Digital intelligent detection method and system based on pneumatic powder conveying

Abstract

The invention relates to the technical field of intelligent detection, and particularly discloses a digital intelligent detection method and system based on powder pneumatic conveying, and the method comprises the steps: monitoring various state parameters in a powder pneumatic conveying process through deploying sensor assemblies on a plurality of key nodes of a powder pneumatic conveying pipeline; a data analysis algorithm based on deep learning is introduced to carry out association interaction analysis on the multi-source state parameters of all the position nodes so as to capture the pneumatic conveying state characteristics of all the position nodes, and then global context sensing of the pneumatic conveying state characteristics is further carried out on the fluctuation condition of the pneumatic conveying state characteristics; the global feature description of the whole pneumatic conveying state of the pipeline is excavated, so that whether the pipeline is blocked or not is intelligently recognized, and corresponding early warning prompt is carried out. According to the invention, real-time monitoring and early warning of the blockage condition of the powder pneumatic conveying pipeline can be realized, so that corresponding maintenance measures can be taken in time, and production interruption and economic loss are avoided.

Description

Technical Field

[0001] The present application relates to the field of intelligent detection technology, and more specifically, to a digital intelligent detection method and system based on powder pneumatic conveying. Background Art

[0002] Powder pneumatic conveying is an efficient way of transporting powdered and granular materials from one place to another through pipelines using gas (usually compressed air) as a medium. This system is widely used in industries such as chemical, food processing, pharmaceutical, and cement. Compared with traditional mechanical conveying, pneumatic conveying has the advantages of good sealing, reduced environmental pollution, high degree of automation, and the ability to achieve long-distance and vertical conveying.

[0003] In pneumatic conveying systems, pipe blockage is a common and serious problem. Due to factors such as powder characteristics (such as particle size distribution, humidity), operating conditions (such as wind speed, pressure) or equipment failure, materials may be deposited on the inner wall of the pipe, gradually forming a blockage, which in turn affects production efficiency and even causes production line shutdown, resulting in economic losses.

[0004] At present, the monitoring of pneumatic conveying systems mainly relies on manual inspections and empirical judgments, or the use of sensors to monitor certain specific parameters (such as pressure difference, temperature). However, manual inspection methods are not only time-consuming and labor-intensive, but also difficult to accurately detect potential blockage risks in real time. Although traditional sensor monitoring methods can provide certain data support, they can usually only detect whether blockage has occurred through simple threshold judgments, and it is difficult to fully reflect the complex powder conveying status inside the pipeline. Especially when material properties change or operating conditions fluctuate, it is easy to misreport or miss blockages.

[0005] Therefore, an optimized digital intelligent detection method and system based on powder pneumatic conveying is expected. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a digital intelligent detection method and system based on powder pneumatic conveying, which monitors various state parameters in the powder pneumatic conveying process by deploying sensor components at multiple key nodes of the powder pneumatic conveying pipeline, and introduces a data analysis algorithm based on deep learning to perform correlation and interactive analysis on the multi-source state parameters of each position node to capture the pneumatic conveying state characteristics of each position node, and then further perform global context perception of the pneumatic conveying state characteristics based on the fluctuation of the pneumatic conveying state characteristics between adjacent position nodes of the pipeline, so as to dig out the global feature description of the overall pneumatic conveying state of the pipeline, so as to intelligently identify whether the pipeline is blocked and provide corresponding early warning prompts. In this way, real-time monitoring and early warning of the blockage of the powder pneumatic conveying pipeline can be achieved, so that corresponding maintenance measures can be taken in time to avoid production interruptions and economic losses.

[0007] According to one aspect of the present application, a digital intelligent detection method based on powder pneumatic conveying is provided, which includes: Acquire pneumatic conveying state data collected by sensor components deployed at multiple key node positions of the pipeline to obtain a node set of pneumatic conveying state data; Extracting pneumatic conveying state features from each pneumatic conveying state data in the node set of pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors; Performing context association coding based on neighborhood location node transportation state perception on the node set of the single-point pneumatic transportation state implicit coding vector to obtain a pipeline pneumatic transportation state node context coding vector; Based on the pipeline pneumatic conveying state node context coding vector, it is determined whether to generate a pipeline blockage warning prompt.

[0008] According to another aspect of the present application, a digital intelligent detection system based on powder pneumatic conveying is provided, which includes: A state data acquisition module, used to acquire pneumatic conveying state data collected by sensor components deployed at multiple key node positions of the pipeline to obtain a node set of pneumatic conveying state data; A state feature extraction module, used to extract pneumatic conveying state features from each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors; A context-associated coding module, used for performing context-associated coding on the node set of the single-point pneumatic conveying state implicit coding vector based on the conveying state perception of the neighboring position node to obtain a pipeline pneumatic conveying state node context coding vector; The early warning prompt generation module is used to determine whether to generate a pipeline blockage early warning prompt based on the pipeline pneumatic conveying state node context coding vector.

