent to which the newly added effective reaction area deviates from the state of continuous coverage by the reaction liquid.

[0185] In one example, the method for determining the degree of feeding disturbance includes:

[0186] Based on the candidate feeding scheme, determine the newly added effective reaction area region of metallic cobalt in the predicted bed structure after feeding;

[0187] Based on the predicted reaction data after feeding, determine the area within the cobalt fixed-bed reaction zone that can be continuously covered by the reaction liquid through which the reaction liquid continuously flows.

[0188] Determine the overlap and connectivity between the newly added effective reaction area of ​​cobalt metal and the continuously covered area of ​​the reaction liquid;

[0189] Based on the overlapping state and the connectivity state, the effective contact degree of the newly added reaction area corresponding to the candidate feeding scheme is determined, and the effective contact degree of the newly added reaction area is used as the feeding disturbance degree corresponding to the candidate feeding scheme.

[0190] In this embodiment, the newly added effective reaction area of ​​cobalt can be directly determined based on the predicted bed structure after feeding. Specifically, the location, stacking range, and morphological distribution of the fed cobalt in the calculation unit corresponding to the candidate feeding scheme can be marked, and the calculation unit where the effective reaction surface area increases before and after feeding is determined as the unit containing the newly added reaction area. For this newly added portion, the theoretical geometric surface area of ​​the fed cobalt is not directly used; instead, it is corrected by considering the pressing relationship between the fed cobalt, the shielding relationship with the original cobalt, the local packing density, and the reachable state of the reaction liquid. For example, if the fed sheet-like cobalt is locally adhered and stacked, its adhered surface can be excluded from the newly added effective reaction area; if the fed granular cobalt enters the gaps between the original bulk cobalt and can be flowed through by the liquid phase, its exposed surface can be included in the newly added effective reaction area; although the total surface area of ​​the fed bulk cobalt is low, if a continuous liquid phase channel is maintained around it, its contactable surface can still be considered as an effective newly added area.

[0191] The continuous coverage area of ​​the reaction liquid can be determined within the cobalt fixed-bed reaction zone based on the predicted reaction data after feeding. In practice, the calculation units can be considered as nodes with spatial adjacency, and the existence of liquid-phase channels through which the reaction liquid can pass between adjacent calculation units is used as a connectivity condition. Based on this, starting from the reaction liquid inlet side or the downstream area of ​​the feed distributor, a search is conducted along the main flow direction of the reaction liquid to find continuously reachable calculation units. Calculation units that can be continuously flowed through by the reaction liquid should typically simultaneously satisfy the following conditions: the local liquid phase velocity is not in a near-stagnant state, the reactant concentration is not lower than the supply lower limit for maintaining the cobalt surface reaction, and there is a continuous adjacent path between them and the upstream supply area. The aforementioned supply lower limit can be determined based on the concentration distribution of the effective reaction area in a stable operating batch. For example, an operating range with stable outlet cobalt acetate concentration, stable bed pressure drop, and no temperature alarm can be selected, and the low-end concentration level corresponding to the calculation units participating in the normal reaction within this range can be used as a reference. The local liquid phase velocity lower limit can also be determined by cold-state flow tests or stable operating condition simulation results, for example, using the low-end range of the velocity distribution in the normal flow area within the same axial cross-section as a judgment criterion.

