SCR denitration catalyst and preparation method thereof

Abstract

The preparation method comprises the following steps: adding a manganese-containing precursor solution into anatase type titanium dioxide, stirring, standing, uniformly infiltrating, and carrying out image monitoring and state analysis on the infiltration state of the solution by adopting an intelligent industrial control technology based on deep learning. According to the method, the solution local texture features are extracted to extract infiltration state information based on the solution local texture features, then feature dominant enhancement and uniformity scoring are carried out on the texture features of all local areas of the image, the overall infiltration uniformity of the solution is revealed in combination with a mathematical statistical method, and therefore reasonable control over the infiltration degree and infiltration time is facilitated. In this way, it can be ensured that nano-manganese dioxide is evenly distributed on the catalyst carrier, the infiltration time is effectively controlled, and therefore the preparation efficiency and quality stability of the catalyst are improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent preparation, and more specifically, to an SCR denitration catalyst and a preparation method thereof. Background Art

[0002] With the increasing attention paid to environmental protection, reducing nitrogen oxides (NOx) in industrial emissions has become one of the focuses of environmental protection work. As a highly efficient denitrification technology, SCR (Selective Catalytic Reduction) technology has been widely used in many industries such as power and chemical industry. Traditional SCR denitrification catalysts are mainly based on vanadium-titanium materials, but this type of catalyst is easily affected by heavy metal components in exhaust gas, resulting in decreased catalyst activity or even poisoning failure.

[0003] In this regard, the invention patent with publication number CN113877566B discloses a method for preparing an SCR denitration catalyst that is resistant to heavy metal poisoning. Nano manganese dioxide is used as a modifier. A certain amount of nano manganese dioxide is introduced into a vanadium-titanium denitration catalyst. Under working conditions, manganese dioxide reacts with heavy metals in dust, thereby avoiding the deposition of heavy metals on active components and causing catalyst poisoning, while also improving the catalytic denitration performance of the catalyst.

[0004] However, although the above methods have improved the anti-poisoning performance of the catalyst to a certain extent, in the actual operation process, how to ensure the uniform distribution of nanomaterials on the catalyst carrier (such as anatase titanium dioxide) is still one of the key factors affecting the performance of the catalyst. Traditional preparation methods often rely on manual operation and empirical judgment, and it is difficult to achieve precise control of the infiltration process, which may lead to uneven distribution of nano manganese dioxide on the carrier surface, affecting the overall performance and stability of the catalyst.

[0005] Therefore, a method for preparing an optimized SCR denitration catalyst is desired. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an SCR denitration catalyst and a preparation method thereof, which adds a manganese-containing precursor solution to anatase titanium dioxide, stirs it and then allows it to be evenly infiltrated, and uses deep learning-based intelligent industrial control technology to perform image monitoring and state analysis on the infiltration state of the solution to extract the infiltration state information based on the local texture features of the solution, and then, by performing feature explicit enhancement and uniformity scoring on the texture features of each local area of ​​the image, and combining mathematical statistical methods, the overall infiltration uniformity of the solution is revealed, which helps to achieve reasonable control of the infiltration degree and infiltration time. In this way, the uniform distribution of nano-manganese dioxide on the catalyst carrier can be ensured, and the infiltration time can be effectively controlled, thereby improving the preparation efficiency and quality stability of the catalyst.

[0007] According to one aspect of the present application, a method for preparing an SCR denitration catalyst is provided, which comprises: adding a manganese-containing precursor solution to anatase titanium dioxide according to an equal volume impregnation method, stirring, and standing to make it evenly infiltrated to obtain a manganese-containing wet material; heating the manganese-containing wet material from room temperature to 160° C. and keeping it warm for drying, and then grinding it to obtain a manganese-containing powder; adding a vanadium source precursor solution to the manganese-containing powder, preparing it into a slurry, standing, drying, and calcining it after mixing, to obtain an SCR denitration catalyst that prevents heavy metal poisoning, characterized in that the manganese-containing precursor solution is added to the anatase titanium dioxide, stirring, and standing to make it evenly infiltrated, comprising:

[0008] Acquire a solution infiltration state image captured by a camera;

[0009] Extracting local depth texture features from the solution infiltration state image to obtain a set of local depth texture feature maps of the solution infiltration state;

[0010] Performing feature explicit hierarchical enhancement on each of the local depth texture feature maps in the set of local depth texture feature maps in the solution infiltration state to obtain a set of enhanced local depth texture feature maps in the solution infiltration state;

[0011] Performing infiltration uniformity evaluation on each of the local depth texture enhancement feature maps in the set of local depth texture enhancement feature maps in the solution infiltration state to obtain a set of local infiltration uniformity score values;

[0012] The mean and variance of the set of the solution local infiltration state uniformity score values ​​are calculated, and whether the infiltration is uniform is determined based on the mean and variance.

