Online monitoring system and method for fiber cement board production line

Abstract

The present application discloses an online monitoring system and method for a fiber cement board production line, which relates to the field of intelligent monitoring. It uses a time series of multiple real-time steam temperatures collected by multiple temperature sensors deployed inside the cylinder, and uses data processing and analysis technology based on deep learning to perform time series encoding and spatial aggregation of the real-time temperatures of each steam, so as to automatically determine whether steam needs to be introduced according to the temperature distribution representation characteristics calculated between the time series characteristics of the real-time steam temperatures and the temperature spatial aggregation representation characteristics. In this way, through the automated temperature monitoring system, the dependence on manual experience is reduced, and the accuracy and response speed of temperature control are improved. In addition, the system can automatically determine the need for steam introduction, thereby optimizing the temperature distribution and ensuring the temperature uniformity of each area in the curing room, thereby improving the level of intelligence in cement board curing production control.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring, and more specifically, to an online monitoring system and method for a fiber cement board production line. Background Art

[0002] Fiber cement board is widely used in building decoration, exterior wall hangings, ceilings and partitions due to its high strength, light weight, fire resistance and corrosion resistance.

[0003] Chinese patent CN110357661B discloses a fiber cement board production method, which firstly involves loading the boards by pushing a trailer loaded with fiber cement boards into a steaming cylinder; then, the board is sealed and heated to ensure that the cylinder is sealed and then steam is introduced for preliminary curing; then, temperature control is performed by circulating steam in the cylinder through a top compression mechanism to achieve balanced heat distribution; finally, quality inspection is performed by placing the steamed boards horizontally and conducting an appearance inspection to timely screen and process defective products.

[0004] Steam curing during the manufacturing process is a key step to ensure product strength and durability. In order to achieve high-quality steam curing, the temperature distribution inside the cylinder needs to be precisely controlled to ensure that all parts of the fiber cement board are heated evenly. However, in the traditional steam curing process, temperature control mainly relies on the experience and technical level of the operator. This method is not only inefficient, but also difficult to ensure the uniformity of temperature distribution. Specifically, operators must frequently check the temperature and manually adjust the steam supply, which not only consumes a lot of manpower, but is also prone to human errors, resulting in unstable temperature control. In addition, due to the lack of accurate temperature monitoring methods, operators often find it difficult to accurately grasp the true situation of the temperature distribution inside the cylinder, which may cause the temperature in some areas to be too high or too low, thereby affecting the strength and uniformity of the fiber cement board.

[0005] Therefore, an online monitoring system and method for a fiber cement board production line is desired. Summary of the invention

[0006] In order to solve the above technical problems, this application is proposed.

[0007] According to one aspect of the present application, an online monitoring system for a fiber cement board production line is provided, which includes: a fiber cement board transportation and loading module, used to place the fiber cement board into a cylinder of a production device; a steam curing module, used to seal the cylinder and pass steam to steam-cure the fiber cement board; a steam control module, used to control the steam through a compression device during the steam curing process to ensure that the heat inside the cylinder is evenly transferred to obtain a fiber cement board after steam curing; an appearance inspection module, used to perform an appearance inspection on the steam-cured fiber cement board to obtain a test result;

[0008] Wherein, the steam control module comprises:

[0009] A steam real-time temperature acquisition unit, used to acquire a time series of multiple steam real-time temperatures acquired by multiple temperature sensors deployed inside the cylinder;

[0010] A real-time temperature time series encoding unit, used for inputting each steam real-time temperature time series of the multiple steam real-time temperature time series into a real-time temperature sequence encoder to obtain multiple steam real-time temperature time series feature vectors;

[0011] A real-time temperature time series feature aggregation unit, used for inputting the plurality of steam real-time temperature time series feature vectors into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector;

[0012] A temperature distribution offset coefficient calculation unit is used to calculate the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients as a temperature distribution topological representation feature;

[0013] The steam control unit is used to obtain a control instruction based on the topological representation characteristics of the temperature distribution, and the control instruction is used to indicate whether steam needs to be introduced.

[0014] In the online monitoring system of the above-mentioned fiber cement board production line, the real-time temperature time series encoding unit is used to: input each steam real-time temperature time series in the multiple steam real-time temperature time series into the real-time temperature sequence encoder based on the Bi-LSTM model to obtain multiple steam real-time temperature time series feature vectors.

[0015] In the online monitoring system of the fiber cement board production line, the real-time temperature time series feature aggregation unit includes: a steam real-time temperature time series feature energy level coefficient calculation subunit, which is used to determine the characteristic energy level coefficient of each steam real-time temperature time series feature vector based on the maximum value, minimum value, mean value and variance of each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors to obtain a set of steam real-time temperature time series feature energy level coefficients; a steam real-time temperature time series feature energy level significance descriptor calculation subunit, which is used to determine the scale of the local neighborhood, calculate the local significance of each steam real-time temperature time series feature energy level coefficient in the set of steam real-time temperature time series feature energy level coefficients. The method comprises the following steps: a first step is to input the set of steam real-time temperature time series feature energy level significance descriptors into a mask module based on a gating function to obtain a set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors; and a temperature space domain time series distribution clustering subunit is used to use the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors as a sequence of weights to calculate the position-weighted sum of the multiple steam real-time temperature time series feature vectors to obtain the temperature space domain time series distribution clustering representation vector.