[0009] Technical effects: Compared with the prior art, the digital intelligent detection method and system based on pneumatic conveying of powders provided in the present application monitor various state parameters in the process of pneumatic conveying of powders by deploying sensor components at multiple key nodes of the pneumatic conveying pipeline of powders, and introduce a data analysis algorithm based on deep learning to perform correlation and interactive analysis on the multi-source state parameters of each position node to capture the pneumatic conveying state characteristics of each position node. Then, the global context perception of the pneumatic conveying state characteristics is further performed based on the fluctuation of the pneumatic conveying state characteristics between adjacent position nodes of the pipeline, so as to mine the global characteristic description of the overall pneumatic conveying state of the pipeline, thereby intelligently identifying whether the pipeline is blocked and issuing corresponding early warning prompts, which can realize real-time monitoring and early warning of blockage of the pneumatic conveying pipeline of powders, so as to take corresponding maintenance measures in time to avoid production interruption and economic losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0011] Figure 1 It is a flow chart of a digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application.

[0012] Figure 2 Schematic diagram of data flow of a digital intelligent detection method based on pneumatic conveying of powders according to an embodiment of the present application.

[0013] Figure 3 This is a flowchart of sub-step S2 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application.

[0014] Figure 4 This is a flowchart of sub-step S3 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application.

[0015] Figure 5 This is a flowchart of sub-step S31 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application.

[0016] Figure 6This is a flowchart of sub-step S32 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application.

[0017] Figure 7 It is a block diagram of a digital intelligent detection system based on powder pneumatic conveying according to an embodiment of the present application. DETAILED DESCRIPTION

[0018] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0019] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0020] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0021] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0022] It should be noted that in this application, all actions to obtain data are carried out in compliance with the relevant data protection laws and policies of the country where the data is located, and with the authorization given by the owner of the corresponding device.

[0023] In response to the technical problems described in the above background technology, this application proposes a digital intelligent detection method based on pneumatic conveying of powders, which monitors various state parameters in the process of pneumatic conveying of powders by deploying sensor components at multiple key nodes of the powder pneumatic conveying pipeline, and introduces a data analysis algorithm based on deep learning to perform correlation and interactive analysis on the multi-source state parameters of each position node to capture the pneumatic conveying state characteristics of each position node. Then, the global context perception of the pneumatic conveying state characteristics is further performed based on the fluctuation of the pneumatic conveying state characteristics between adjacent position nodes of the pipeline, so as to mine the global feature description of the overall pneumatic conveying state of the pipeline, thereby intelligently identifying whether the pipeline is blocked and providing corresponding early warning prompts. In this way, real-time monitoring and early warning of blockage of the powder pneumatic conveying pipeline can be achieved, so that corresponding maintenance measures can be taken in time to avoid production interruptions and economic losses.

[0024] Figure 1 It is a flow chart of a digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of a digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the digital intelligent detection method based on powder pneumatic conveying includes the following steps: S1, acquiring pneumatic conveying state data collected by sensor components deployed at multiple key node positions of a pipeline to obtain a node set of pneumatic conveying state data; S2, extracting pneumatic conveying state features from each pneumatic conveying state data in the node set of pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors; S3, performing context association coding on the node set of the single-point pneumatic conveying state implicit coding vector based on the perception of the conveying state of the neighboring position nodes to obtain a pipeline pneumatic conveying state node context coding vector; S4, determining whether to generate a pipeline blockage warning prompt based on the pipeline pneumatic conveying state node context coding vector.

[0025] In the above-mentioned digital intelligent detection method based on pneumatic conveying of powders, the step S1 obtains the pneumatic conveying state data collected by the sensor components deployed at multiple key node positions of the pipeline to obtain a node set of pneumatic conveying state data. In a specific example of the present application, the pneumatic conveying state data includes powder flow value, gas flow rate value and pressure value. It should be understood that the powder pneumatic conveying process involves complex gas-solid two-phase flow, so that the conveying state at different positions may be different. Therefore, the present application collects pneumatic conveying state data in the pipeline at multiple key nodes of the pipeline, such as powder flow value, gas flow rate value and pressure value, so as to fully capture the dynamic changes of powder conveying in the entire pipeline system, deeply understand the inherent laws of the pneumatic conveying process, and thus achieve accurate assessment of the pipeline state.