[0192] After determining the effective reaction area region of newly added cobalt and the continuous coverage area of ​​the reaction liquid, their spatial relationship can be compared within the same computational grid. If the cell containing the effective reaction area of ​​newly added cobalt is also within the continuous coverage area of ​​the reaction liquid, and this area can be connected to the upstream supply path and the downstream drainage path through adjacent computational cells, then the corresponding newly added reaction area can be considered as an effective contact area. If the newly added reaction area is concentrated outside the continuous coverage area of ​​the reaction liquid, or although it partially overlaps, it only forms isolated contact points and cannot form a continuous liquid phase coverage along the bed flow direction, then it indicates that the newly added cobalt surface formed by the candidate feeding scheme has a risk of being unusable or having low utilization. In specific judgment, the part of the newly added effective reaction area within the continuous coverage area of ​​the reaction liquid can be regarded as the effective contact part, and the part not continuously covered can be regarded as the insufficient contact part. The degree of effective contact of the newly added reaction area can be determined by combining whether the insufficient contact part is located in the core position of the feeding target area, whether it forms a continuous shielding area, and whether it blocks the original liquid phase channel. The effective contact level can be used as the feeding disturbance level of the candidate feeding scheme; its value or level can be calibrated by the historical operation results after feeding, for example, the recovery time of cobalt acetate concentration at the outlet after feeding, the change in bed pressure drop and the temperature fluctuation in the feeding area can be used as the basis for screening qualified feeding samples, and then the range required for continuous coverage of the newly added reaction area can be determined in reverse.

[0193] In yet another example, the scheme for generating a fixed-bed reaction flow field also includes:

[0194] The degree of material feeding disturbance is compared with the preset disturbance allowable conditions;

[0195] When the degree of feeding disturbance does not meet the preset disturbance allowable condition, at least one of the following in the candidate feeding scheme is updated according to the overlap and connection state between the newly added effective reaction area of ​​metallic cobalt and the continuous coverage area of ​​the reaction liquid: feeding position, feeding amount, particle size range, morphology type, filling density and filling layer.

[0196] Based on the updated candidate feeding schemes, the predicted bed structure after feeding is redefined, and the predicted reaction data after feeding corresponding to the updated candidate feeding schemes are recalculated.

[0197] Based on the recalculated predicted reaction data after feeding, the feeding disturbance level corresponding to the updated candidate feeding scheme is determined again until the feeding disturbance level meets the preset disturbance allowance condition.

[0198] The candidate feeding scheme that meets the preset disturbance allowance condition is determined as the fixed bed reaction flow field control scheme.

[0199] The preset disturbance allowance condition can be understood as follows: after the candidate feeding scheme is executed, the newly added effective reaction area of ​​cobalt metal can be located within the continuous coverage area of ​​the reaction liquid, and this newly added effective reaction area will not form a large area of ​​unusable surface due to local accumulation, obstruction, or compaction after feeding, nor will it cause the feeding area to lose its connection with the original reaction liquid flow path. In other words, the preset disturbance allowance condition is not simply a restriction on changes in flow rate, temperature, or concentration before and after feeding, but rather a constraint on whether the newly added cobalt metal after feeding truly becomes an effective reaction interface that can be continuously contacted by the reaction liquid and participate in the formation of cobalt acetate.

[0200] Specifically, the proportion of the newly added effective cobalt reaction area located within the continuous coverage area of ​​the reaction liquid should not be less than a preset coverage ratio, and the newly added effective cobalt reaction area should be connected to the continuous flow path on the inlet side of the reaction liquid. As one possible implementation, the preset coverage ratio can be set to 70% to 90%, for example, 80%. When, in the predicted structure after feeding, more than 80% of the newly added effective cobalt reaction area falls within the continuous coverage area of ​​the reaction liquid, and this newly added effective cobalt reaction area can be connected to the upstream supply area through continuously adjacent computing units, the feeding disturbance level can be considered to meet the preset disturbance allowable condition. If the coverage ratio of the newly added effective reaction area is less than 70%, or the newly added effective reaction area is mainly concentrated in locally compacted areas that are difficult for the reaction liquid to reach, even if the feeding amount meets the requirements, the candidate feeding scheme can be considered not to meet the preset disturbance allowable condition.