[0013] According to another aspect of the present application, there is provided an SCR denitration catalyst, comprising:

[0014] A solution infiltration state image acquisition module, used to acquire a solution infiltration state image collected by a camera;

[0015] A local depth texture feature extraction module, used for extracting local depth texture features from the solution infiltration state image to obtain a set of local depth texture feature maps of the solution infiltration state;

[0016] A feature-explicit hierarchical enhancement module is used to perform feature-explicit hierarchical enhancement on each of the solution-infiltrated state local depth texture feature maps in the set of solution-infiltrated state local depth texture feature maps to obtain a set of solution-infiltrated state local depth texture enhancement feature maps;

[0017] A wetting uniformity evaluation module, used to respectively perform wetting uniformity evaluation on each of the solution infiltration state local depth texture enhancement feature maps in the set of solution infiltration state local depth texture enhancement feature maps to obtain a set of local infiltration uniformity score values;

[0018] The wetting state uniformity judgment module is used to calculate the mean and variance of the set of local wetting state uniformity score values ​​of the solution, and determine whether the infiltration is uniform based on the mean and variance.

[0019] Compared with the prior art, the present application provides an SCR denitration catalyst and a preparation method thereof, which adds a manganese-containing precursor solution to anatase titanium dioxide, stirs it and then allows it to be evenly infiltrated, and uses deep learning-based intelligent industrial control technology to perform image monitoring and state analysis on the infiltration state of the solution to extract the infiltration state information based on the local texture features of the solution, and then, by performing feature explicit enhancement and uniformity scoring on the texture features of each local area of ​​the image, and combining mathematical statistical methods, to reveal the overall infiltration uniformity of the solution, thereby helping to achieve reasonable control of the degree of infiltration and infiltration time. In this way, the uniform distribution of nano manganese dioxide on the catalyst carrier can be ensured, and the infiltration time can be effectively controlled, thereby improving the preparation efficiency and quality stability of the catalyst. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1 is a flow chart of a method for preparing an SCR denitration catalyst according to an embodiment of the present application;

[0022] Figure 2It is a data flow diagram of a method for preparing an SCR denitration catalyst according to an embodiment of the present application;

[0023] Figure 3 It is a flowchart of sub-step S1 of the method for preparing an SCR denitration catalyst according to an embodiment of the present application;

[0024] Figure 4 4 is a block diagram of an SCR denitration catalyst according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0026] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0027] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0028] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0029] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0030] Invention patent CN113877566B proposes a method for preparing a SCR denitration catalyst that is resistant to heavy metal poisoning, which comprises: adding a manganese-containing precursor solution to anatase-type titanium dioxide by an equal volume impregnation method, stirring, and standing to allow it to be evenly infiltrated to obtain a manganese-containing wet material; heating the manganese-containing wet material from room temperature to 160°C and keeping it warm for drying, and then grinding it to obtain a manganese-containing powder; adding a vanadium source precursor solution to the manganese-containing powder, preparing it into a slurry, mixing, standing, drying, and calcining to obtain a SCR denitration catalyst that is resistant to heavy metal poisoning.

[0031] Although the above method improves the anti-poisoning performance of the catalyst to a certain extent, in the actual operation process, how to ensure the uniform distribution of nanomaterials on the catalyst carrier (such as anatase titanium dioxide) is still one of the key factors affecting the performance of the catalyst. Traditional preparation methods often rely on manual operation and empirical judgment, which makes it difficult to achieve precise control of the infiltration process, which may lead to uneven distribution of nano manganese dioxide on the carrier surface, affecting the overall performance and stability of the catalyst.

[0032] Based on this, in the technical solution of the present application, a method for preparing an SCR denitration catalyst is proposed. Figure 1 The present invention is a flow chart of a method for preparing an SCR denitration catalyst according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of the method for preparing the SCR denitration catalyst according to the embodiment of the present application. Figure 1 and Figure 2 As shown, the preparation method of the SCR denitration catalyst according to the embodiment of the present application includes the following steps: S1, adding a manganese-containing precursor solution to anatase titanium dioxide according to an equal volume impregnation method, stirring, and standing to allow it to be evenly wetted to obtain a manganese-containing wet material; S2, heating the manganese-containing wet material from room temperature to 160°C and keeping it warm for drying, and then grinding it to obtain a manganese-containing powder; S3, adding a vanadium source precursor solution to the manganese-containing powder, preparing it into a slurry, mixing, standing, drying, and calcining to obtain an SCR denitration catalyst that is resistant to heavy metal poisoning.