[0016] In the online monitoring system of the fiber cement board production line, the steam real-time temperature time series characteristic energy level coefficient calculation subunit is used to: respectively extract the maximum value and the minimum value in the steam real-time temperature time series characteristic vector to obtain the maximum value of the steam real-time temperature time series characteristic and the minimum value of the steam real-time temperature time series characteristic; respectively calculate the mean and variance of the steam real-time temperature time series characteristic vector to obtain the mean value of the steam real-time temperature time series characteristic and the variance of the steam real-time temperature time series characteristic; add the steam real-time temperature time series characteristic variance and the preset hyperparameter to obtain the first energy level coefficient of the steam real-time temperature time series characteristic; calculate the maximum value of the steam real-time temperature time series characteristic The maximum difference value of the steam real-time temperature time series is obtained by calculating the square of the difference between the minimum value of the steam real-time temperature time series characteristic and the mean value of the steam real-time temperature time series characteristic to obtain the minimum difference value of the steam real-time temperature time series; the maximum difference value of the steam real-time temperature time series is added to the minimum difference value of the steam real-time temperature time series and the preset hyperparameter to obtain the second energy level coefficient of the steam real-time temperature time series characteristic; the first energy level coefficient of the steam real-time temperature time series characteristic is divided by the second energy level coefficient of the steam real-time temperature time series characteristic to obtain the energy level coefficient of the steam real-time temperature time series characteristic.

[0017] In the online monitoring system of the above-mentioned fiber cement board production line, the steam real-time temperature time series feature energy level significance descriptor mask subunit includes: a normalization secondary subunit, which is used to normalize each steam real-time temperature time series feature energy level significance descriptor in the set of the steam real-time temperature time series feature energy level significance descriptors to obtain a set of normalized steam real-time temperature time series feature energy level significance descriptors; a steam real-time temperature time series feature energy level significance descriptor mask probabilistic secondary subunit, which is used to input each normalized steam real-time temperature time series feature energy level significance descriptor in the set of the normalized steam real-time temperature time series feature energy level significance descriptors into a gating function for masking processing to obtain the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors.

[0018] In the online monitoring system of the above-mentioned fiber cement board production line, the steam real-time temperature time series feature energy level significance descriptor mask probabilistic secondary sub-unit is used to: set each normalized steam real-time temperature time series feature energy level significance descriptor greater than a predetermined threshold in the set of normalized steam real-time temperature time series feature energy level significance descriptors to the original value, and set the rest to zero, so as to obtain the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors.

[0019] In the online monitoring system of the fiber cement board production line, the temperature distribution offset coefficient calculation unit includes: a steam real-time temperature difference calculation subunit, which is used to calculate the position difference between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a set of steam real-time temperature difference feature vectors; a steam real-time temperature inverse covariance calculation subunit, which is used to calculate the covariance matrix between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector, and calculate the The inverse of the covariance matrix is ​​used to obtain a set of steam real-time temperature inverse covariance matrices; a steam real-time temperature offset value calculation subunit is used to obtain a set of steam real-time temperature offset values ​​based on the offset value between each group of corresponding steam real-time temperature differential eigenvectors and the steam real-time temperature inverse covariance matrix in the set of the steam real-time temperature differential eigenvectors and the steam real-time temperature inverse covariance matrix; a temperature distribution topology simulation calculation subunit is used to calculate the square root of each steam real-time temperature offset value in the set of steam real-time temperature offset values ​​to obtain the temperature distribution topology simulation representation matrix composed of multiple temperature time series distribution offset coefficients.

[0020] In the online monitoring system of the above-mentioned fiber cement board production line, the steam real-time temperature offset value calculation subunit is used to: calculate the transposed vector of the steam real-time temperature difference eigenvector and the product of the steam real-time temperature inverse covariance matrix and the steam real-time temperature difference eigenvector to obtain the steam real-time temperature offset value.

[0021] In the online monitoring system of the fiber cement board production line, the steam control unit is used to input the temperature distribution topological representation matrix into the classifier-based steam control module to obtain the control instruction, which is used to indicate whether steam needs to be introduced.

[0022] According to another aspect of the present application, there is provided an online monitoring method for a fiber cement board production line, comprising:

[0023] Acquire a time series of multiple real-time steam temperatures collected by multiple temperature sensors deployed inside the cylinder;

[0024] Inputting each steam real-time temperature time series of the plurality of steam real-time temperature time series into a real-time temperature sequence encoder to obtain a plurality of steam real-time temperature time series feature vectors;

[0025] Inputting the plurality of steam real-time temperature time series feature vectors into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector;

[0026] Calculating the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients as a temperature distribution topological representation feature;

[0027] Based on the temperature distribution topological representation characteristics, a control instruction is obtained, and the control instruction is used to indicate whether steam needs to be introduced.