[0026] Specifically, the sensor components deployed at multiple key nodes of the pipeline begin to collect various data related to the pneumatic conveying status in real time, forming a data set containing rich information. Each element in the set represents the conveying status at a certain location at a specific point in time. When selecting the type of sensor, measurement accuracy, response speed, durability, and the ability to adapt to different environmental conditions must be considered. For the measurement of powder flow values, a mass flow meter or volume flow meter may be required; for the determination of gas flow rate values, a thermal anemometer or ultrasonic anemometer can be used; and the monitoring of pressure values ​​depends on equipment such as pressure transmitters. Each type of sensor has its own unique working principle and technical characteristics. Choosing the right tool is crucial to ensuring data quality and reliability.

[0027] The data types collected by the sensor cover the key parameters that affect the flow of powders, including but not limited to powder flow value, gas flow rate value and pressure value. The powder flow value reflects the amount of material passing through a certain section of the pipeline per unit time. This value is directly related to production efficiency and whether there is a possibility of blockage due to abnormal flow. In actual operation, changes in powder flow may be caused by a variety of factors, such as adjustment of the opening of the feed port, changes in downstream process requirements, or changes in the characteristics of the powder itself. Therefore, by continuously monitoring the powder flow value, not only can the current conveying rate be grasped, but also the trend in the future can be predicted, providing a decision-making basis for preventive maintenance.

[0028] The gas flow rate value refers to the speed of the compressed air or other gas medium that propels the powder forward. This speed needs to be maintained within an appropriate range. Too low a speed may cause powder sedimentation, while too high a speed may cause unnecessary energy consumption or equipment wear. In pneumatic conveying systems, the control of gas flow rate is the key to maintaining stable conveying. In order to achieve the best results, engineers usually set the ideal flow rate range based on factors such as the properties of the powder (such as particle size distribution, density) and conveying distance. At the same time, since gas flow rate has a direct impact on energy consumption, reasonable flow rate management can also help reduce operating costs and improve economic benefits.

[0029] Pressure is another critical parameter that indicates the pressure conditions inside the pipeline, including the pressure level during positive pressure delivery or the vacuum level during negative pressure suction. Under normal operating conditions, maintaining a stable pressure is essential to ensure smooth material transfer, and any significant pressure fluctuations may be a sign of impending problems. Especially during long-distance transportation, pressure loss is an issue that must be taken seriously. If the pressure in a certain section of the pipeline suddenly drops, it may mean a leak or a local blockage. Conversely, if the pressure rises abnormally, it indicates that there may be a risk of material accumulation. Therefore, accurate pressure monitoring not only helps to detect potential faults in a timely manner, but also can guide the optimal design of the system to a certain extent.

[0030] When the sensor assembly starts working, each key node position will continuously generate data records about the above parameters. These data accumulate over time, forming a dynamically changing data stream. Each set of data not only carries information about the current moment, but also implies trend changes over a period of time, such as whether the powder flow rate is gradually decreasing, whether the gas flow rate fluctuates irregularly, whether the pressure rises or falls unexpectedly, etc. All these details together constitute a multi-dimensional perspective describing the entire pneumatic conveying process, providing a solid foundation for subsequent analysis.

[0031] In the above-mentioned digital intelligent detection method based on powder pneumatic conveying, the step S2 extracts the pneumatic conveying state features from each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors. Figure 3 FIG. 1 is a flowchart of sub-step S2 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application. Figure 3As shown, the step S2 includes the steps of: S21, performing structured coding on each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector}; S22, performing multi-source data association interactive coding on each {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} in the node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} to obtain a node set of implicit coding vectors of the single-point pneumatic conveying state.

[0032] Specifically, the step S21 performs structured coding on each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector}. It should be understood that since the parameters such as powder flow, gas flow rate and pressure in the pneumatic conveying state data have different physical dimensions and data distributions, in order to eliminate the influence of such dimensional differences on subsequent data processing and analysis, the present application further performs structured coding on each pneumatic conveying state data to map the pneumatic conveying state parameters with different dimensions and distributions to a unified feature space, thereby obtaining a node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector}. In a specific example of the present application, first, a fine-grained numerical interval is established based on the data distribution characteristics of each pneumatic conveying state parameter, and the One-Hot Encoding technology is used to generate a unique binary vector according to the numerical interval to which each pneumatic conveying state parameter belongs, as a structured feature representation of each pneumatic conveying state parameter, thereby eliminating data dimension differences, enabling the model to better distinguish the pneumatic conveying states represented by different values, and improving the accuracy of subsequent analysis.

[0033] Specifically, in a specific example of the present application, the step S22 includes: inputting each {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} in the node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} into a single-point pneumatic conveying state sensor based on a feedforward neural network model to obtain a node set of the single-point pneumatic conveying state implicit coding vector. In particular, since parameters such as powder flow, gas flow rate and pressure are interrelated in the powder pneumatic conveying process, the three together reflect the conveying state of the powder in the pipeline at the current position node. Therefore, in order to comprehensively consider the intrinsic relationship between powder flow, gas flow rate and pressure, so as to achieve comprehensive perception and description of the pneumatic conveying state of each node position, the present application further adopts a feedforward neural network model to interactively fuse the {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} of each node position. Through the nonlinear mapping capability of the feedforward neural network model, the structured characteristics of each state parameter are interacted and fused, thereby generating a single-point pneumatic conveying state implicit coding vector that can comprehensively reflect the pneumatic conveying state of the current node position.

