Intelligent anti-blocking device and method for phosphogypsum calcining furnace

Abstract

The present invention provides an intelligent anti-blocking device and method for a phosphogypsum calcining furnace, which relates to the technical field of calcining furnaces, and includes: a monitoring demand analysis module for determining a set of demand parameters for sensors at key positions; a working data stream acquisition module for obtaining a calcining furnace monitoring node network and obtaining a set of working data streams at key positions of the calcining furnace; a distributed blockage assessment module for constructing a set of node data processing modules and obtaining a set of key node blockage characteristic parameters; an association matching module for constructing a calcining furnace blockage characteristic tree and determining the characteristic parameters of the cause of the calcining furnace blockage; and an anti-blockage warning control module for obtaining the anti-blockage coordination control parameters of the calcining furnace and performing anti-blockage warning control. The present invention solves the technical problems that the prior art cannot obtain calcining furnace data in real time and accurately, lacks accurate analysis means for a large amount of data, and is difficult to determine the specific cause and location of the blockage, resulting in delayed blockage reaction and poor anti-blockage effect.

Description

Technical Field

[0001] The invention relates to the technical field of calcining furnaces, and in particular to an intelligent anti-blocking device and method for a phosphogypsum calcining furnace. Background Art

[0002] Phosphogypsum calciners are widely used in industrial production, especially in the fields of phosphate fertilizer, metallurgy and fertilizer, for processing raw materials such as phosphogypsum. These calciners operate under high temperature and complex reaction conditions, and require effective anti-blocking technology to ensure the stability and safety of production. However, in phosphogypsum calciners and other similar equipment, the prediction and identification of blockage problems usually rely on manual experience or regular inspections. This method may not be able to detect hidden dangers in time, resulting in equipment shutdown or production interruption. In addition, traditional monitoring systems are unable to process and analyze large amounts of real-time data, resulting in the inability to identify potential problems in a timely manner and difficulty in accurately locating the cause of the blockage. Usually, they can only be checked through indirect indicators or after the equipment is shut down, resulting in delayed blockage response and poor anti-blocking effect. Summary of the invention

[0003] The present application provides an intelligent anti-blocking device and method for a phosphogypsum calcining furnace, aiming to solve the technical problems that the prior art is unable to obtain calcining furnace data in real time and accurately, lacks accurate analysis means for a large amount of data, and is difficult to determine the specific cause and location of the blockage, resulting in delayed blockage reaction and poor anti-blocking effect.

[0004] The first aspect disclosed in the present application provides an intelligent anti-blocking device for a phosphogypsum calcining furnace, the device comprising: a monitoring demand analysis module, the monitoring demand analysis module is used to obtain the structural design information of the target phosphogypsum calcining furnace, identify the key positions of the structural design information, obtain the key position set of the calcining furnace, perform monitoring demand analysis based on the key position set of the calcining furnace, and determine the key position sensor requirement parameter set; a working data stream acquisition module, the working data stream acquisition module is used to respectively perform monitoring network deployment on the key position set of the calcining furnace according to the key position sensor requirement parameter set, obtain the calcining furnace monitoring node network, and acquire the calcining furnace key position working data stream set based on the calcining furnace monitoring node network; a distributed congestion assessment module, the distributed congestion assessment module is used to A node data processing module set is constructed according to the calciner monitoring node network, and a distributed congestion assessment is performed on the working data stream set of the key position of the calciner through the node data processing module set to obtain a key node congestion characteristic parameter set; an association matching module is used to construct a calciner congestion characteristic tree, and an association matching is performed based on the key node congestion characteristic parameter set and the calciner congestion characteristic tree to determine the calciner congestion cause characteristic parameters; an anti-blocking warning control module is used to trigger an anti-blocking strategy mechanism to perform anti-blocking control analysis on the calciner congestion cause characteristic parameters and the key node congestion characteristic parameter set to obtain a calciner anti-blocking coordination control parameter, and an anti-blocking warning control is performed on the target phosphogypsum calciner based on the calciner anti-blocking coordination control parameter.

[0005] The second aspect disclosed in the present application provides an intelligent anti-blocking method for a phosphogypsum calcining furnace, the method is implemented based on the above-mentioned intelligent anti-blocking device for a phosphogypsum calcining furnace, the method comprises: obtaining structural design information of a target phosphogypsum calcining furnace, identifying key positions of the structural design information, obtaining a set of key positions of the calcining furnace, performing monitoring demand analysis based on the set of key positions of the calcining furnace, and determining a set of sensor demand parameters for key positions; performing monitoring network deployment on the set of key positions of the calcining furnace respectively according to the set of sensor demand parameters for key positions, obtaining a calcining furnace monitoring node network, and acquiring a set of working data streams of key positions of the calcining furnace based on the collection of working data streams of key positions of the calcining furnace; and performing monitoring demand analysis on the set of key positions of the calcining furnace according to the set of sensor demand parameters for key positions. The calciner monitoring node network constructs a node data processing module set, and performs distributed congestion assessment on the working data flow set of the key position of the calciner through the node data processing module set to obtain a key node congestion characteristic parameter set; constructs a calciner congestion characteristic tree, and performs association matching with the calciner congestion characteristic tree based on the key node congestion characteristic parameter set to determine the calciner congestion cause characteristic parameters; triggers the anti-blocking strategy mechanism to perform anti-blocking control analysis on the calciner congestion cause characteristic parameters and the key node congestion characteristic parameter set to obtain the calciner anti-blocking coordination control parameters, and performs anti-blocking early warning control on the target phosphogypsum calciner based on the calciner anti-blocking coordination control parameters.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] By analyzing the structural design information of the phosphogypsum calcining furnace, the key positions in the calcining furnace, such as the feed port, calcining area, and discharge port, are identified. Based on the identification results of the key positions, the monitoring requirements are analyzed and the required sensor types and specifications are determined. This ensures that suitable sensors are deployed at key positions, which can effectively monitor key parameters and improve the monitoring capability and response speed of the calcining furnace status. According to the sensor requirement parameter set, a monitoring network is deployed to collect working data streams at key positions through the monitoring network to ensure that comprehensive and real-time operating data is obtained, which is helpful for analyzing the actual working status and operating trends of the calcining furnace. A node data processing module set is constructed to efficiently process the working data streams obtained from the monitoring network, improve the efficiency and accuracy of data processing, and perform distributed congestion assessment on the working data at key positions, which can identify and analyze possible congestion points and obtain key The blockage characteristic parameters of key nodes can realize early detection of potential blockage problems and avoid affecting the operation of the entire calciner; construct a blockage characteristic tree of the calciner. Through the characteristic tree, we can clearly understand the blockage characteristics of different types and their associations. By associating and matching the blockage characteristic parameters of key nodes with the blockage characteristic tree, we can determine the specific blockage cause characteristic parameters of the calciner, which helps to accurately locate the root cause of the blockage and provide a basis for the subsequent formulation of anti-blocking strategies; trigger the anti-blocking strategy mechanism, match the blockage cause characteristic parameters with the anti-blocking strategy library, and determine the appropriate anti-blocking strategy plan. We can formulate anti-blocking measures in a targeted manner and improve the anti-blocking effect. Based on the anti-blocking strategy plan, we can optimize parameters and obtain anti-blocking coordination control parameters to carry out anti-blocking early warning control of the calciner. Through optimized control parameters, we can effectively prevent blockage problems and improve the overall operation efficiency and stability of the calciner.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A schematic structural diagram of an intelligent anti-blocking device for a phosphogypsum calcining furnace provided in an embodiment of the present application;

