Catalyst design scheme update iteration method and device based on multi-cluster algorithm

Abstract

The embodiment of the specification provides a catalyst design scheme updating iteration method and device based on a multi-clustering algorithm, the method comprises the following steps: collecting and obtaining an initial data set; adopting a K-means algorithm to classify the initial data set to obtain a first data set; constructing a density training and verification model through a DBSCAN algorithm to cluster the first data set to obtain a second data set; generating a third data set through a catalyst design system according to a customer demand parameter; determining a similarity value through similarity determination based on the second data set and the third data set; and determining a catalyst design scheme if the similarity value is greater than or equal to 95%. The catalyst parameters, the neighborhood radius and the core point threshold value in the DBSCAN algorithm are optimized through the set training condition, the similarity value reaching 95% is used as the redundancy value of whether to optimize the design software, the catalyst design scheme updating iteration is realized, and the performance of product operation is improved.

Description

Technical Field

[0001] This invention relates to the field of thermal power unit optimization technology, and in particular to a method and apparatus for updating and iterating catalyst design schemes based on a multi-clustering algorithm. Background Technology

[0002] As human understanding deepens, emission standards and technical specifications for thermal power units continue to improve, requiring corresponding upgrades to denitrification catalysts. In this process, product design and on-site environment are crucial. Therefore, it is necessary to compare design simulation results with actual conditions to obtain the optimal product formulation and solve problems such as non-compliance of flue gas emissions with standards.

[0003] In existing technologies, the ratio of various coal types to catalysts is continuously optimized based on problems encountered during production. However, the data used to train the catalyst design model is mostly randomly acquired. Therefore, how to reasonably optimize the catalyst design scheme to further reduce gas pollution during production is an urgent problem to be solved. Summary of the Invention

[0004] To overcome the problems existing in related technologies, this disclosure provides a method and apparatus for updating and iterating catalyst design schemes based on multi-clustering algorithms, in order to solve the technical problem of how to reasonably optimize the training data of catalyst design models to further reduce gas pollution in the production process.

[0005] This specification provides one or more embodiments of a catalyst design scheme update and iteration method based on a multi-clustering algorithm, including:

[0006] Step S101: Collect data on the composition of flue gas emitted after treatment by the flue gas treatment tower, and preprocess the data to obtain an initial dataset;

[0007] Step S102: Classify the initial dataset using the K-means algorithm to obtain the first dataset;

[0008] Step S103: Construct a density training and validation model using the DBSCAN algorithm, and use the density training and validation model to cluster the first dataset based on a preset neighborhood radius and a threshold for the number of data objects in the neighborhood to obtain the second dataset.

[0009] Step S104: Generate a third dataset using the catalyst design system based on customer requirement parameters;

[0010] Step S105: Determine the similarity value based on the second and third datasets;

[0011] Step S106: If the similarity value is ≥95%, a catalyst design scheme is obtained; if the similarity value reaches the preset radius and threshold correction conditions, radius and threshold correction threshold optimization is performed, and the process proceeds to step S103; if the similarity value reaches the preset catalyst parameter optimization conditions, catalyst parameter optimization design is performed, and the process proceeds to step S104.

[0012] This specification provides one or more embodiments of a catalyst design scheme update and iteration device based on a multi-clustering algorithm, including:

[0013] Data acquisition module: used to collect data on the composition of flue gas emitted after treatment by the flue gas treatment tower, and to preprocess the data to obtain an initial dataset;

[0014] Distance training module: The K-means algorithm is used to classify the initial dataset to obtain the first dataset;

[0015] Density training and validation module: used to build a density training and validation model using the DBSCAN algorithm, and to obtain a second dataset by clustering the first dataset based on a preset neighborhood radius and a threshold for the number of data objects in the neighborhood using the density training and validation model;

[0016] Catalyst design module: Used to generate a third dataset based on customer requirements parameters through the catalyst design system;

[0017] Similarity calculation module: used to calculate similarity between the second and third datasets to determine similarity values;

[0018] The judgment execution module is used to determine whether the similarity value calculated by the similarity calculation module is ≥95%, and thus obtain the catalyst design scheme; if the similarity value reaches the preset radius and threshold correction conditions, then the radius and threshold correction threshold optimization is performed and fed back to the density training and validation model; if the similarity value reaches the preset catalyst parameter optimization conditions, then the catalyst parameter optimization design is performed and fed back to the catalyst design module.