[0034] In the above-mentioned digital intelligent detection method based on powder pneumatic conveying, the step S3 performs context-associated coding on the node set of the implicit coding vector of the single-point pneumatic conveying state based on the perception of the conveying state of the neighboring node position to obtain the context coding vector of the pipeline pneumatic conveying state node. It should be understood that it is often difficult to fully reflect the overall conveying state of the pipeline by focusing only on the pneumatic conveying state characteristics of a single node. There is a mutual correlation between the pneumatic conveying states of adjacent nodes in the pipeline, especially when there is a local blockage or uneven conveying conditions in the pipeline, the pneumatic conveying states between adjacent nodes will fluctuate significantly. Therefore, in order to capture the global characteristics of the overall pneumatic conveying state of the pipeline, the present application further performs global context perception by analyzing the fluctuations in the conveying state characteristics between the neighboring node positions, and performs context-associated coding on the node set of the implicit coding vector of the single-point pneumatic conveying state, thereby mining the global feature description of the overall pneumatic conveying state of the pipeline, so as to more accurately judge whether there are abnormal conditions such as blockage in the pipeline. Among them, Figure 4 FIG. 1 is a flowchart of sub-step S3 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application. Figure 4As shown, the step S3 includes the steps of: S31, performing feature transfer significance measurement based on neighborhood position node transportation state perception on each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector to obtain a node set of single-point pneumatic conveying state feature transfer significance factors; S32, based on the node set of single-point pneumatic conveying state feature transfer significance factors, performing attention-driven feature transfer aggregation coding on the node set of the single-point pneumatic conveying state implicit coding vector to obtain the pipeline pneumatic conveying state node context coding vector.

[0035] Figure 5 FIG. 1 is a flowchart of sub-step S31 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application. Figure 5 As shown, the step S31 includes the steps of: S311, calculating the characteristic jump degree of each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector to obtain the node set of the single-point pneumatic conveying state characteristic jump degree; S312, calculating the characteristic transfer space span of each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector to obtain the node set of the single-point pneumatic conveying state characteristic transfer space span; S313, based on the characteristic jump degree and characteristic transfer space span of each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector, calculating the characteristic transfer significance factor of each single-point pneumatic conveying state implicit coding vector to obtain the node set of the single-point pneumatic conveying state characteristic transfer significance factor.

[0036] More specifically, the step S311 is expressed as follows:

[0037]

[0038]

[0039] in, A node set representing the implicit encoding vector of the single-point pneumatic conveying state, , , and The first, second, and third nodes in the node set of the single-point pneumatic conveying state implicit coding vector are represented respectively. and A single-point pneumatic conveying state implicit encoding vector, is the number of feature vectors in the node set of the single-point pneumatic conveying state implicit coding vector, Indicates the The first implicit encoding vector of the single-point pneumatic conveying state The eigenvalues ​​at the positions, For the The characteristic scale value of the implicit encoding vector of a single-point pneumatic conveying state, For the The characteristic energy intensity factor of the implicit encoding vector of a single-point pneumatic conveying state, Indicates The characteristic energy intensity factor of the implicit encoding vector of a single-point pneumatic conveying state, Indicates the The characteristic jump degree of the implicit encoding vector of a single-point pneumatic conveying state.

[0040] Specifically, in order to accurately capture the subtle changes in the pneumatic conveying state of each node in the pipeline, the present application uses a sliding window technique to traverse the implicit coding vectors of the pneumatic conveying state of all adjacent nodes, and evaluates the drastic degree of characteristic changes in the pneumatic conveying state characteristics of each node position relative to the adjacent node position by calculating the characteristic jump degree between the adjacent single-point pneumatic conveying state implicit coding vectors. It should be understood that when the overall pneumatic conveying state of the pipeline is globally perceived in the context, the implicit coding vectors of the pneumatic conveying state of each single point exchange information in sequence according to the order of their positions in the pipeline. When the characteristic difference between the implicit coding vector of a single-point pneumatic conveying state and its adjacent nodes increases significantly, it indicates that the pneumatic conveying state at the node has changed significantly, which may be a sign of local blockage or inconsistent conveying conditions. At this time, special attention and in-depth analysis of this node are required to confirm whether there is a risk of blockage. This method ensures that the system can identify potential problems in a timely manner and provides a scientific basis for taking preventive measures.