[0010] Figure 2 A schematic flow chart of an intelligent anti-blocking method for a phosphogypsum calcining furnace provided in an embodiment of the present application.

[0011] Explanation of the accompanying drawings: monitoring demand analysis module 10, working data flow acquisition module 20, distributed congestion assessment module 30, association matching module 40, anti-congestion warning control module 50. DETAILED DESCRIPTION

[0012] The embodiment of the present application provides an intelligent anti-blocking device and method for a phosphogypsum calcining furnace, thereby solving the technical problems that the prior art is unable to obtain calcining furnace data in real time and accurately, lacks accurate analysis means for a large amount of data, and is difficult to determine the specific cause and location of the blockage, resulting in delayed blockage reaction and poor anti-blocking effect.

[0013] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically introduced below in conjunction with the accompanying drawings of the specification. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0014] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides an intelligent anti-blocking device for a phosphogypsum calcining furnace, the device comprising:

[0015] A monitoring demand analysis module 10, the monitoring demand analysis module 10 is used to obtain the structural design information of the target phosphogypsum calcining furnace, identify the key positions of the structural design information, obtain the key position set of the calcining furnace, perform monitoring demand analysis based on the key position set of the calcining furnace, and determine the key position sensor requirement parameter set; a working data flow acquisition module 20, the working data flow acquisition module 20 is used to respectively perform monitoring network deployment on the key position set of the calcining furnace according to the key position sensor requirement parameter set, obtain the calcining furnace monitoring node network, and acquire the calcining furnace key position working data flow set based on the calcining furnace monitoring node network; a distributed congestion assessment module 30, the distributed congestion assessment module 30 is used to construct a node according to the calcining furnace monitoring node network A data processing module set, which performs distributed congestion assessment on the working data stream set at the key position of the calciner through the node data processing module set to obtain a set of key node congestion characteristic parameters; an association matching module 40, which is used to construct a calciner congestion characteristic tree, and performs association matching with the calciner congestion characteristic tree based on the key node congestion characteristic parameter set to determine the characteristic parameters of the cause of congestion of the calciner; an anti-blocking warning control module 50, which is used to trigger an anti-blocking strategy mechanism to perform anti-blocking control analysis on the characteristic parameters of the cause of congestion of the calciner and the set of key node congestion characteristic parameters, obtain calciner anti-blocking coordination control parameters, and perform anti-blocking warning control on the target phosphogypsum calciner based on the calciner anti-blocking coordination control parameters.

[0016] Furthermore, the distributed congestion assessment module 30 comprises the following steps:

[0017] Data attributes are extracted from each monitoring node in the calcining furnace monitoring node network to obtain a node collection data attribute information set; data preprocessing step analysis is performed based on the node collection data attribute information set to determine a node data preprocessing multi-channel set; data mining is performed on each monitoring node to obtain a node working data set, and a node congestion assessment model set is constructed based on the node working data set; based on the node data preprocessing multi-channel set and the node congestion assessment model set, a data processing program is loaded on each monitoring node to construct the node data processing module set.

[0018] Furthermore, the distributed congestion assessment module 30 comprises the following steps:

[0019] According to the node collection data attribute information set, a node data attribute threshold set is determined, and outlier identification is performed on the node working data set according to the node data attribute threshold set to obtain a node abnormal working attribute data set; based on the node abnormal working attribute data set, a congestion feature association assessment is performed to determine an abnormal data congestion risk assessment parameter set; the node working data set is identified according to the abnormal data congestion risk assessment parameter set to obtain a node working congestion assessment data set; congestion assessment training is performed on the node working congestion assessment data set to construct the node congestion assessment model set.

[0020] Furthermore, the association matching module 40 includes the following steps:

[0021] Obtain the calciner blockage characteristic factor information, which includes the blockage node location, node blockage risk parameters and corresponding blockage generation causes; perform associated topological identification on the phosphogypsum calciner blockage fault library based on the calciner blockage characteristic factor information to obtain the calciner blockage characteristic fault library; define the calciner blockage fault top node, classify the blockage fault causes of the calciner blockage characteristic fault library based on the calciner blockage fault top node, and obtain the calciner first-level blockage cause node set; perform blockage feature cascade analysis based on the calciner first-level blockage cause node set to construct the calciner blockage characteristic tree.