[0201] In one example, this application embodiment provides a fixed-bed reaction process coupling control system based on industrial cloud computing, the system comprising:

[0202] The numerical simulation module is used to acquire the physical properties and operating parameters of the reaction liquid in the fixed-bed reactor, and to obtain the reaction data in the fixed-bed reactor by multi-field coupling solution through a preset numerical simulation model. The numerical simulation model includes a porous media model for characterizing the flow resistance of the reaction liquid in the cobalt bed, a component transfer model for characterizing the mass transfer process between cobalt and the reaction liquid, a heat transfer model for characterizing the exothermic or endothermic process of the reaction, and a reaction kinetic model for characterizing the process of cobalt forming cobalt acetate.

[0203] The reaction state identification module is used to identify the reaction uneven region in the cobalt metal bed based on the reaction data, and to determine whether there is a need for cobalt metal replenishment in the reaction uneven region based on the effective reaction specific surface area, local reaction rate, reactant concentration value, cobalt acetate generation concentration value and cobalt metal consumption corresponding to the reaction uneven region.

[0204] The control scheme generation module is used to generate candidate feeding schemes when there is a demand for cobalt metal feeding, update the porosity distribution and effective reaction specific surface area distribution in the cobalt metal fixed bed reaction zone based on the candidate feeding schemes, obtain the predicted bed structure after feeding, and input the predicted bed structure after feeding into the numerical simulation model to obtain predicted reaction data after feeding; and determine the feeding disturbance degree corresponding to the candidate feeding scheme based on the predicted reaction data after feeding, and update the candidate feeding scheme based on the feeding disturbance degree until the updated candidate feeding scheme meets the preset disturbance allowable condition, and determine the candidate feeding scheme that meets the preset disturbance allowable condition as the fixed bed reaction flow field control scheme.

[0205] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for coupled control of a fixed-bed reaction process based on industrial cloud computing, characterized in that, The method includes: The physical properties and operating parameters of the reaction liquid in the fixed-bed reactor are obtained, and the reaction data in the fixed-bed reactor are obtained by multi-field coupling solution through a preset numerical simulation model. The numerical simulation model includes a porous media model for characterizing the flow resistance of the reaction liquid in the cobalt bed, a component transfer model for characterizing the mass transfer process between cobalt and the reaction liquid, a heat transfer model for characterizing the exothermic or endothermic process of the reaction, and a reaction kinetic model for characterizing the process of cobalt forming cobalt acetate. Based on the reaction data, identify regions of uneven reaction within the cobalt metal bed; Based on the location and extent of the uneven reaction region and its corresponding flow field parameters, mass transfer parameters, heat transfer parameters, and local reaction rate parameters, a fixed-bed reaction flow field control scheme is generated.

2. The fixed-bed reaction process coupling control method based on industrial cloud computing according to claim 1, characterized in that, The numerical simulation model is constructed in the following ways: The basic parameters of the fixed-bed reactor used for the reaction of metallic cobalt to prepare cobalt acetate are obtained, including reactor structural parameters and metallic cobalt bed packing parameters; Based on the aforementioned cobalt bed loading parameters, the morphological distribution, particle size distribution, bulk density distribution, porosity distribution, and effective reaction specific surface area distribution of the cobalt raw material in the fixed bed are determined. A numerical simulation model of the cobalt fixed-bed reaction zone is constructed based on the reactor structural parameters, the porosity distribution, and the effective reaction surface area distribution.

3. The fixed-bed reaction process coupling control method based on industrial cloud computing according to claim 2, characterized in that, The method for constructing the numerical simulation model also includes: The cobalt fixed-bed reaction zone is divided into multiple computational units; For each calculation unit, the permeability parameters and inertial drag parameters of the calculation unit are determined according to the porosity, cobalt particle size, cobalt morphology parameters and packing density of the calculation unit, and the porous medium model is established based on the permeability parameters and inertial drag parameters. For each calculation unit, the local mass transfer coefficient between the cobalt metal surface and the reaction liquid is determined based on the effective reaction surface area, reaction liquid flow rate and reactant concentration corresponding to the calculation unit, and the component transfer model is established based on the local mass transfer coefficient. For each calculation unit, the local heat source term of the calculation unit is determined according to the effective reaction specific surface area, local reaction rate and reaction heat parameters corresponding to the calculation unit, and the heat transfer model is established based on the local heat source term. For each calculation unit, the local reaction rate of cobalt metal to cobalt acetate is determined based on the effective reaction surface area, reactant concentration and temperature corresponding to the calculation unit, and the reaction kinetic model is established based on the local reaction rate.