[0041] More specifically, the step S312 is expressed as follows:

[0042] in, Indicates the The characteristic transfer space span of the implicit encoding vector of a single-point pneumatic conveying state, Indicates the and stated The number of eigenvectors between them.

[0043] Specifically, in view of the continuity of the powder flow in the pipeline, there is a certain spatial correlation between the pneumatic conveying states of the nodes at each position. Specifically, the correlation between the pneumatic conveying states of adjacent nodes is closer, while the correlation between nodes that are farther apart is relatively weak. In order to make full use of this spatial correlation, the application takes the last node position at the end of the pipeline as the anchor center, and evaluates the spatial correlation degree of the pneumatic conveying state of each node relative to the state at the end of the pipeline by calculating the feature transfer space span of the implicit coding vector of each single-point pneumatic conveying state relative to the anchor center, revealing the spatial distribution characteristics of the pneumatic conveying state in the pipeline, which helps to further evaluate the importance and influence of each node position in the perception of the overall pneumatic conveying state of the pipeline.

[0044] More specifically, the step S313 is expressed by the formula:

[0045] in, and is a preset weight parameter used to balance the influence of feature jump degree and feature transfer space span. Indicates the The characteristic transfer significance factor of the implicit encoding vector of a single-point pneumatic conveying state.

[0046] Specifically, in order to more comprehensively evaluate the two important factors of feature jump degree and feature transfer space span, so as to more accurately measure the importance of each single-point pneumatic conveying state implicit coding vector in the context information transmission process, the present application further introduces the calculation of feature transfer significance factor, which is determined based on the feature jump degree and feature transfer space span of each single-point pneumatic conveying state implicit coding vector. In this way, in the subsequent pneumatic conveying state feature context transfer coding process, more focus can be placed on those important node positions, effectively reducing the interference of noise information, thereby improving the accuracy and efficiency of the overall pneumatic conveying state perception of the pipeline.

[0047] Figure 6 FIG. 1 is a flow chart of sub-step S32 of the digital intelligent detection method based on powder pneumatic conveying according to an embodiment of the present application. Figure 6 As shown, the step S32 includes the steps of: S321, inputting the node set of the single-point pneumatic conveying state feature transfer significant factors into a gated transfer unit containing a softmax normalization function to obtain a node set of the single-point pneumatic conveying state feature transfer significant weights; S322, based on the node set of the single-point pneumatic conveying state feature transfer significant weights, weighted aggregation is performed on the node set of the single-point pneumatic conveying state implicit coding vector to obtain the pipeline pneumatic conveying state node context coding vector.

[0048] More specifically, the step S321 is expressed by the formula:

[0049] in, is the normalized exponential function, is the gated mask function, is the gating threshold, For the The single-point pneumatic conveying state characteristics transfer significant weight.

[0050] Specifically, this application introduces a gating mechanism to screen the feature transfer significance factors of each single-point pneumatic conveying state implicit coding vector, and generates a node set with significant weights of single-point pneumatic conveying state feature transfer. Specifically, the gating logic is used to evaluate and select nodes with higher importance, ensuring that only those nodes with significant feature transfer are included in the final analysis, thereby optimizing the data processing process and improving the accuracy of the analysis.

[0051] More specifically, the step S322 is expressed by the formula:

[0052] in, A node context encoding vector representing the pipeline pneumatic transportation state.

[0053] Specifically, the generated weights are further used to perform weighted aggregation on the node set of the original single-point pneumatic conveying state implicit coding vector to generate the pipeline pneumatic conveying state node context coding vector. It should be understood that through weighted aggregation, appropriate weights can be given to different nodes, the influence of important nodes can be highlighted, and the interference of noise data can be reduced. It can ensure that the pneumatic conveying state characteristics of each node in the pipeline are reasonably reflected, enhance the ability to capture the dynamic changes of the entire conveying process, and provide a solid foundation for more accurate state monitoring and fault warning.

[0054] In this way, not only the pneumatic conveying status information of each location node is effectively integrated, but also the spatial correlation between nodes and the degree of change of the pneumatic conveying status are fully considered, thereby achieving a more accurate and comprehensive understanding of the overall pneumatic conveying status of the pipeline.

[0055] In the above-mentioned digital intelligent detection method based on powder pneumatic conveying, the step S4 determines whether to generate a pipeline blockage warning prompt based on the pipeline pneumatic conveying state node context coding vector. In a specific example of the present application, the step S4 includes: inputting the pipeline pneumatic conveying state node context coding vector into a classifier-based intelligent detection module to obtain a detection result, and the detection result is used to indicate whether there is a pipeline blockage; in response to the detection result that there is a pipeline blockage, the pipeline blockage warning prompt is generated. Specifically, the classifier is based on a neural network architecture, and analyzes whether there is a blockage in the pipeline by performing feature learning on the pipeline pneumatic conveying state node context coding vector, and outputs the final detection result through the Softmax function of the classification layer. Then, when the pipeline blockage is detected, a warning prompt is automatically generated. In this way, the staff can deal with pipeline abnormalities in a timely manner according to the warning prompt, ensure the normal operation of the production line, and reduce economic losses.