[0022] Furthermore, the association matching module 40 includes the following steps:

[0023] The first-level blockage cause node set of the calcining furnace is respectively subjected to blockage feature correlation analysis to obtain a first-level node blockage feature parameter set; based on the first-level node blockage feature parameter set, fault cause subdivision association is performed to construct a second-level blockage cause node set of the calcining furnace; the second-level blockage cause node set of the calcining furnace is sequentially subjected to blockage association analysis until a K-level blockage cause node set of the calcining furnace is obtained, wherein K≥2; based on the top node of the calcining furnace blockage fault, the first-level blockage cause node set of the calcining furnace and the second-level blockage cause node set of the calcining furnace, cascade deep clustering is performed until the K-level blockage cause node set of the calcining furnace is obtained to obtain the calcining furnace blockage feature tree.

[0024] Furthermore, the anti-blocking warning control module 50 includes the following steps:

[0025] The anti-blocking strategy mechanism is triggered to call and obtain the anti-blocking strategy library of the phosphogypsum calcining furnace; the anti-blocking strategy scheme of the target calcining furnace is determined based on the characteristic parameters of the cause of the calcining furnace blockage and the anti-blocking strategy library of the phosphogypsum calcining furnace; the control parameters of the anti-blocking strategy scheme of the target calcining furnace are analyzed based on the set of key node blockage characteristic parameters to obtain the optimization threshold of the anti-blocking control parameters; the anti-blocking effect fitness function is constructed, and a global optimization is performed within the optimization threshold of the anti-blocking control parameters based on the anti-blocking effect fitness function to obtain the anti-blocking coordination control parameters of the calcining furnace.

[0026] Furthermore, the anti-blocking warning control module 50 includes the following steps:

[0027] A plurality of anti-blocking control parameters are randomly selected within the anti-blocking control parameter optimization threshold, and the plurality of anti-blocking control parameters are calculated and evaluated using the anti-blocking effect fitness function to obtain a plurality of anti-blocking control effect fitnesses; the plurality of anti-blocking control parameters are clustered and optimized according to the plurality of anti-blocking control effect fitnesses to obtain an anti-blocking control parameter cluster, the cluster center of the anti-blocking control parameter cluster being the anti-blocking control parameter with the largest fitness; cross-mutation update and iterative optimization evaluation are performed within the anti-blocking control parameter cluster until a preset convergence requirement is reached to obtain the anti-blocking coordination control parameters of the calcining furnace.

[0028] Through the subsequent detailed description of an intelligent anti-blocking method for a phosphogypsum calcining furnace in this specification, those skilled in the art can clearly understand an intelligent anti-blocking device for a phosphogypsum calcining furnace in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0029] Embodiment 2, based on the same inventive concept as the intelligent anti-blocking device for phosphogypsum calcining furnace in the above embodiment, Figure 2As shown, the embodiment of the present application provides an intelligent anti-blocking method for a phosphogypsum calcining furnace, the method comprising:

[0030] Acquire the structural design information of the target phosphogypsum calcining furnace, identify key positions of the structural design information, obtain a set of key positions of the calcining furnace, perform monitoring demand analysis based on the set of key positions of the calcining furnace, and determine a set of sensor demand parameters for key positions;

[0031] A phosphogypsum calcining furnace is an industrial furnace used to calcine phosphogypsum. The structural design information is obtained through the design drawings, structural models, operation manuals, etc. of the target phosphogypsum calcining furnace. This information includes the size, shape, internal structure, etc. of the furnace body. The key positions of the structural design information are identified. The key positions are areas where blockages are prone to occur during the flow or reaction of materials inside the calcining furnace, such as the feed port, calcination zone, and discharge port. The feed port is the entrance for materials to enter the furnace. Due to the nature of the material or uneven supply, blockages are prone to occur. The calcination zone is the area where the calcination reaction is most intense, with a high temperature, and the material is prone to agglomeration, resulting in blockage. The discharge port is the outlet for materials to be discharged from the furnace. If the material is not completely calcined or the particle size is large, it is easy to cause blockage. These key positions are summarized as a set of key positions of the calcining furnace.

[0032] Conduct monitoring demand analysis, including sensor type analysis and sensor specification analysis. Specifically, select the appropriate sensor type according to the characteristics of each key position. For example, a flow sensor is needed at the feed port to detect the material flow state, while a temperature sensor and a pressure sensor are needed in the calcination area to monitor the material reaction and gas pressure. Select the sensor range, sensitivity, accuracy, response speed and other specifications according to the characteristics of each key position. For example, sensors working in high temperature environments need to have high temperature tolerance, while sensors used to detect material flow need to have sufficient sensitivity to identify flow changes. Through the above analysis, a set of sensor demand parameters for key positions of the calciner is finally formed.

[0033] According to the key position sensor requirement parameter set, monitoring network deployment is performed on the key position set of the calciner respectively to obtain a calciner monitoring node network, and based on the calciner monitoring node network, a working data stream set of the key position of the calciner is acquired;

[0034] According to the determined sensor type and specifications, the sensors are installed at key locations of the calciner, such as the feed port, calcination area, and discharge port. A network topology is established, and the installed sensors are connected to a unified monitoring network. Each monitoring node, that is, the sensor installed at the key location, is configured. The configuration process includes setting the sensor's sampling frequency, data transmission rate, and assigning a unique identifier to each node to form a complete calciner monitoring node network.

[0035] After the monitoring node network is deployed, each sensor collects real-time data at key locations according to the configuration, such as temperature, pressure, flow, etc. By summarizing the data collected by different sensors, a complete set of working data streams at key locations of the calciner is formed. These data streams contain time series information, which can reflect the operating status of the calciner and potential blockage risks, and serve as the basis for subsequent analysis.