[0033] In particular, in S1, according to the equal volume impregnation method, the manganese-containing precursor solution is added to the anatase titanium dioxide, stirred, and then allowed to stand for uniform infiltration to obtain a manganese-containing wet material. In a specific example of the present application, Figure 3As shown, the S1 includes: S11, acquiring a solution infiltration state image collected by a camera; S12, performing local depth texture feature extraction on the solution infiltration state image to obtain a set of solution infiltration state local depth texture feature maps; S13, performing feature explicit hierarchical enhancement on each solution infiltration state local depth texture feature map in the set of solution infiltration state local depth texture feature maps to obtain a set of solution infiltration state local depth texture enhancement feature maps; S14, performing infiltration uniformity evaluation on each solution infiltration state local depth texture enhancement feature map in the set of solution infiltration state local depth texture enhancement feature maps to obtain a set of local infiltration uniformity score values; S15, calculating the mean and variance of the set of solution local infiltration state uniformity score values, and determining whether the infiltration is uniform based on the mean and variance.

[0034] Specifically, the step S11 is to obtain a solution infiltration state image captured by a camera. It should be understood that by capturing the solution infiltration state image by a camera, the dynamic changes in the solution infiltration process can be captured in real time, and the infiltration of the manganese-containing precursor solution in the anatase titanium dioxide can be accurately reflected, thereby providing reliable data support for the evaluation of the uniformity of the solution infiltration.

[0035] Specifically, the S12 extracts local depth texture features from the solution infiltration state image to obtain a set of local depth texture feature maps of the solution infiltration state. That is, in the technical solution of the present application, first, the solution infiltration state image is evenly divided along the width direction of the solution infiltration state image to obtain a set of local depth images of the solution infiltration state; here, considering that in the solution infiltration state image, the solution infiltration state at different positions may be different, for example, the infiltration state of the edge area may not be as uniform as the central area. Therefore, in order to more carefully analyze the infiltration state of the solution in different areas, the present application further evenly divides the solution infiltration state image along the width direction of the solution infiltration state image to form multiple image slices to obtain a set of local depth images of the solution infiltration state. In this way, independent texture feature analysis can be performed for each specific position in the solution infiltration state image, which helps to more accurately evaluate the infiltration uniformity of the solution in different areas and provide more detailed data support for the subsequent overall infiltration uniformity evaluation.

[0036] Then, each local depth image of the solution infiltration state in the set of local depth images of the solution infiltration state is input into a texture feature extractor based on a convolutional neural network model to obtain a set of local depth texture feature maps of the solution infiltration state, wherein the texture feature extractor based on the convolutional neural network model uses a strip pooling layer. In particular, the convolutional neural network processes the spatial information of the image by performing a sliding convolution operation on the local depth image of the solution infiltration state, effectively extracting the detail texture features in the image to obtain a set of local depth texture feature maps of the solution infiltration state. In particular, in order to better adapt to the texture features of the local depth image of the solution infiltration state, the texture feature extractor based on the convolutional neural network model described in the present application uses a strip pooling layer for feature pooling processing. Specifically, the strip pooling layer (Strip Pooling Layer) uses a long strip pooling core extending along the spatial dimension to perform a pooling operation, which can effectively retain the potential directional texture features in the solution infiltration process, thereby improving the accuracy of texture feature extraction.

[0037] Specifically, the S13 performs feature explicit hierarchical enhancement on each local depth texture feature map of the solution infiltration state in the set of local depth texture feature maps of the solution infiltration state to obtain a set of enhanced local depth texture feature maps of the solution infiltration state. Considering that in the local depth texture feature map of the solution infiltration state, the texture features at different positions may show different infiltration states, therefore, in order to further enhance the texture feature differentiation within the local depth texture feature map of the solution infiltration state, so as to improve the recognition accuracy of subtle texture structure changes in the image, the present application proposes a feature enhancement method, which performs pixel-level texture feature explicit enhancement on each local depth texture feature map of the solution infiltration state by layered precipitation, so as to enhance the difference between texture features at different positions, so as to more clearly distinguish the specific infiltration conditions of the solution on the surface or inside the solution.