[0019] This specification provides one or more embodiments of a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the catalyst design scheme update and iteration method based on the multi-clustering algorithm described above.

[0020] This specification provides one or more embodiments of a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the catalyst design scheme update and iteration method based on the multi-clustering algorithm described above.

[0021] This embodiment uses a catalyst design scheme update and iteration method based on K-means and DBSCAN clustering algorithms. By setting training conditions (whether the similarity value reaches the preset radius, threshold correction threshold, and catalyst parameter optimization threshold), the catalyst parameters and the neighborhood radius and core point threshold in the DBSCAN algorithm are optimized. When the similarity value reaches 95%, it is used as a criterion for whether to optimize the redundancy value of the design software or optimize the theoretical value, thereby realizing the catalyst design scheme update and iteration, and improving the performance of product operation. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in one or more embodiments of this specification or in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating a catalyst design scheme update and iteration method based on a multi-clustering algorithm, provided for one or more embodiments of this specification;

[0024] Figure 2 A framework diagram of a catalyst design scheme update and iteration device based on a multi-clustering algorithm provided for one or more embodiments of this specification;

[0025] Figure 3 A schematic diagram illustrating the execution flow of a catalyst design scheme update and iteration device based on a multi-clustering algorithm, provided for one or more embodiments of this specification;

[0026] Figure 4 This is a schematic diagram of the structure of a computer provided for one or more embodiments of this specification. Detailed Implementation

[0027] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this invention.

[0028] The present invention will now be described in detail with reference to specific embodiments and accompanying drawings.

[0029] Method Implementation Examples

[0030] According to embodiments of the present invention, a catalyst design scheme update and iteration method based on a multi-clustering algorithm is provided, such as... Figure 1 The diagram shows a flowchart of a catalyst design scheme update and iteration method based on a multi-clustering algorithm provided in this embodiment. The catalyst design scheme update and iteration method based on a multi-clustering algorithm according to this embodiment includes:

[0031] Step S101: Collect data on the composition of flue gas emitted after treatment by the flue gas treatment tower, and preprocess the data to obtain an initial dataset;

[0032] Step S102: Use the K-means algorithm to classify the initial dataset to obtain the first dataset; for example, the distance between samples represents different intervals of flue gas content.

[0033] The first dataset contains different classification sets such as physical properties, mechanical properties, and thermal properties;

[0034] In this embodiment, the K-means algorithm is used to classify the initial dataset. Samples are classified according to the distance between them (different ranges of smoke content). After classification, training stops, forming a new first dataset. The specific calculation of the K-means algorithm is shown in the following formula:

[0035]

[0036] In the formula, J is the distortion function, and μ k The cluster centroid; using C (i) It refers to training example x (i) The distance from the cluster centroid, and the determination of x. (i) Which cluster centroid μ k The nearest cluster centroid is then assigned to it. k It is the location of the cluster centroid, μ k(i) This refers to the training example x that has already been assigned. (i) The location of the cluster centroid is determined. Through continuous correction using this distortion function, the first dataset with optimal data is obtained.

[0037] Step S103: Construct a density training and validation model using the DBSCAN algorithm, and use the density training and validation model to cluster the first dataset based on a preset neighborhood radius and a threshold for the number of data objects in the neighborhood to obtain the second dataset.