[0036] A node data processing module set is constructed according to the calciner monitoring node network, and a distributed congestion assessment is performed on a set of working data streams at key positions of the calciner through the node data processing module set to obtain a set of key node congestion characteristic parameters;

[0037] Corresponding data processing modules are constructed for the data stream of each monitoring node. The main task of these modules is to analyze the collected data and identify possible blockages in the key positions of the calciner. Specifically, the node data processing module is designed according to the data type and data characteristics collected by the node. For example, image data needs to be filtered and convolution features extracted, etc. Multiple data processing modules are integrated to form a complete set of node data processing modules. This set can cover the data processing requirements of all key positions of the calciner and provide comprehensive data support for the blockage assessment of the entire calcination system.

[0038] Through a set of node data processing modules, a distributed blockage assessment is performed on the working data flow set at key positions of the calciner. Distributed assessment means that the data processing module of each node can independently complete preliminary data analysis locally, reducing the burden of data transmission and centralized processing. Specifically, the data processing module of each node identifies potential blockage features based on the preprocessed data. These features include abnormal temperature increase, sudden pressure drop, unstable flow, etc. These blockage features can be used to determine whether there are abnormalities through set thresholds. At the local node, the processing module performs a local assessment of the identified blockage features to determine their severity and possibility. For example, if the temperature continues to exceed the normal range for a period of time, it is assessed as a high-risk blockage feature. The assessment results of multiple nodes are integrated to obtain a set of key node blockage feature parameters.

[0039] Constructing a calciner blockage feature tree, and performing correlation matching based on the key node blockage feature parameter set and the calciner blockage feature tree to determine the calciner blockage cause feature parameters;

[0040] Obtain the characteristic factor information related to the blockage, including the location where the blockage occurs, the blockage risk parameters of each node, and the specific reasons that may cause these blockages. Based on the blockage characteristic factor information, build a calciner blockage fault library, which records various blockage situations and their corresponding causes, and associates these causes with the actual blockage phenomenon through the topological structure. For example, if the temperature at a certain location continues to rise, it may be caused by overheating of the material or uneven heat distribution.

[0041] Define the top node. In the feature tree, the top node usually represents the most serious or most extensive blockage situation. The nodes below the top node represent more specific and detailed blockage causes. The top node is specifically the top-level event of calciner blockage.

[0042] The factors in the blockage fault library are divided into different levels according to severity or scope of impact, forming a set of first-level, second-level or even deeper blockage cause nodes. There are usually causal or dependent relationships between these nodes, forming a complete calciner blockage feature tree. Through the feature tree, the cause of the blockage can be analyzed and judged more intuitively.

[0043] Use the feature matching algorithm to compare the key node blockage feature parameters with the nodes in the feature tree one by one. For example, if the temperature and pressure parameters of a node are abnormal at the same time, they can be matched with the corresponding nodes in the feature tree to find out the possible causes of the blockage. The matching algorithm is used to find the blockage cause node that best suits the current situation, determine the specific cause of the current blockage, and mark it as the blockage cause feature parameter of the node. These parameters reflect the causes of blockage in actual operations, which may include equipment aging, unqualified material quality, operational errors, etc.

[0044] The anti-blocking strategy mechanism is triggered to perform anti-blocking control analysis on the characteristic parameters of the calcining furnace blockage cause and the set of key node blockage characteristic parameters, obtain the calcining furnace anti-blocking coordination control parameters, and perform anti-blocking early warning control on the target phosphogypsum calcining furnace based on the calcining furnace anti-blocking coordination control parameters.

[0045] Automatically call the pre-set anti-blocking strategy library for phosphogypsum calcining furnaces. This strategy library contains a series of solutions and countermeasures for different blockage situations, including adjusting material ratios, optimizing calcination parameters, and starting emergency dredging equipment. According to the characteristic parameters of the currently detected blockage cause, it is matched with the preset strategies in the strategy library. Through the matching algorithm, the most suitable anti-blocking strategy solution for the current blockage situation is automatically selected. For example, if a certain position is blocked due to material accumulation, choose to increase the heat supply at that position or increase the material flow rate at that position to solve the blockage problem.

[0046] After determining the anti-blocking strategy scheme, the control parameters of the scheme are analyzed according to the set of key node blockage characteristic parameters to determine the optimized threshold range of these control parameters. For example, an optimal temperature adjustment range is determined based on historical data, within which blockage can be effectively avoided. Random selection of parameters, clustering optimization, cross-mutation update and other operations are performed within the optimized threshold range, and the anti-blocking effect is evaluated through the fitness function. The fitness function can be based on the evaluation criteria of the anti-blocking effect, including anti-blocking efficiency, cost, anti-blocking quality, etc., to ensure that the selected parameter combination can achieve the best anti-blocking effect in actual application. According to the fitness evaluation results, the parameter combination with the highest fitness is selected as the calciner anti-blocking coordination control parameter.

[0047] Furthermore, the node data processing module set is constructed according to the calcining furnace monitoring node network, including:

[0048] Data attributes are extracted from each monitoring node in the calcining furnace monitoring node network to obtain a node collection data attribute information set; data preprocessing step analysis is performed based on the node collection data attribute information set to determine a node data preprocessing multi-channel set; data mining is performed on each monitoring node to obtain a node working data set, and a node congestion assessment model set is constructed based on the node working data set; based on the node data preprocessing multi-channel set and the node congestion assessment model set, a data processing program is loaded on each monitoring node to construct the node data processing module set.

[0049] Each monitoring node will continuously collect working data at its location. These data include temperature, pressure, material flow rate, image information and other types. Different types of data have different processing requirements. Therefore, it is necessary to determine the appropriate preprocessing method based on the attributes of the data. Specifically, the data collected by each node is identified by attributes to determine which type the data belongs to. Common data types include time series data (such as temperature and pressure changes), image data (such as material stacking images), and status data (such as equipment operating status). The attribute information set of the node collected data is obtained based on the attribute identification results. Through data attribute identification, it can be clarified how to perform subsequent processing on this data.