[0056] Here, since each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector respectively represents the implicit associated coding feature representation of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} at the key node position, when performing feature sequence message transmission based on feature jump degree in the full node space, the parameter semantic association feature population attributes of different key nodes will have sequence transmission fairness differences based on the feature jump degree level, thereby affecting the expression certainty of the pipeline pneumatic conveying state node context coding vector relative to the classifier-based intelligent detection module, and affecting the accuracy of the detection result obtained by its input to the classifier-based intelligent detection module.

[0057] In a preferred example, inputting the pipeline pneumatic conveying state node context encoding vector into a classifier-based intelligent detection module to obtain a detection result includes: decoding a pneumatic conveying parameter spectrum tensor from the classifier-based intelligent detection module; Embedding the pipeline pneumatic transport state node context encoding vector into the tangent bundle space of the pneumatic transport parameter spectrum tensor to generate a pneumatic transport mimic criterion topological encoding;

[0058] in, represents the pipeline pneumatic transport state node context encoding vector, represents the pneumatic conveying parameter spectrum tensor, Represents the topological coding of pneumatic conveying mimetic criteria; Applying a polynomial gain operation to the pneumatic conveying mimic criterion topological code to output a pneumatic conveying dynamic scaling mapping code;

[0059] where represents the natural exponential function, Represents the pneumatic conveying dynamic scaling mapping code; Generate a pneumatic transport hyperdimensional clustering coding manifold based on the pneumatic transport dynamic scaling mapping coding and the pneumatic transport mimicry criterion topological coding;

[0060] in, represents the transpose operation, represents matrix multiplication, represents the length of the pneumatic conveying dynamic scaling map encoding, Represents the pneumatic conveying hyperdimensional clustering encoding manifold; performing curvature regularization on the pneumatic transport hyperdimensional clustering encoding manifold to generate a pneumatic transport steady-state convergence architecture;

[0061] in, It represents the steady-state convergence architecture of pneumatic conveying; Reparameterize the pipeline pneumatic transport state node context encoding vector to the fiber bundle space of the pneumatic transport steady-state convergence architecture, and extract the pneumatic transport discrimination compensation tensor;

[0062] in, represents the pneumatic conveying discrimination compensation tensor; The pipeline pneumatic transportation state node context encoding vector and the pneumatic transportation discrimination compensation tensor are fused by tensor bundle coupling to output an optimized pipeline pneumatic transportation state node context encoding vector;

[0063] in, and represents the weighted hyperparameter, A node context encoding vector representing an optimized pipeline pneumatic conveying state; Finally, the optimized pipeline pneumatic conveying state node context encoding vector is input into a classifier-based intelligent detection module to obtain a detection result.

[0064] Here, based on the segmentation hypercurvature gradient set determined by the implicit rule topology of the model, the context encoding vector of the pipeline pneumatic conveying state node is guided to evolve towards the topological separability space, and the correlation dimension with the category criterion is enhanced through the polynomial gain converter. The hyperdimensional clustering convergence field of the classification feature is constructed as the information-intensive topological capture discriminant compensation tensor, and the tensor is fused using the spectrum adaptive convergence mechanism. The optimized context encoding vector of the pipeline pneumatic conveying state node is adapted to the optimization of the parameter spectrum of the classifier-based intelligent detection module on the dynamic scaling manifold to improve the accuracy of the detection results obtained by its input to the classifier-based intelligent detection module.

[0065] After confirming that the test results show that there is a pipeline blockage, the system will start to build specific warning prompts according to the predefined set of rules. The warning prompt is not just a simple warning message, but a notification with rich details, which is designed to convey a comprehensive and easy-to-understand description of the situation to the operator. To achieve this goal, the content design of the warning prompt must take into account both professionalism and readability. On the one hand, the notification should include key technical parameters, such as the specific location of the blockage, the main sensor data involved (powder flow value, gas flow rate value, pressure value, etc.), the possible scope of impact and the estimated duration, etc.; on the other hand, the expression should be simplified as much as possible, using intuitive language and graphical elements, so that even operators without a deep technical background can quickly grasp the core information. In addition, you can also consider adding some guiding suggestions, such as recommended initial measures or which experts to contact for help, so that relevant personnel can respond quickly.