[0038] Specifically, the specific steps of performing feature explicit hierarchical enhancement on each local depth texture feature map of the solution infiltration state in the set of local depth texture feature maps of the solution infiltration state include: first, performing pixel-based texture structure semantic saliency measurement on the local depth texture feature map of the solution infiltration state to obtain a solution infiltration state fine-grained texture structure semantic saliency distribution matrix; that is, in the technical solution of the present application, firstly performing pixel-grained feature decoupling on the local depth texture feature map of the solution infiltration state to obtain a set of solution infiltration state pixel-grained local depth texture decoupling feature vectors; through feature decoupling, the local depth texture feature map of the solution infiltration state is decomposed into a series of independent solution infiltration state pixel-grained local depth texture decoupling feature vectors to eliminate unnecessary coupling relationships between pixels and improve the ability to understand the texture features of each local position. Then, each of the solution infiltration state pixel granularity local depth texture decoupling feature vectors in the set of the solution infiltration state pixel granularity local depth texture decoupling feature vectors is input into the hyperbolic space mapper to obtain a set of hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vectors; considering that the hyperbolic geometric space has better adaptability than the Euclidean space when processing a hierarchical data set, therefore, the present application further sends the obtained set of solution infiltration state pixel granularity local depth texture decoupling feature vectors into the hyperbolic space mapper for domain transformation, that is, mapping from the Euclidean geometric space to the hyperbolic geometric space, so as to utilize the characteristics of the hyperbolic geometric space, better retain the hierarchical relationship and topological structure between the texture features of each local position, reveal the hidden hierarchical structure between the features, and thus improve the feature expression capability. Wherein, the hyperbolic space mapper is used to use the first weight parameter matrix and the second weight parameter matrix to continuously linearly transform the solution infiltration state pixel granularity local depth texture decoupling feature vector to obtain the hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vector. Furthermore, the fine-grained structure semantic saliency metric coefficients of each hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vector in the set of the hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vectors are calculated to obtain the solution infiltration state fine-grained texture structure semantic saliency distribution matrix formed by arranging multiple solution infiltration state fine-grained texture structure semantic saliency metric coefficients. That is, in the hyperbolic space, by defining a metric function, the semantic saliency of each pixel granularity texture feature is quantitatively evaluated, and by arranging the obtained semantic saliency metric coefficients into a matrix form according to the corresponding pixel space position, the solution infiltration state fine-grained texture structure semantic saliency distribution matrix is ​​obtained to reveal the overall semantic saliency distribution of the solution infiltration state local depth texture feature map, intuitively display the semantic saliency difference of the solution infiltration state in different regions, and thus provide an important basis for subsequent feature selection.

[0039] More specifically, the semantic saliency measurement formula is used to measure the texture structure semantic saliency based on pixel granularity on the local depth texture feature map of the solution infiltration state to obtain the semantic saliency distribution matrix of the fine-grained texture structure of the solution infiltration state; wherein the semantic saliency measurement formula is:

[0040] decouple(F i )={v1,v2,...,v i ,...,v n}

[0041] h i =W1v i W2

[0042]

[0043] Among them, F i represents the local depth texture feature map of the solution infiltration state, decouple(·) represents the pixel granularity feature decoupling, v1, v2, v i and v n denote the first, second, i-th and n-th solution infiltration state pixel size local depth texture decoupling feature vectors, W1 and W2 denote the first weight parameter matrix and the second weight parameter matrix, respectively. i represents the local depth texture decoupling feature vector of the pixel granularity of the ith hyperbolic space modulation solution infiltration state, ‖·‖ 2 represents the square of the norm of the eigenvector, represents the logarithmic function with base 2, s i Represents the semantic saliency coefficient of the fine-grained texture structure of the i-th solution infiltration state.

[0044] Furthermore, based on the semantic saliency distribution matrix of the fine-grained texture structure of the solution infiltration state, the feature explicit hierarchical modulation is performed on the local depth texture feature map of the solution infiltration state to obtain the local depth texture enhancement feature map of the solution infiltration state. That is, in the technical solution of the present application, the semantic saliency distribution matrix of the fine-grained texture structure of the solution infiltration state is input into the feature explicit hierarchical precipitation module based on the multi-layer mask function to obtain the semantic saliency distribution weight matrix of the fine-grained texture structure of the solution infiltration state; and the semantic saliency distribution weight matrix of the fine-grained texture structure of the solution infiltration state is calculated and multiplied by the position points of each feature matrix of the local depth texture feature map of the solution infiltration state along the channel dimension to obtain the local depth texture enhancement feature map of the solution infiltration state. Here, a multi-layer mask function is used to perform layered precipitation on the semantic significance distribution matrix of the fine-grained texture structure of the solution infiltration state. By applying different masks to features at different levels, selective amplification of texture features at different positions in the local depth texture feature map of the solution infiltration state is achieved, and a corresponding weight matrix is ​​generated. In addition, by performing positional point multiplication operations on each feature matrix of the original local depth texture feature map of the solution infiltration state along the channel dimension, weighted adjustment of the original features is achieved, thereby significantly enhancing the differences in texture features at different spatial positions in the image, helping to identify more subtle texture structure changes and improving the accuracy of uniformity assessment.

[0045] More specifically, based on the semantic saliency distribution matrix of the fine-grained texture structure of the solution infiltration state, the feature-explicit hierarchical modulation is performed on the local depth texture feature map of the solution infiltration state by the following feature-explicit hierarchical modulation formula to obtain the local depth texture enhancement feature map of the solution infiltration state; wherein the feature-explicit hierarchical modulation formula is:

[0046]

[0047] F f =F i ⊙S

[0048] Where mask(·) represents a multi-layer mask function, θ represents the mask threshold, S represents the semantic saliency distribution weight matrix of the fine-grained texture structure in the solution wetting state, and s i ' represents the weight value of the ith position in the semantic saliency distribution weight matrix of the fine-grained texture structure in the solution infiltration state, ⊙ represents the point multiplication by position, F f A local depth texture enhancement feature map representing the solution infiltration state.