[0050] Different data types, such as time series data, image data, state data, etc., require different preprocessing methods. For example, for time series data, the preprocessing steps include denoising, smoothing, differential processing, normalization, etc.; for image data, the preprocessing steps include image filtering, edge detection, image enhancement, convolution feature extraction, etc., which help to remove noise and extract key features; for state data, the preprocessing steps include data discretization, outlier detection and processing, missing value filling, etc., to ensure the continuity and accuracy of the state data.

[0051] According to the results of the preprocessing step analysis, multiple data preprocessing channels are defined. Each channel corresponds to one or more data types and includes a series of specific preprocessing steps. For example, the time series data channel includes denoising, smoothing, normalization, etc., and the image data channel includes filtering, convolution feature extraction, etc. After the preprocessing channels are defined, parameters are configured for each channel, including the type and size of the filter, the size of the convolution kernel, the size of the smoothing window, etc. After all channels are configured, they are integrated to obtain a multi-channel set of node data preprocessing.

[0052] Data mining is performed on each monitoring node, including feature extraction, pattern recognition, and anomaly detection. According to the mined features, the data set is divided into different categories, such as normal operation data set, abnormal operation data set, etc., and the data points in each data set are labeled to indicate their corresponding operating status or congestion risk level to obtain the node working data set.

[0053] Select a suitable evaluation model based on the characteristics of the node data. Commonly used models include decision trees, support vector machines, random forests, neural networks, etc. Use the node working data set to train the selected model so that it can identify and predict different operating states and congestion risks. The training process is iterated multiple times to optimize the accuracy of the model. The performance of the model is evaluated through methods such as cross-validation, and the model performance is optimized by adjusting model parameters and selecting the best feature combination to finally obtain a set of node congestion evaluation models.

[0054] The node data preprocessing multi-channel set is organically combined with the congestion assessment model set to form a complete data processing program, which can automatically process the data collected from the monitoring nodes and output the congestion assessment results. The data processing program is loaded onto the edge computing device of each monitoring node so that it can process the collected data in real time and perform on-site assessment to form the node data processing module set.

[0055] Furthermore, the step of constructing a node congestion assessment model set based on the node working data set includes:

[0056] According to the node collection data attribute information set, a node data attribute threshold set is determined, and outlier identification is performed on the node working data set according to the node data attribute threshold set to obtain a node abnormal working attribute data set; based on the node abnormal working attribute data set, a congestion feature association assessment is performed to determine an abnormal data congestion risk assessment parameter set; the node working data set is identified according to the abnormal data congestion risk assessment parameter set to obtain a node working congestion assessment data set; congestion assessment training is performed on the node working congestion assessment data set to construct the node congestion assessment model set.

[0057] According to the data attributes collected by each monitoring node, the normal working range of each attribute is analyzed, the standard working threshold of each data attribute is determined, and the node data attribute threshold set is determined. For example, the feed port needs to monitor pressure and temperature, so the standard working thresholds of these parameters can be set by referring to the calciner design specifications, industry standards or historical data. The collected node working data is compared with the determined node data attribute threshold. If the working data of a node exceeds the corresponding data attribute threshold range, the data is identified as an abnormal value. The identified abnormal values ​​are integrated to obtain the node abnormal working attribute data set.

[0058] Relevant features are extracted from the node abnormal working attribute data set. These features include the type of abnormal data, frequency of occurrence, duration, related operating conditions, etc. Statistical methods or data mining techniques, such as correlation analysis, regression analysis, cluster analysis, etc., are used to evaluate the relationship between abnormal data and congestion characteristics, and analyze which abnormal data features are significantly associated with congestion events. Based on the results of the association analysis, the congestion risk of each abnormal data is evaluated. For example, according to the strength of the association between the abnormal data and the congestion event, a risk level is assigned to each abnormal data, such as low risk, medium risk or high risk. The evaluation results are summarized to form an abnormal data congestion risk assessment parameter set, which contains the risk assessment results of all node abnormal data.

[0059] The node working data set is identified according to the abnormal data congestion risk assessment parameter set. Specifically, the assessment parameters are used to determine which node working data meets or exceeds the set risk threshold. For example, if the pressure data of a certain node exceeds the preset risk threshold, the data is marked as potential congestion risk data. This means that the node working data set is classified and labeled to obtain a new data set with identification, namely the node working congestion assessment data set, which includes all node data identified as having a congestion risk.

[0060] Use the identified node work congestion assessment data set to prepare a training data set, which includes the work data identified as congestion risks and their related features, as well as the actual congestion conditions corresponding to these features. Select a suitable machine learning model for training, such as decision trees, support vector machines, random forests, neural networks, etc. Use the node work data set to train the selected model so that it can identify and predict different operating states and congestion risks. The training process is iterated multiple times to optimize the accuracy of the model. The performance of the model is evaluated through methods such as cross-validation, and the model performance is optimized by adjusting model parameters and selecting the best feature combination. Finally, a node congestion assessment model set is obtained, which is used to monitor and evaluate the congestion risk of the node in real time.

[0061] Furthermore, the construction of the calciner blockage feature tree includes:

[0062] Obtain the calciner blockage characteristic factor information, which includes the blockage node location, node blockage risk parameters and corresponding blockage generation causes; perform associated topological identification on the phosphogypsum calciner blockage fault library based on the calciner blockage characteristic factor information to obtain the calciner blockage characteristic fault library; define the calciner blockage fault top node, classify the blockage fault causes of the calciner blockage characteristic fault library based on the calciner blockage fault top node, and obtain the calciner first-level blockage cause node set; perform blockage feature cascade analysis based on the calciner first-level blockage cause node set to construct the calciner blockage characteristic tree.