[0066] In addition to the content itself, the presentation form of the warning is equally important. In order to ensure that the information can be conveyed to the target audience in a timely and effective manner, the system supports multiple output channels, including but not limited to screen display, sound alarm, SMS notification, email sending, etc. Each channel has its own characteristics and applicable scenarios, and can be flexibly selected according to specific needs during actual deployment. For example, inside the control room, large display screens and high-volume alarms can be used to attract the attention of field staff; while in remote monitoring centers, it is more suitable to deliver detailed reports through electronic communication means. At the same time, in order to improve the efficiency of information transmission, attention should also be paid to consistency and collaborative work between different channels to avoid duplication or conflict of information. For example, links can be attached in text messages and emails to directly point to the complete warning page, so that the recipient can get more information.

[0067] In order to ensure the efficiency and stability of the early warning prompt generation process, a strict quality control system is set up inside the system. From the initial input of the test results to the final output of the prompt, each link is strictly monitored to ensure that each link is smoothly connected and error-free. Especially in the case of multi-threaded concurrent processing, in order to avoid data loss or delay caused by resource competition, advanced queue management and locking mechanisms are adopted. In addition, the system also has a self-diagnosis function, which can monitor its own health status in real time. Once a potential problem is found, it will automatically trigger the repair process or issue a maintenance reminder, thereby maximizing service continuity.

[0068] It is worth mentioning that in order to adapt to the needs of different application scenarios, the system provides highly configurable early warning strategies. This means that parameters such as early warning thresholds, priority settings, and response modes can be flexibly adjusted according to the actual working conditions and management requirements of a specific plant. For example, in some cases, you may want to receive a more conservative early warning in advance so that there is enough time for preventive maintenance; at other times, you can choose more stringent conditions and only issue warnings for situations where there are indeed major risks.

[0069] In summary, a digital intelligent detection method based on pneumatic conveying of powders according to the embodiment of the present application is explained, which monitors various state parameters in the pneumatic conveying process of powders by deploying sensor components at multiple key nodes of the pneumatic conveying pipeline of powders, and introduces a data analysis algorithm based on deep learning to perform correlation and interactive analysis on the multi-source state parameters of each position node to capture the pneumatic conveying state characteristics of each position node, and then further performs global context perception of the pneumatic conveying state characteristics based on the fluctuation of the pneumatic conveying state characteristics between adjacent position nodes of the pipeline, so as to dig out the global characteristic description of the overall pneumatic conveying state of the pipeline, thereby intelligently identifying whether the pipeline is blocked and providing corresponding early warning prompts. In this way, real-time monitoring and early warning of blockage of the powder pneumatic conveying pipeline can be achieved, so that corresponding maintenance measures can be taken in time to avoid production interruptions and economic losses.

[0070] Furthermore, a digital intelligent detection system based on powder pneumatic conveying is also provided.

[0071] Figure 7 FIG. 1 is a block diagram of a digital intelligent detection system based on powder pneumatic conveying according to an embodiment of the present application. Figure 7As shown, according to the embodiment of the present application, the digital intelligent detection system 100 based on pneumatic conveying of powders includes: a state data acquisition module 110, which is used to obtain the pneumatic conveying state data collected by the sensor components deployed at multiple key node positions of the pipeline to obtain a node set of pneumatic conveying state data; a state feature extraction module 120, which is used to extract the pneumatic conveying state features from each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors; a context association coding module 130, which is used to perform context association coding on the node set of the single-point pneumatic conveying state implicit coding vector based on the perception of the conveying state of the neighborhood position node to obtain a pipeline pneumatic conveying state node context coding vector; an early warning prompt generation module 140, which is used to determine whether to generate a pipeline blockage early warning prompt based on the pipeline pneumatic conveying state node context coding vector.

[0072] The specific operation of each module in the above-mentioned digital intelligent detection system based on powder pneumatic conveying has been referred to above. Figures 1 to 6 The digital intelligent detection method based on powder pneumatic conveying has been introduced in detail, and therefore, its repeated description will be omitted.

[0073] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0074] In the above embodiments, the description of each embodiment has its own emphasis. For the parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0075] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0076] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0077] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A digital intelligent detection method based on powder pneumatic conveying, characterized in that: include: Acquire pneumatic conveying state data collected by sensor components deployed at multiple key node positions of the pipeline to obtain a node set of pneumatic conveying state data; Extracting pneumatic conveying state features from each pneumatic conveying state data in the node set of pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors; Performing context association coding based on neighborhood location node transportation state perception on the node set of the single-point pneumatic transportation state implicit coding vector to obtain a pipeline pneumatic transportation state node context coding vector; Based on the pipeline pneumatic conveying state node context coding vector, it is determined whether to generate a pipeline blockage warning prompt.

2. The digital intelligent detection method based on powder pneumatic conveying according to claim 1 is characterized in that: The pneumatic conveying state data includes a powder flow value, a gas flow rate value and a pressure value.