[0049] Specifically, the S14 is to evaluate the uniformity of the infiltration uniformity of each local depth texture enhancement feature map in the set of local depth texture enhancement feature maps of the solution infiltration state to obtain a set of local infiltration uniformity score values. That is, in the technical solution of the present application, each local depth texture enhancement feature map in the set of local depth texture enhancement feature maps of the solution infiltration state is input into a uniformity score module based on a feedforward neural network model to obtain a set of local infiltration uniformity score values. It should be understood that after the above-mentioned internal texture structure difference enhancement processing, the set of local depth texture enhancement feature maps of the solution infiltration state comprehensively and accurately maps the infiltration uniformity information contained in the local depth images of each solution infiltration state. In order to further realize the quantitative evaluation of the uniformity of the infiltration uniformity of each local depth image of the solution infiltration state, the present application uses a feedforward neural network model to construct a uniformity score module, and each local depth texture enhancement feature map of the solution infiltration state is scored for uniformity. Those of ordinary skill in the art should know that a feedforward neural network is a one-way propagation neural network with a simple structure, easy to implement, and excellent performance in dealing with linearly separable problems. In the present application, the feedforward neural network model is designed to have multiple hidden layers, each of which contains a certain number of neurons, and the neurons perform nonlinear transformations on the input texture features through activation functions to capture the complex patterns of the solution infiltration state. Through layer-by-layer transmission and transformation, the mapping relationship between the input texture structure features and the infiltration uniformity score is learned, thereby outputting the uniformity scores of the local depth texture enhancement feature maps of each solution infiltration state, and obtaining a set of local infiltration uniformity score values.

[0050] Specifically, the S15 calculates the mean and variance of the set of local wetting state uniformity score values ​​of the solution, and determines whether the wetting is uniform based on the mean and variance. That is, by calculating the mean and variance of the set of local wetting state uniformity score values ​​of the solution, the average wetting uniformity level of the nanomaterial on the surface or inside of the solution and the degree of fluctuation of the solution wetting uniformity in different regions are revealed, thereby achieving a comprehensive evaluation of the overall wetting uniformity of the nanomaterial. For example, by performing threshold judgment on the mean and the variance, it is determined whether the nanomaterial is uniformly infiltrated.

[0051] In particular, in S2, the manganese-containing wet material is heated from room temperature to 160° C. and kept warm for drying, and then ground to obtain manganese-containing powder. In SCR (selective catalytic reduction) technology, the performance of the catalyst is crucial to reducing nitrogen oxides (NOx) in industrial emissions.

[0052] In one example, first, the manganese-containing wet material is placed in a drying device that can precisely control the temperature, such as a rotary evaporator, an oven, or a more professional heat treatment furnace. The key to this stage is to set a reasonable heating rate, generally 1 to 5 degrees Celsius per minute. The purpose of this is to ensure that the temperature inside and outside the wet material rises evenly to avoid physical structure damage or uneven composition caused by sudden temperature changes. This step not only helps to effectively remove moisture, but also lays the foundation for the subsequent chemical changes. As the temperature gradually rises to the predetermined target temperature, that is, 160°C, the temperature is stopped and maintained at this temperature for a certain period of time, which is usually called the insulation period. The insulation time will be adjusted according to factors such as the amount and shape of the wet material, and can range from several hours to more than ten hours. This insulation stage is very important to ensure that the wet material is thoroughly dried and to promote chemical reactions. For example, during this process, the precursor may undergo decomposition, crystallization, or other forms of chemical transformation, which are necessary to form the final active substance. During the entire heating and insulation process, the temperature changes need to be closely monitored to ensure that the temperature is stable near the set value. If an intelligent control system is used, the heating power can be automatically adjusted to maintain a constant temperature, thus ensuring the continuity and stability of the entire process.

[0053] After the insulation is completed, the heating device is turned off and the manganese-containing wet material is allowed to cool naturally to near room temperature. In some cases, forced ventilation or other methods can be used to accelerate the cooling process. After the sample is completely cooled, it is removed from the drying equipment. This cooling process also needs to be carried out carefully to prevent the hot material from suddenly contacting the cold air and causing cracks or other physical damage. In addition, the appropriate cooling rate also helps to maintain the chemical stability and structural integrity of the material. The cooled block or flake manganese-containing material needs to be further processed into fine particles, that is, manganese-containing powder. For this purpose, appropriate grinding tools should be selected, such as ball mills, vibrating mills or planetary mills, to ensure that the particle size distribution of the final product is uniform and suitable for subsequent processing requirements. The grinding operation involves placing the cooled material into the selected grinding tool and crushing it until the required fine particles are obtained. Sometimes it is also necessary to screen the ground powder through a screen to remove oversized particles and ensure that all particle sizes are within the specified range. This step is crucial to the performance of the final catalyst, because the appropriate particle size distribution can directly affect the specific surface area and pore structure of the catalyst, and thus affect its catalytic efficiency.