[0063] Obtain information on characteristic factors of calciner blockage, including blockage node locations, node blockage risk parameters, and corresponding blockage causes, wherein the blockage node locations are key node locations where blockage may occur in the calciner, such as the feed port, calcination area, and discharge port, etc.; node blockage risk parameters include pressure, temperature, flow rate, speed, etc., which have a threshold range under normal operating conditions; blockage causes include material accumulation, equipment wear, improper operation, etc.

[0064] Obtain a phosphogypsum calcining furnace blockage fault library, which includes all blockage fault events in the historical records. By associating the acquired blockage characteristic factor information with the fault library, identify the specific node blockage location and fault risk that may be caused by each blockage cause. Based on the association results, establish a calcining furnace blockage characteristic fault library, which includes a description of each blockage cause, the location of the affected node, and the associated risk parameters.

[0065] The top node of the calciner blockage fault is defined. In the feature tree, the top node usually represents the most serious or widespread blockage situation. The nodes at each level below the top node represent more specific and detailed blockage causes. In this embodiment, it is specifically the top-level event of calciner blockage. Based on the top node, all blockage features in the fault library are classified into blockage fault causes, and similar blockage causes are classified under corresponding nodes to form a set of first-level blockage cause nodes. These first-level nodes represent the largest category of blockage causes, such as improper temperature control, gas flow, etc.

[0066] Starting from each primary blockage cause node, trace the secondary problems that may be caused. For example, improper temperature control may lead to heat source failure, and fan failure may further affect gas flow problems. According to the cascade analysis results, a hierarchical blockage feature tree is constructed. The feature tree starts from the top node and expands downward, layer by layer. Each layer represents different blockage causes and their interrelationships.

[0067] Furthermore, the construction of the calciner blockage feature tree includes:

[0068] The first-level blockage cause node set of the calcining furnace is respectively subjected to blockage feature correlation analysis to obtain a first-level node blockage feature parameter set; based on the first-level node blockage feature parameter set, fault cause subdivision association is performed to construct a second-level blockage cause node set of the calcining furnace; the second-level blockage cause node set of the calcining furnace is sequentially subjected to blockage association analysis until a K-level blockage cause node set of the calcining furnace is obtained, wherein K≥2; based on the top node of the calcining furnace blockage fault, the first-level blockage cause node set of the calcining furnace and the second-level blockage cause node set of the calcining furnace, cascade deep clustering is performed until the K-level blockage cause node set of the calcining furnace is obtained to obtain the calcining furnace blockage feature tree.

[0069] Obtain relevant data of each primary blockage cause node, covering the specific blockage problems caused by it, and compare it with the actual monitoring data of the calciner to analyze the blockage characteristics of each primary node, including the performance of different nodes when blockage occurs. For example, improper temperature control may cause agglomeration blockage in the relevant calcination area, wall blockage of the discharge port, etc. For example, gas flow problems may cause uneven gas distribution and cold spot blockage in the calcination area, etc. According to the analysis results, a set of primary node blockage characteristic parameters is obtained.

[0070] Based on the set of first-level node blockage characteristic parameters, the blockage characteristics of each first-level node are subdivided and associated with specific reasons. For example, the blockage of the discharge port due to wall sticking is caused by the furnace temperature being too high, and the blockage of the calcination zone due to the furnace temperature being too low. According to the association results, under the first-level node, the second-level blockage cause nodes are added. These second-level nodes specifically describe the different blockage modes under the first-level reasons.

[0071] By analyzing the relationship between each secondary node and establishing them with lower-level nodes, such as analyzing how high furnace temperature leads to different types of blockages, each secondary node is subdivided into three-level nodes. For example, if the secondary node indicates that the temperature is too high and causes the discharge port to stick to the wall, the impact of different temperature ranges on the degree of sticking can be further analyzed, and so on, until the required K-level node is reached. After completing the node analysis of all levels, a complete K-level blockage cause node set is formed, K≥2, and each node represents a specific blockage cause and contains its detailed description and impact.

[0072] The first-level, second-level nodes and even K-level nodes are connected to the top node through a cascade relationship to form a hierarchical fault model. Each node is connected through clear association rules to show the complete chain from the bottom-level fault to the top-level cause. The nodes at all levels are deeply clustered. By analyzing the similarities and relationships between nodes, the hierarchical structure of the nodes is optimized. All the optimized nodes and their relationships are represented in a tree structure to generate a calciner blockage feature tree. The feature tree intuitively displays the hierarchical structure from the top-level cause to the specific blockage problem.

[0073] Furthermore, the step of obtaining the calcining furnace anti-blocking coordination control parameters includes:

[0074] The anti-blocking strategy mechanism is triggered to call and obtain the anti-blocking strategy library of the phosphogypsum calcining furnace; the anti-blocking strategy scheme of the target calcining furnace is determined based on the characteristic parameters of the cause of the calcining furnace blockage and the anti-blocking strategy library of the phosphogypsum calcining furnace; the control parameters of the anti-blocking strategy scheme of the target calcining furnace are analyzed based on the set of key node blockage characteristic parameters to obtain the optimization threshold of the anti-blocking control parameters; the anti-blocking effect fitness function is constructed, and a global optimization is performed within the optimization threshold of the anti-blocking control parameters based on the anti-blocking effect fitness function to obtain the anti-blocking coordination control parameters of the calcining furnace.

[0075] When a blockage problem is detected in the calciner, the anti-blocking strategy mechanism is automatically triggered to call the predefined anti-blocking strategy library for the phosphogypsum calciner. The anti-blocking strategy library includes a series of anti-blocking strategies for different types of blockage problems. The anti-blocking strategy contains detailed anti-blocking measures, such as adjusting operating parameters, optimizing process flow or increasing equipment maintenance.

[0076] The characteristic parameters of the blocking cause are matched with the blocking problems in the anti-blocking strategy library to find the solution closest to the current blocking cause. Based on the matching results, the anti-blocking strategy solution that best suits the target calciner is determined as the anti-blocking strategy solution for the target calciner.