4. The fixed-bed reaction process coupling control method based on industrial cloud computing according to claim 1, characterized in that, The multi-field coupling solution includes: The physical properties of the reaction liquid are used as boundary conditions, and the reaction operating parameters are used as initial conditions input into the numerical simulation model. Based on the coupling relationship between the porous medium model, the component transport model, the heat transfer model and the reaction kinetics model, the reaction process of reaction liquid flow, reactant mass transfer, heat transfer and the formation of cobalt acetate from metal cobalt is solved synchronously and iteratively in the metal cobalt fixed bed reaction zone. During the synchronous iterative solution process, the local mass transfer coefficient and local reaction rate are updated based on the local flow rate, local temperature and local reactant concentration obtained in the current iteration step. The reactant consumption, cobalt acetate production and local heat source terms are corrected based on the updated local reaction rate until the changes in flow rate, temperature, concentration and local reaction rate between adjacent iteration steps meet the preset convergence conditions. The bed spatial distribution result obtained when the preset convergence condition is met is used as the reaction data.

5. The fixed-bed reaction process coupling control method based on industrial cloud computing according to claim 4, characterized in that, The methods for identifying uneven reaction regions include: The reaction data is spatially gridded to obtain the flow rate, pressure, temperature, reactant concentration, cobalt acetate concentration, and local reaction rate values ​​corresponding to each calculation unit. Based on the flow rate, reactant concentration, temperature, and local reaction rate values ​​corresponding to each calculation unit, the flow field deviation index, mass transfer deviation index, temperature deviation index, and reaction rate deviation index are calculated respectively. The flow field deviation index, the mass transfer deviation index, the temperature deviation index, and the reaction rate deviation index are weighted to obtain the reaction uniformity evaluation value of each calculation unit. The reaction uniformity evaluation value is compared with a preset uniformity threshold, and the calculation unit whose reaction uniformity evaluation value is greater than the preset uniformity threshold is identified as an abnormal calculation unit. By merging spatially adjacent abnormal computational units into connected regions, the non-uniform reaction region within the cobalt metal bed is obtained.

6. The fixed-bed reaction process coupling control method based on industrial cloud computing according to claim 2, characterized in that, The fixed-bed reaction flow field control scheme includes: Based on the effective specific surface area of ​​the reaction, local reaction rate, reactant concentration, cobalt acetate formation concentration and cobalt metal consumption corresponding to the uneven reaction region, it is determined whether there is a need for cobalt metal replenishment in the uneven reaction region. When the effective reaction surface area in the uneven reaction region is lower than a preset specific surface area threshold, and the local reaction rate is lower than a preset reaction rate threshold, the corresponding uneven reaction region is identified as a region with insufficient cobalt metal reaction capacity, and cobalt metal replenishment demand information is generated. The cobalt metal replenishment demand information includes at least one of the following: replenishment area, replenishment amount demand, replenishment timing demand, and target effective reaction surface area after replenishment.

7. The fixed-bed reaction process coupling control method based on industrial cloud computing according to claim 6, characterized in that, The fixed-bed reaction flow field control scheme further includes: Based on the cobalt metal feeding requirement information, candidate feeding schemes are generated. The candidate feeding schemes include at least one of the following: feeding location, feeding amount, particle size range, morphology type, packing density, and packing layers of cobalt metal. Based on the candidate feeding scheme, the porosity distribution and effective reaction specific surface area distribution in the cobalt fixed bed reaction zone are updated to obtain the predicted bed structure after feeding. Input the predicted bed structure after feeding into the numerical simulation model to calculate the predicted reaction data after feeding corresponding to the candidate feeding scheme; The degree of feed disturbance corresponding to the candidate feed scheme is determined based on the predicted reaction data after feed feeding, and the fixed bed reaction flow field control scheme is determined based on the degree of feed disturbance.