[0054] Throughout the process, safety regulations must be strictly followed, especially when working in high temperature environments, fire and explosion prevention must be taken into account. For specific types of catalyst precursors, other factors may need to be considered, such as atmosphere conditions (such as drying under nitrogen protection), humidity control, etc., to prevent unwanted side reactions. These additional measures can further improve the quality and performance of the catalyst, ensuring its reliability and effectiveness in practical applications.

[0055] Through the above-mentioned carefully designed drying and powder preparation process, not only can the manganese-containing wet material be effectively dried, but also the formation of the desired catalytically active substances is promoted through the precise control of temperature, time and environmental conditions, providing high-quality basic materials for the subsequent catalyst synthesis.

[0056] In particular, the S3, adds a vanadium source precursor solution to the manganese-containing powder, modulates it into a slurry, and after mixing, stands, dries, and roasts to obtain an SCR denitration catalyst that is resistant to heavy metal poisoning. In the selective catalytic reduction (SCR) technology, the development of efficient denitration catalysts is crucial to reducing nitrogen oxides (NOx) in industrial emissions. The present application proposes an optimized SCR denitration catalyst preparation method, especially for the uniform distribution of nano manganese dioxide on anatase titanium dioxide carrier, so as to improve the stability of the catalyst and its resistance to heavy metal poisoning.

[0057] In order to further enhance the overall performance of the catalyst, a vanadium source precursor solution needs to be introduced into the manganese-containing powder. First, select an appropriate vanadium source precursor solution, such as ammonium metavanadate (NH4VO3), vanadium trioxide (V2O5) or other suitable compound aqueous solution. Calculate the corresponding solution amount according to the required vanadium content and ensure that its concentration is suitable for the subsequent impregnation process. Subsequently, the vanadium source precursor solution is slowly added to the manganese-containing powder while continuously stirring to ensure uniform mixing between the two. It is very important to control the liquid addition speed and stirring intensity to avoid excessive bubbles or agglomeration, which will affect the quality of the final product. As the liquid is gradually absorbed by the powder, the material begins to transform into a viscous slurry. This slurry should have good fluidity and dispersibility for subsequent operations. In addition, ultrasonic treatment or mechanical shearing can be used to assist mixing to ensure that all particles are evenly covered. The purpose of this is to enable the vanadium source to be evenly distributed throughout the powder, thereby ensuring the uniformity of the active ingredients throughout the catalyst.

[0058] Once a uniform slurry is formed, it is placed in a container and allowed to stand at room temperature for a period of time. The purpose of standing is to allow the components in the solution to further diffuse and penetrate between the manganese-containing powder particles, ensuring that the vanadium source can be distributed as evenly as possible on the carrier. The standing time usually depends on the specific material properties and the required reaction conditions, and can generally range from a few hours to one or two days. During this time, the infiltration process can be assisted by slight vibration, but care should be taken not to destroy the structure that has been formed. The standing stage not only helps to improve the uniformity of the distribution of the vanadium source on the carrier, but also provides the necessary preparation time for subsequent chemical changes. This stage allows the components in the solution to have enough time to fully physically and chemically combine, thereby ensuring that the active components can be firmly attached to the surface of the carrier.

[0059] After the standing is completed, the drying stage begins. The main goal of drying is to remove moisture and other volatile components from the slurry and solidify the material into a solid form. Common drying methods include natural drying, oven drying, or more advanced vacuum freeze-drying technology. For this step, oven drying is more common. Set an appropriate temperature (such as 80-120°C) and keep it under constant conditions for a long enough time until the sample is completely dry. During the drying process, care should also be taken to prevent physical damage or chemical changes caused by local overheating. In addition, a staged heating strategy can be considered, first pre-drying at low temperature, and then gradually increasing the temperature to the final drying temperature to maximize the protection of the structural integrity of the material. The dried product should be in a loose and uniform powder state for subsequent operation. The drying process is critical to removing excess moisture and solvents, as the presence of these components may affect the structure and performance of the catalyst.

[0060] The dried solid material needs to undergo a high-temperature calcination process in order to promote the decomposition or phase transformation of the vanadium source precursor to form catalytically active vanadium compounds. The calcination temperature is usually high, which may reach 400-600°C or even higher, depending on the type of vanadium source selected and the target phase to be generated. Atmospheric conditions, such as air, nitrogen protection or reducing gas environment, also need to be considered during calcination, which will affect the crystal structure and surface properties of the final product. The calcination time should also be adjusted according to actual conditions to ensure that all precursors have been fully converted and no unreacted substances remain. In addition, a multi-stage heating method can be used, that is, starting from a lower temperature and gradually rising to the highest calcination temperature, to promote the gradual decomposition of the precursor and form a stable active phase. During the calcination process, the temperature changes need to be closely monitored to ensure that the heating rate is moderate and to avoid adverse consequences caused by excessive temperature fluctuations. The calcination process is crucial for the formation and stabilization of the active components of the catalyst, which determines the structure and catalytic performance of the final catalyst.