3. The digital intelligent detection method based on powder pneumatic conveying according to claim 2 is characterized in that: Extracting pneumatic conveying state features from each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors, including: Performing structured coding on each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector}; Multi-source data association interactive coding is performed on each {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} in the node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} to obtain the node set of the single-point pneumatic conveying state implicit coding vector.

4. The digital intelligent detection method based on powder pneumatic conveying according to claim 3 is characterized in that: Each {powder flow rate embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} in the node set of {powder flow rate embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} is respectively subjected to multi-source data association interactive coding to obtain a node set of the single-point pneumatic conveying state implicit coding vector, including: Each {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} in the node set of {powder flow embedded coding vector, gas flow rate embedded coding vector, pressure embedded coding vector} is respectively input into the single-point pneumatic conveying state sensor based on the feedforward neural network model to obtain the node set of the single-point pneumatic conveying state implicit coding vector.

5. The digital intelligent detection method based on powder pneumatic conveying according to claim 4 is characterized in that: The node set of the single-point pneumatic conveying state implicit coding vector is subjected to context association coding based on the conveying state perception of the neighboring position node to obtain the pipeline pneumatic conveying state node context coding vector, including: Performing feature transfer significance measurement based on neighborhood position node transport state perception on each single-point pneumatic transport state implicit coding vector in the node set of the single-point pneumatic transport state implicit coding vector to obtain a node set of single-point pneumatic transport state feature transfer significance factors; Based on the node set of the single-point pneumatic conveying state feature transfer significant factors, the node set of the single-point pneumatic conveying state implicit encoding vector is subjected to attention-driven feature transfer aggregation encoding to obtain the pipeline pneumatic conveying state node context encoding vector.

6. The digital intelligent detection method based on powder pneumatic conveying according to claim 5 is characterized in that: The node set of the single-point pneumatic conveying state implicit coding vectors in the node set of the single-point pneumatic conveying state implicit coding vectors is subjected to feature transfer significance measurement based on the conveying state perception of the neighboring position node to obtain the node set of the single-point pneumatic conveying state feature transfer significance factor, including: Calculating the characteristic jump degree of each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector to obtain a node set of the single-point pneumatic conveying state characteristic jump degree; Calculating the characteristic transfer space span of each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector to obtain the node set of the single-point pneumatic conveying state characteristic transfer space span; Based on the feature jump degree and feature transfer space span of each single-point pneumatic conveying state implicit coding vector in the node set of the single-point pneumatic conveying state implicit coding vector, the feature transfer significance factor of each single-point pneumatic conveying state implicit coding vector is calculated to obtain the node set of the single-point pneumatic conveying state feature transfer significance factor.

7. The digital intelligent detection method based on powder pneumatic conveying according to claim 6 is characterized in that: Based on the node set of the single-point pneumatic conveying state feature transfer significant factor, the node set of the single-point pneumatic conveying state implicit encoding vector is subjected to attention-driven feature transfer aggregation encoding to obtain the pipeline pneumatic conveying state node context encoding vector, including: Inputting the node set of the single-point pneumatic conveying state characteristic transfer significant factors into a gated transfer unit including a softmax normalization function to obtain a node set of the single-point pneumatic conveying state characteristic transfer significant weights; Based on the node set whose single-point pneumatic transportation state feature transfers significant weights, the node set of the single-point pneumatic transportation state implicit coding vector is weighted aggregated to obtain the pipeline pneumatic transportation state node context coding vector.

8. The digital intelligent detection method based on powder pneumatic conveying according to claim 7 is characterized in that: Determining whether to generate a pipeline blockage warning prompt based on the pipeline pneumatic conveying state node context encoding vector includes: Inputting the pipeline pneumatic conveying state node context encoding vector into a classifier-based intelligent detection module to obtain a detection result, wherein the detection result is used to indicate whether there is a pipeline blockage; In response to the detection result indicating that the pipeline is clogged, an early warning prompt of the pipeline clog is generated.

9. A digital intelligent detection system based on powder pneumatic conveying, characterized in that: include: A state data acquisition module, used to acquire pneumatic conveying state data collected by sensor components deployed at multiple key node positions of the pipeline to obtain a node set of pneumatic conveying state data; A state feature extraction module, used to extract pneumatic conveying state features from each pneumatic conveying state data in the node set of the pneumatic conveying state data to obtain a node set of single-point pneumatic conveying state implicit coding vectors; A context-associated coding module, used for performing context-associated coding on the node set of the single-point pneumatic conveying state implicit coding vector based on the conveying state perception of the neighboring position node to obtain a pipeline pneumatic conveying state node context coding vector; The early warning prompt generation module is used to determine whether to generate a pipeline blockage early warning prompt based on the pipeline pneumatic conveying state node context coding vector.

Images