8. The method for coupled control of a fixed-bed reaction process based on industrial cloud computing according to claim 7, characterized in that, The methods for determining the degree of material feeding disturbance include: Based on the candidate feeding scheme, determine the newly added effective reaction area region of metallic cobalt in the predicted bed structure after feeding; Based on the predicted reaction data after feeding, determine the area within the cobalt fixed-bed reaction zone that can be continuously covered by the reaction liquid through which the reaction liquid continuously flows. Determine the overlap and connectivity between the newly added effective reaction area of ​​cobalt metal and the continuously covered area of ​​the reaction liquid; Based on the overlapping state and the connectivity state, the effective contact degree of the newly added reaction area corresponding to the candidate feeding scheme is determined, and the effective contact degree of the newly added reaction area is used as the feeding disturbance degree corresponding to the candidate feeding scheme.

9. The fixed-bed reaction process coupling control method based on industrial cloud computing according to claim 8, characterized in that, The fixed-bed reaction flow field control scheme further includes: The degree of material feeding disturbance is compared with the preset disturbance allowable conditions; When the degree of feeding disturbance does not meet the preset disturbance allowable condition, at least one of the following in the candidate feeding scheme is updated according to the overlap and connection state between the newly added effective reaction area of ​​metallic cobalt and the continuous coverage area of ​​the reaction liquid: feeding position, feeding amount, particle size range, morphology type, filling density and filling layer. Based on the updated candidate feeding schemes, the predicted bed structure after feeding is redefined, and the predicted reaction data after feeding corresponding to the updated candidate feeding schemes are recalculated. Based on the recalculated predicted reaction data after feeding, the feeding disturbance level corresponding to the updated candidate feeding scheme is determined again until the feeding disturbance level meets the preset disturbance allowance condition. The candidate feeding scheme that meets the preset disturbance allowance condition is determined as the fixed bed reaction flow field control scheme.

10. A fixed-bed reaction process coupling control system based on industrial cloud computing, used to implement the fixed-bed reaction process coupling control method based on industrial cloud computing as described in any one of claims 1-9, characterized in that, The system includes: The numerical simulation module is used to acquire the physical properties and operating parameters of the reaction liquid in the fixed-bed reactor, and to obtain the reaction data in the fixed-bed reactor by multi-field coupling solution through a preset numerical simulation model. The numerical simulation model includes a porous media model for characterizing the flow resistance of the reaction liquid in the cobalt bed, a component transfer model for characterizing the mass transfer process between cobalt and the reaction liquid, a heat transfer model for characterizing the exothermic or endothermic process of the reaction, and a reaction kinetic model for characterizing the process of cobalt forming cobalt acetate. The reaction state identification module is used to identify the reaction uneven region in the cobalt metal bed based on the reaction data, and to determine whether there is a need for cobalt metal replenishment in the reaction uneven region based on the effective reaction specific surface area, local reaction rate, reactant concentration value, cobalt acetate generation concentration value and cobalt metal consumption corresponding to the reaction uneven region. The control scheme generation module is used to generate candidate feeding schemes when there is a demand for cobalt metal feeding, update the porosity distribution and effective reaction specific surface area distribution in the cobalt metal fixed bed reaction zone based on the candidate feeding schemes, obtain the predicted bed structure after feeding, and input the predicted bed structure after feeding into the numerical simulation model to obtain predicted reaction data after feeding; and determine the feeding disturbance degree corresponding to the candidate feeding scheme based on the predicted reaction data after feeding, and update the candidate feeding scheme based on the feeding disturbance degree until the updated candidate feeding scheme meets the preset disturbance allowable condition, and determine the candidate feeding scheme that meets the preset disturbance allowable condition as the fixed bed reaction flow field control scheme.

Images