[0061] After the above series of treatments, a new type of SCR denitrification catalyst was finally obtained. This catalyst not only contains the original nano manganese dioxide modified component, but also adds an appropriate amount of vanadium-based active components, making it more resistant to heavy metal poisoning and more efficient.

[0062] In summary, the preparation method of the SCR denitration catalyst according to the embodiment of the present application is explained, which adds a manganese-containing precursor solution to anatase titanium dioxide, stirs it and then allows it to be evenly infiltrated, and uses deep learning-based intelligent industrial control technology to perform image monitoring and state analysis on the infiltration state of the solution to extract the infiltration state information based on the local texture features of the solution, and then, by performing feature explicit enhancement and uniformity scoring on the texture features of each local area of ​​the image, and combining mathematical statistical methods, to reveal the overall infiltration uniformity of the solution, thereby helping to achieve reasonable control of the degree of infiltration and infiltration time. In this way, the uniform distribution of nano manganese dioxide on the catalyst carrier can be ensured, and the infiltration time can be effectively controlled, thereby improving the preparation efficiency and quality stability of the catalyst.

[0063] Furthermore, an SCR denitration catalyst is also provided.

[0064] Figure 4 FIG. 1 is a block diagram of an SCR denitration catalyst according to an embodiment of the present application. Figure 4 As shown, the SCR denitration catalyst 300 according to the embodiment of the present application includes: a solution wetting state image acquisition module 310, which is used to acquire a solution wetting state image collected by a camera; a local depth texture feature extraction module 320, which is used to perform local depth texture feature extraction on the solution wetting state image to obtain a set of solution wetting state local depth texture feature maps; a feature explicit hierarchical enhancement module 330, which is used to perform feature explicit hierarchical enhancement on each solution wetting state local depth texture feature map in the set of solution wetting state local depth texture feature maps to obtain a set of solution wetting state local depth texture enhancement feature maps; an infiltration uniformity evaluation module 340, which is used to perform infiltration uniformity evaluation on each solution wetting state local depth texture enhancement feature map in the set of solution wetting state local depth texture enhancement feature maps to obtain a set of local wetting uniformity score values; and an infiltration state uniformity judgment module 350, which is used to calculate the mean and variance of the set of solution local wetting state uniformity score values, and determine whether the infiltration is uniform based on the mean and variance.

[0065] As described above, the SCR denitration catalyst 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server having a preparation algorithm of the SCR denitration catalyst. In a possible implementation, the SCR denitration catalyst 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the SCR denitration catalyst 300 can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the SCR denitration catalyst 300 can also be one of the many hardware modules of the wireless terminal.

[0066] Alternatively, in another example, the SCR denitration catalyst 300 and the wireless terminal may also be separate devices, and the SCR denitration catalyst 300 may be connected to the wireless terminal via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0067] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for preparing an SCR denitration catalyst, comprising: According to the equal volume impregnation method, the manganese-containing precursor solution is added to the anatase titanium dioxide, stirred, and then allowed to stand for uniform infiltration to obtain a manganese-containing wet material; The manganese-containing wet material is heated from room temperature to 160° C. and kept warm for drying, and then ground to obtain a manganese-containing powder; a vanadium source precursor solution is added to the manganese-containing powder to prepare a slurry, and after mixing, the mixture is allowed to stand, dried, and calcined to obtain a SCR denitration catalyst that prevents heavy metal poisoning, characterized in that the manganese-containing precursor solution is added to anatase titanium dioxide, stirred, and allowed to stand for uniform infiltration, comprising: Acquire a solution infiltration state image captured by a camera; Extracting local depth texture features from the solution infiltration state image to obtain a set of local depth texture feature maps of the solution infiltration state; Performing feature explicit hierarchical enhancement on each of the local depth texture feature maps in the set of local depth texture feature maps in the solution infiltration state to obtain a set of enhanced local depth texture feature maps in the solution infiltration state; Performing infiltration uniformity evaluation on each of the local depth texture enhancement feature maps in the set of local depth texture enhancement feature maps in the solution infiltration state to obtain a set of local infiltration uniformity score values; The mean and variance of the set of the solution local infiltration state uniformity score values ​​are calculated, and whether the infiltration is uniform is determined based on the mean and variance.

2. The method for preparing the SCR denitration catalyst according to claim 1, characterized in that: Extracting local depth texture features from the solution infiltration state image to obtain a set of local depth texture feature maps of the solution infiltration state includes: Uniformly dividing the solution infiltration state image along the width direction of the solution infiltration state image to obtain a set of local depth images of the solution infiltration state; The texture features of each solution infiltration state local depth image in the set of solution infiltration state local depth images are extracted respectively to obtain a set of solution infiltration state local depth texture feature maps.