[0038] In this embodiment, a density training and validation model is constructed using the DBSCAN algorithm. The density training and validation model clusters the first dataset according to the defined neighborhood radius and the threshold number of data objects in the neighborhood, generating a corresponding training set. Training then stops, and the classification set is returned to the data warehouse to generate the second dataset. The DBSCAN algorithm includes two important parameters: core points and border points, specifically set as follows:

[0039] (1) Core point: For a dataset D, if the ε-neighborhood of sample p contains at least MinPts samples (including sample p), then sample p is called a core point;

[0040] N ε (p)≥MinPts (2)

[0041] Where ε-domain N ε The expression for (p) is:

[0042] N ε (p)={q∈D|dist(p,q)≤ε} (3)

[0043] (2) Boundary point: For a sample b that is not a core point, if b is within the ε-neighborhood of any core point p, then sample b is called a boundary point.

[0044] Right now:

[0045]

[0046] Step S104: Based on the customer requirement parameters, a third dataset is generated through the catalyst design system. In this embodiment, the customer requirement parameters mainly refer to the flue gas conditions that need to be achieved after using the catalyst, including fuel parameters, SCR inlet flue gas concentration parameters, etc. The catalyst design system designs raw materials, module quantity, etc., based on the performance guarantee parameters proposed by the customer.

[0047] Step S105: Determine the similarity value based on the second and third datasets;

[0048] Step S106: If the similarity value is ≥95%, a catalyst design scheme is obtained; if the similarity value reaches the preset radius and threshold correction conditions, radius and threshold correction and threshold optimization are performed, and the process proceeds to step S103; if the similarity value reaches the preset catalyst parameter optimization conditions, catalyst parameter optimization is performed, and the process proceeds to step S104.

[0049] This embodiment provides a catalyst design scheme update and iteration method based on K-means and DBSCAN clustering algorithms. By setting training conditions (whether the similarity value reaches the preset radius, threshold correction threshold, and catalyst parameter optimization threshold), the catalyst parameters and the neighborhood radius and core point threshold in the DBSCAN algorithm are optimized. When the similarity value reaches 95%, it is used as a criterion for whether to optimize the redundancy value of the design software or optimize the theoretical value, thereby realizing the catalyst design scheme update and iteration, and thus improving the performance of product operation.

[0050] This embodiment focuses on data, identifies optimization opportunities through data analysis and utilization, promotes interactive innovation and continuous optimization of data, technology, business processes and organizational structure, improves the level of data development and utilization, and provides support for the setting and management decisions of catalysts in thermal power units.

[0051] In this preferred embodiment, in order to further optimize the catalyst design scheme based on the similarity determination result, the catalyst parameter optimization condition in step S106 is 90% ≤ similarity value ≤ 95%;

[0052] In this embodiment, if there is an error of ±5% to ±10%, it is an error in the setting of catalyst design parameters. It is necessary to adjust the function parameters of the catalyst design system to improve the matching of the relevant parameters of the software design with the field environment. Therefore, based on the similarity value results, if the parameter error after similarity judgment between the second dataset and the third dataset is within 5%-10%, it is necessary to optimize the parameters set by the catalyst design system and update the third dataset with the optimized parameters of the catalyst design system.

[0053] In this preferred embodiment, in order to further optimize the catalyst design scheme based on the similarity determination result, the preset radius and threshold correction conditions in step S106 are set to similarity value ≤ 90%, then radius and threshold correction threshold optimization is performed, and the process proceeds to step S103.

[0054] In this embodiment, if the parameter error after similarity determination between the second and third datasets exceeds 15%, it indicates a problem in the processing of the first and second datasets. In this case, it is necessary to optimize the neighborhood radius and the threshold for the number of data objects in the neighborhood, and to find the optimal "neighborhood radius and threshold" through repeated training. That is, to complete a new data model based on the project overview. Therefore, based on the similarity value results, if the parameter error after similarity determination between the second and third datasets is 10% or more, it is necessary to re-plan the density training range of the density training and validation model, that is, to correct the neighborhood radius and threshold. The correction is mainly based on the number of datasets with the neighborhood radius. The initial optimization parameters need to be verified by a large number of experiments so that the average value of the output second dataset conforms to the normal scenario, and the relevant data is verified using a validation system.