[0077] The parameters in the target calciner anti-blocking strategy are optimized according to the key node blockage characteristic parameter set to achieve the goal of effective anti-blocking. Specifically, the optimization threshold of the control parameters is determined by analyzing the key node blockage characteristics. For example, if the temperature of the key node is too high, an upper limit threshold of the temperature control needs to be set. Based on the characteristic parameters and the strategy, the anti-blocking control parameter optimization threshold of each control parameter is determined to ensure the effectiveness of the anti-blocking strategy.

[0078] Construct an anti-blocking effect fitness function, which includes various indicators for evaluating the effectiveness of anti-blocking strategies, such as anti-blocking efficiency, implementation cost, anti-blocking quality, etc. The fitness function can comprehensively consider these factors to evaluate the effects of different control parameter settings.

[0079] Within the optimization threshold range of the anti-blocking control parameters, global optimization algorithms, such as genetic algorithms, particle swarm optimization, etc., are used to optimize the parameters, and the anti-blocking effect fitness function is used to evaluate the optimization parameters. Specifically, an initial control parameter combination is randomly generated within the optimization threshold range of the anti-blocking control parameters, the parameter combination is continuously adjusted, and the fitness function value is calculated to gradually approach the optimal solution. When the preset convergence standard or iteration number limit is reached, the control parameter combination that maximizes the fitness function value is output as the calcining furnace anti-blocking coordination control parameter to achieve the best anti-blocking effect.

[0080] Furthermore, the obtaining of the calcining furnace anti-blocking coordination control parameters includes:

[0081] A plurality of anti-blocking control parameters are randomly selected within the anti-blocking control parameter optimization threshold, and the plurality of anti-blocking control parameters are calculated and evaluated using the anti-blocking effect fitness function to obtain a plurality of anti-blocking control effect fitnesses; the plurality of anti-blocking control parameters are clustered and optimized according to the plurality of anti-blocking control effect fitnesses to obtain an anti-blocking control parameter cluster, the cluster center of the anti-blocking control parameter cluster being the anti-blocking control parameter with the largest fitness; cross-mutation update and iterative optimization evaluation are performed within the anti-blocking control parameter cluster until a preset convergence requirement is reached to obtain the anti-blocking coordination control parameters of the calcining furnace.

[0082] Multiple anti-blocking control parameters are randomly generated within the optimization threshold. Uniform distribution, normal distribution or other random generation methods can be used to ensure that the parameter combination is representative. A pre-built anti-blocking effect fitness function is used to evaluate each randomly generated anti-blocking control parameter to obtain the corresponding fitness value, which indicates the anti-blocking effect of the parameter combination.

[0083] According to the fitness of multiple anti-blocking control effects, multiple anti-blocking control parameters are clustered and optimized. A preset proportion of control parameters with the highest fitness are clustered and optimized, for example, the first 20% of the control parameters. After clustering, an anti-blocking control parameter cluster is obtained, and the cluster center of the anti-blocking control parameter cluster is the anti-blocking control parameter with the largest fitness.

[0084] Determine the cluster center within the anti-blocking control parameter cluster as the initial individual, set the parameters for crossover mutation update and iterative optimization, including crossover rate, mutation rate and number of iterations, and randomly select multiple individuals from the parameter cluster for crossover operation. The crossover operation is to generate new individuals. New individuals are created by combining the characteristics of two or more individuals. The individuals generated by the crossover are mutated to increase the diversity of the population. Mutation can change part of the characteristics of the individual, thereby exploring new solution spaces. The individuals after crossover mutation are used to replace the original individuals to form a new generation of individuals. The fitness value of the newly generated anti-blocking control parameters is calculated through the fitness function to characterize the performance of each individual in the actual anti-blocking control. According to the fitness value, high-quality individuals are selected to enter the next round of iteration, and the process of crossover mutation update and fitness evaluation is repeated to gradually optimize the anti-blocking control parameters and set convergence criteria, such as the change in fitness value is less than the set threshold, or the number of iterations reaches the upper limit. When the convergence criteria are reached, the iterative process is stopped to obtain the final calciner anti-blocking coordination control parameters, which show the best anti-blocking effect during the optimization process.

[0085] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent anti-blocking device for a phosphogypsum calcining furnace, characterized in that: The device comprises: A monitoring demand analysis module, the monitoring demand analysis module is used to obtain the structural design information of the target phosphogypsum calcining furnace, identify key positions of the structural design information, obtain a set of key positions of the calcining furnace, perform monitoring demand analysis based on the set of key positions of the calcining furnace, and determine a set of sensor demand parameters for key positions; A working data stream acquisition module, the working data stream acquisition module is used to respectively perform monitoring network deployment on the key position set of the calciner according to the key position sensor requirement parameter set, obtain the calciner monitoring node network, and acquire the working data stream set of the key position of the calciner based on the calciner monitoring node network collection; A distributed congestion assessment module, the distributed congestion assessment module is used to construct a node data processing module set according to the calciner monitoring node network, and to perform a distributed congestion assessment on the key position working data flow set of the calciner through the node data processing module set to obtain a key node congestion characteristic parameter set; An association matching module, wherein the association matching module is used to construct a calciner blockage feature tree, and to determine a calciner blockage cause feature parameter based on an association matching between the key node blockage feature parameter set and the calciner blockage feature tree; An anti-blocking early warning control module, the anti-blocking early warning control module is used to trigger an anti-blocking strategy mechanism to perform anti-blocking control analysis on the calcining furnace blockage cause characteristic parameters and the key node blockage characteristic parameter set, obtain the calcining furnace anti-blocking coordination control parameters, and perform anti-blocking early warning control on the target phosphogypsum calcining furnace based on the calcining furnace anti-blocking coordination control parameters; The distributed congestion assessment module comprises the following steps: Extracting data attributes from each monitoring node in the calciner monitoring node network to obtain a node collection data attribute information set; Performing data preprocessing step analysis based on the node collection data attribute information set to determine a node data preprocessing multi-channel set; Performing data mining on each monitoring node to obtain a node working data set, and constructing a node congestion assessment model set based on the node working data set; Based on the node data preprocessing multi-channel set and the node congestion assessment model set, loading the data processing program on each monitoring node to construct the node data processing module set; The distributed congestion assessment module comprises the following steps: Determine a node data attribute threshold set according to the node collection data attribute information set, perform outlier identification on the node working data set according to the node data attribute threshold set, and obtain a node abnormal working attribute data set; Performing a congestion feature association assessment based on the node abnormal working attribute data set to determine an abnormal data congestion risk assessment parameter set; The node working data set is identified according to the abnormal data congestion risk assessment parameter set to obtain a node working congestion assessment data set; Congestion assessment training is performed on the node working congestion assessment data sets respectively to construct the node congestion assessment model set.