3. The method for preparing the SCR denitration catalyst according to claim 2, characterized in that: Extracting texture features of each solution infiltration state local depth image in the set of solution infiltration state local depth images respectively to obtain a set of solution infiltration state local depth texture feature maps, including: Each local depth image of the solution infiltration state in the set of local depth images of the solution infiltration state is input into a texture feature extractor based on a convolutional neural network model to obtain a set of local depth texture feature maps of the solution infiltration state, wherein the texture feature extractor based on a convolutional neural network model uses a strip pooling layer.

4. The method for preparing the SCR denitration catalyst according to claim 3, characterized in that: Performing feature explicit hierarchical enhancement on each of the local depth texture feature maps in the set of local depth texture feature maps in the solution infiltration state to obtain a set of enhanced local depth texture feature maps in the solution infiltration state, including: Performing a texture structure semantic saliency measurement based on pixel granularity on the local depth texture feature map of the solution infiltration state to obtain a solution infiltration state fine-grained texture structure semantic saliency distribution matrix; Based on the semantic saliency distribution matrix of the fine-grained texture structure of the solution infiltration state, the local depth texture feature map of the solution infiltration state is subjected to feature explicit hierarchical modulation to obtain the local depth texture enhanced feature map of the solution infiltration state.

5. The method for preparing the SCR denitration catalyst according to claim 4, characterized in that: The local depth texture feature map of the solution infiltration state is subjected to a texture structure semantic saliency measurement based on pixel granularity to obtain a solution infiltration state fine-grained texture structure semantic saliency distribution matrix, including: Performing pixel granularity feature decoupling on the local depth texture feature map of the solution infiltration state to obtain a set of solution infiltration state pixel granularity local depth texture decoupling feature vectors; Inputting each solution infiltration state pixel granularity local depth texture decoupling feature vector in the set of solution infiltration state pixel granularity local depth texture decoupling feature vectors into a hyperbolic space mapper to obtain a set of hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vectors; The fine-grained structure semantic saliency measurement coefficient of each hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vector in the set of the hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vector is calculated to obtain the solution infiltration state fine-grained texture structure semantic saliency distribution matrix formed by arranging multiple solution infiltration state fine-grained texture structure semantic saliency measurement coefficients.

6. The method for preparing the SCR denitration catalyst according to claim 5, characterized in that: The hyperbolic space mapper is used to use the first weight parameter matrix and the second weight parameter matrix to continuously linearly transform the solution infiltration state pixel granularity local depth texture decoupling feature vector to obtain the hyperbolic space modulated solution infiltration state pixel granularity local depth texture decoupling feature vector.

7. The method for preparing the SCR denitration catalyst according to claim 6, characterized in that: Based on the semantic saliency distribution matrix of the fine-grained texture structure of the solution infiltration state, the local depth texture feature map of the solution infiltration state is subjected to feature explicit hierarchical modulation to obtain the local depth texture enhanced feature map of the solution infiltration state, including: Inputting the semantic saliency distribution matrix of the fine-grained texture structure in the solution infiltration state into a feature explicit hierarchical precipitation module based on a multi-layer mask function to obtain a semantic saliency distribution weight matrix of the fine-grained texture structure in the solution infiltration state; The semantic significance distribution weight matrix of the fine-grained texture structure of the solution infiltration state is calculated and multiplied by each feature matrix along the channel dimension of the local depth texture feature map of the solution infiltration state at each position point to obtain the local depth texture enhancement feature map of the solution infiltration state.

8. The method for preparing the SCR denitration catalyst according to claim 7, characterized in that: Performing infiltration uniformity evaluation on each of the local depth texture enhancement feature maps in the set of local depth texture enhancement feature maps in the solution infiltration state to obtain a set of local infiltration uniformity score values, including: Each of the local depth texture enhancement feature maps in the solution infiltration state in the set of local depth texture enhancement feature maps in the solution infiltration state is respectively input into a uniformity score module based on a feedforward neural network model to obtain a set of local infiltration uniformity score values.

9. An SCR denitration catalyst, characterized in that: include: A solution infiltration state image acquisition module, used to acquire a solution infiltration state image collected by a camera; A local depth texture feature extraction module, used for extracting local depth texture features from the solution infiltration state image to obtain a set of local depth texture feature maps of the solution infiltration state; A feature-explicit hierarchical enhancement module is used to perform feature-explicit hierarchical enhancement on each of the solution-infiltrated state local depth texture feature maps in the set of solution-infiltrated state local depth texture feature maps to obtain a set of solution-infiltrated state local depth texture enhancement feature maps; A wetting uniformity evaluation module, used to respectively perform wetting uniformity evaluation on each of the solution infiltration state local depth texture enhancement feature maps in the set of solution infiltration state local depth texture enhancement feature maps to obtain a set of local infiltration uniformity score values; The wetting state uniformity judgment module is used to calculate the mean and variance of the set of local wetting state uniformity score values ​​of the solution, and determine whether the infiltration is uniform based on the mean and variance.

Images