[0055] System Implementation Examples

[0056] According to embodiments of the present invention, a catalyst design scheme update and iteration device based on a multi-clustering algorithm is provided, such as... Figure 2-3 The diagram shows a framework diagram and an execution flow diagram of a catalyst design scheme update and iteration device based on a multi-clustering algorithm, provided in this embodiment. According to this embodiment, the catalyst design scheme update and iteration device based on a multi-clustering algorithm includes:

[0057] Data acquisition module: used to collect data on the composition of flue gas emitted after treatment by the flue gas treatment tower, and to preprocess the data to obtain an initial dataset;

[0058] Distance training module: The K-means algorithm is used to classify the initial dataset to obtain the first dataset; for example, the distance between samples represents different intervals of smoke content.

[0059] Density training and validation model: used to build a density training and validation model using the DBSCAN algorithm. The density training and validation model clusters the first dataset based on a preset neighborhood radius and a threshold for the number of data objects in the neighborhood to obtain the second dataset.

[0060] Catalyst design module: Used to generate a third dataset based on customer requirements parameters through the catalyst design system;

[0061] Similarity calculation module: used to calculate similarity between the second and third datasets to determine similarity values;

[0062] The judgment execution module is used to determine whether the similarity value calculated by the similarity calculation module is ≥95%, and thus obtain the catalyst design scheme; if the similarity value reaches the preset radius and threshold correction threshold, then the radius and threshold correction threshold optimization is performed and fed back to the density training and validation model; if the similarity value reaches the preset catalyst parameter optimization threshold, then the catalyst parameter optimization is performed and fed back to the catalyst design module.

[0063] In this embodiment, the distance training module and density training and verification model use a catalyst design scheme update and iteration method based on K-means and DBSCAN clustering algorithms. By setting training conditions (whether the similarity value reaches the preset radius, threshold correction threshold, and catalyst parameter optimization threshold), the catalyst parameters and the neighborhood radius and core point threshold in the DBSCAN algorithm are optimized. When the similarity value reaches 95%, it is used as a criterion for whether to optimize the redundancy value of the design software or optimize the theoretical value, thereby realizing the catalyst design scheme update and iteration, and thus improving the performance of product operation.

[0064] In this embodiment, the preset radius and threshold correction conditions are 90% ≤ similarity value ≤ 95%;

[0065] The preset optimization condition for catalyst parameters is that the similarity value is ≤90%.

[0066] The embodiments of the present invention are system embodiments corresponding to the above method embodiments. The specific operations of each module processing step can be understood by referring to the description of the method embodiments, and will not be repeated here.

[0067] like Figure 4 As shown, the present invention also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the catalyst design scheme update and iteration method based on the multi-clustering algorithm in the above embodiments, or when the computer program is executed by a processor, it implements the catalyst design scheme update and iteration method based on the multi-clustering algorithm in the above embodiments. When the computer program is executed by the processor, it implements the following method steps:

[0068] Step S101: Collect data on the composition of flue gas emitted after treatment by the flue gas treatment tower, and preprocess the data to obtain an initial dataset;

[0069] Step S102: Use the K-means algorithm to classify the initial dataset to obtain the first dataset; for example, the distance between samples represents different intervals of flue gas content.

[0070] Step S103: Construct a density training and validation model using the DBSCAN algorithm, and use the density training and validation model to cluster the first dataset based on a preset neighborhood radius and a threshold for the number of data objects in the neighborhood to obtain the second dataset.

[0071] Step S104: Generate a third dataset using the catalyst design system based on customer requirement parameters;

[0072] Step S105: Determine the similarity value based on the second and third datasets;

[0073] Step S106: If the similarity value is ≥95%, a catalyst design scheme is obtained; if the similarity value reaches the preset radius and threshold correction conditions, radius and threshold correction threshold optimization is performed, and the process proceeds to step S103; if the similarity value reaches the preset catalyst parameter optimization conditions, catalyst parameter optimization design is performed, and the process proceeds to step S104.