2. The intelligent anti-blocking device for phosphogypsum calcining furnace according to claim 1, characterized in that: The association matching module includes the following steps: Obtaining calciner blockage characteristic factor information, wherein the calciner blockage characteristic factor information includes a blockage node location, a node blockage risk parameter, and a corresponding blockage generation cause; Based on the calcining furnace blockage characteristic factor information, an associated topological identification is performed on the phosphogypsum calcining furnace blockage fault library to obtain a calcining furnace blockage characteristic fault library; A calcining furnace blocking fault top node is defined, and based on the calcining furnace blocking fault top node, blocking fault causes of the calcining furnace blocking characteristic fault library are classified to obtain a calcining furnace first-level blocking cause node set; Based on the first-level blockage cause node set of the calcining furnace, a blockage feature cascade analysis is performed to construct the calcining furnace blockage feature tree.

3. The intelligent anti-blocking device for phosphogypsum calcining furnace according to claim 2, characterized in that: The association matching module includes the following steps: Performing a blockage feature correlation analysis on the first-level blockage cause node set of the calcining furnace to obtain a first-level node blockage feature parameter set; Based on the first-level node blockage characteristic parameter set, the fault cause is subdivided and associated to construct a second-level blockage cause node set of the calciner; Performing blockage association analysis on the secondary blockage cause node set of the calcining furnace in sequence until a K-level blockage cause node set of the calcining furnace is obtained, wherein K≥2; Based on the calciner blockage fault top node, the calciner first-level blockage cause node set and the calciner second-level blockage cause node set, cascade deep clustering is performed until the calciner K-level blockage cause node set to obtain the calciner blockage feature tree.

4. The intelligent anti-blocking device for a phosphogypsum calcining furnace according to claim 1, characterized in that: The anti-blocking early warning control module includes the following steps: Trigger the anti-blocking strategy mechanism to call and obtain the anti-blocking strategy library of the phosphogypsum calcining furnace; Based on the matching of the characteristic parameters of the calcining furnace blockage cause with the anti-blocking strategy library of the phosphogypsum calcining furnace, the anti-blocking strategy scheme of the target calcining furnace is determined; Based on the key node blocking characteristic parameter set, the control parameter of the target calcining furnace blocking prevention strategy solution is analyzed to obtain the blocking prevention control parameter optimization threshold; An anti-blocking effect fitness function is constructed, and based on the anti-blocking effect fitness function, a global optimization is performed within the anti-blocking control parameter optimization threshold to obtain the anti-blocking coordination control parameters of the calcining furnace.

5. The intelligent anti-blocking device for phosphogypsum calcining furnace according to claim 4, characterized in that: The anti-blocking early warning control module includes the following steps: Randomly select and obtain multiple anti-blocking control parameters within the anti-blocking control parameter optimization threshold, and use the anti-blocking effect fitness function to calculate and evaluate the multiple anti-blocking control parameters to obtain multiple anti-blocking control effect fitnesses; Clustering and optimizing the multiple anti-blocking control parameters according to the multiple anti-blocking control effect fitnesses to obtain an anti-blocking control parameter cluster, wherein the cluster center of the anti-blocking control parameter cluster is the anti-blocking control parameter with the greatest fitness; Cross-mutation updating and iterative optimization evaluation are performed within the anti-blocking control parameter cluster until a preset convergence requirement is reached to obtain the anti-blocking coordination control parameters of the calcining furnace.

6. An intelligent anti-blocking method for a phosphogypsum calcining furnace, characterized in that: The method is implemented based on the intelligent anti-blocking device for phosphogypsum calcining furnace according to any one of claims 1 to 5, and the method comprises: Acquire the structural design information of the target phosphogypsum calcining furnace, identify key positions of the structural design information, obtain a set of key positions of the calcining furnace, perform monitoring demand analysis based on the set of key positions of the calcining furnace, and determine a set of sensor demand parameters for key positions; According to the key position sensor requirement parameter set, monitoring network deployment is performed on the key position set of the calciner respectively to obtain a calciner monitoring node network, and based on the calciner monitoring node network, a working data stream set of the key position of the calciner is acquired; A node data processing module set is constructed according to the calciner monitoring node network, and a distributed congestion assessment is performed on a set of working data streams at key positions of the calciner through the node data processing module set to obtain a set of key node congestion characteristic parameters; Constructing a calciner blockage feature tree, and performing correlation matching based on the key node blockage feature parameter set and the calciner blockage feature tree to determine the calciner blockage cause feature parameters; The anti-blocking strategy mechanism is triggered to perform anti-blocking control analysis on the characteristic parameters of the calcining furnace blockage cause and the set of key node blockage characteristic parameters, obtain the calcining furnace anti-blocking coordination control parameters, and perform anti-blocking early warning control on the target phosphogypsum calcining furnace based on the calcining furnace anti-blocking coordination control parameters.

Images