[0074] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0075] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0076] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A catalyst design scheme update and iteration method based on multi-clustering algorithm, characterized in that, include: Step S101: Collect data on the composition of flue gas emitted after treatment by the flue gas treatment tower, and preprocess the data to obtain an initial dataset; Step S102: Classify the initial dataset using the K-means algorithm to obtain the first dataset; Step S103: Construct a density training and validation model using the DBSCAN algorithm, and use the density training and validation model to cluster the first dataset based on a preset neighborhood radius and a threshold for the number of data objects in the neighborhood to obtain the second dataset. Step S104: Generate a third dataset using the catalyst design system based on customer requirement parameters; Step S105: Determine the similarity value based on the second and third datasets; Step S106: If the similarity value is ≥95%, then the catalyst design scheme is obtained; If the similarity value reaches the preset radius and threshold correction conditions, then the radius and threshold correction threshold optimization is performed, and the process proceeds to step S103; if the similarity value reaches the preset catalyst parameter optimization conditions, then the catalyst parameter optimization design is performed, and the process proceeds to step S104. In step S106, the catalyst parameter optimization threshold condition is 90% ≤ similarity value ≤ 95%; The preset radius and threshold correction conditions in step S106 are similarity values ​​≤ 90%.

2. The catalyst design scheme update and iteration method based on multi-clustering algorithm as described in claim 1, characterized in that, Step S103 specifically involves: The DBSCAN algorithm is used to construct a density training and validation model. The density training and validation model is used to cluster the first dataset according to the neighborhood radius and the threshold of the number of data objects in the neighborhood when the density is defined, and the corresponding training set is generated. Then the training stops and the classification set is put back into the data warehouse to generate the second dataset.

3. A catalyst design scheme update and iteration device based on multi-clustering algorithm, characterized in that, include: Data acquisition module: used to collect data on the composition of flue gas emitted after treatment by the flue gas treatment tower, and to preprocess the data to obtain an initial dataset; Distance training module: The K-means algorithm is used to classify the initial dataset to obtain the first dataset; Density training and validation module: used to build a density training and validation model using the DBSCAN algorithm, and to obtain a second dataset by clustering the first dataset based on a preset neighborhood radius and a threshold for the number of data objects in the neighborhood using the density training and validation model; Catalyst design module: Used to generate a third dataset based on customer requirements parameters through the catalyst design system; Similarity calculation module: used to calculate the similarity between the second and third datasets to determine the similarity value; Judgment Execution Module: Used to determine if the similarity value calculated by the similarity calculation module is ≥95%, and then obtain the catalyst design scheme; If the similarity value reaches the preset radius and threshold correction conditions, then the radius and threshold correction threshold optimization is performed and fed back to the density training and validation model; if the similarity value reaches the preset catalyst parameter optimization conditions, then the catalyst parameter optimization design is performed and fed back to the catalyst design module. The preset radius and threshold correction conditions are that the similarity value is ≤90%; The preset catalyst parameter optimization conditions are 90% ≤ similarity value ≤ 95%.

4. The catalyst design scheme update and iteration device based on multi-clustering algorithm as described in claim 3, characterized in that, The density training and validation module constructs a density training and validation model using the DBSCAN algorithm. The density training and validation model clusters the first dataset according to the neighborhood radius and the threshold of the number of data objects in the neighborhood when the density is defined, generating a corresponding training set. Then, training stops, and the classification set is put back into the data warehouse to generate the second dataset.

5. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the catalyst design scheme update and iteration method based on the multi-clustering algorithm as described in any one of claims 1 to 2.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the catalyst design scheme update and iteration method based on the multi-clustering algorithm as described in any one of claims 1 to 2.

Images