[0028] Compared with the prior art, the online monitoring system and method of a fiber cement board production line provided by the present application uses a time series of multiple real-time steam temperatures collected by multiple temperature sensors deployed inside the cylinder, and uses data processing and analysis technology based on deep learning to perform time series encoding and spatial aggregation of the real-time steam temperatures, so as to automatically determine whether steam needs to be introduced according to the temperature distribution representation characteristics calculated between the time series characteristics of the real-time steam temperatures and the temperature spatial aggregation representation characteristics. In this way, through the automated temperature monitoring system, the dependence on manual experience is reduced, and the accuracy and response speed of temperature control are improved. In addition, the system can automatically determine the need for steam introduction, thereby optimizing the temperature distribution and ensuring the temperature uniformity of each area in the curing room, thereby improving the intelligent level of cement board curing production control. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] 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.

[0030] Figure 1 is a block diagram of an online monitoring system for a fiber cement board production line according to an embodiment of the present application;

[0031] Figure 2 A data flow diagram of an online monitoring system for a fiber cement board production line according to an embodiment of the present application;

[0032] Figure 3 is a block diagram of a steam control module in an online monitoring system of a fiber cement board production line according to an embodiment of the present application;

[0033] Figure 4 The figure is a flow chart of an online monitoring method for a fiber cement board production line according to an embodiment of the present application. DETAILED DESCRIPTION

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] 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.

[0039] In the traditional steam curing process, temperature control mainly depends on the experience and technical level of the operator. This method is not only inefficient, but also difficult to ensure the uniformity of temperature distribution. Specifically, the operator must frequently check the temperature and manually adjust the steam supply, which not only consumes a lot of manpower, but is also prone to human errors, resulting in unstable temperature control. In addition, due to the lack of accurate temperature monitoring methods, operators often find it difficult to accurately grasp the true situation of temperature distribution in the cylinder, which may cause the temperature in some areas to be too high or too low, thereby affecting the strength and uniformity of the fiber cement board. Therefore, an online monitoring system for a fiber cement board production line is desired.

[0040] In the technical solution of the present application, an online monitoring system for a fiber cement board production line is proposed. Figure 1 4 is a block diagram of an online monitoring system for a fiber cement board production line according to an embodiment of the present application. Figure 2 FIG. 1 is a data flow diagram of an online monitoring system for a fiber cement board production line according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the online monitoring system 300 of the fiber cement board production line of the embodiment of the present application, it includes: a fiber cement board transportation and loading module 310, used to place the fiber cement board into the cylinder of the production equipment; a steam curing module 320, used to close the cylinder and pass steam to steam-cure the fiber cement board; a steam control module 330, used to control the steam through a compression device during the steam curing process to ensure uniform heat transfer inside the cylinder to obtain the fiber cement board after steam curing; an appearance inspection module 340, used to perform an appearance inspection on the steam-cured fiber cement board to obtain a test result.

[0041] In particular, the fiber cement board transport loading module 310 is used to place the fiber cement board into a cylinder of a production device. High pressure is usually applied in the cylinder, which helps to compact and densify the components in the fiber cement board. High pressure can improve the strength, durability and water resistance of the fiber cement board.

[0042] In particular, the steam curing module 320 is used to seal the cylinder and pass steam to steam-curing the fiber cement board. It should be understood that by passing steam into the cylinder, the fiber cement board is steam-cured to promote the cement hydration reaction in the fiber cement board, so that the board hardens and obtains strength.

[0043] In particular, the steam control module 330 is used to control the steam through a compression device during the steam curing process to ensure uniform heat transfer inside the cylinder to obtain a steam-cured fiber cement board. In particular, in a specific example of the present application, Figure 3As shown, the steam control module 330 includes: a steam real-time temperature acquisition unit 331, which is used to obtain a time series of multiple steam real-time temperatures collected by multiple temperature sensors deployed inside the cylinder; a real-time temperature time series encoding unit 332, which is used to input the time series of each steam real-time temperature in the multiple steam real-time temperature time series into a real-time temperature sequence encoder to obtain multiple steam real-time temperature time series feature vectors; a real-time temperature time series feature aggregation unit 333, which is used to input the multiple steam real-time temperature time series feature vectors into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector; a temperature distribution offset coefficient calculation unit 334, which is used to calculate the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients as a temperature distribution topological representation feature; a steam control unit 335, which is used to obtain a control instruction based on the temperature distribution topological representation feature, and the control instruction is used to indicate whether steam needs to be introduced.

[0044] Specifically, the steam real-time temperature acquisition unit 331 and the real-time temperature time series encoding unit 332 are used to obtain the time series of multiple steam real-time temperatures collected by multiple temperature sensors deployed inside the cylinder; and the time series of each steam real-time temperature in the time series of the multiple steam real-time temperatures are respectively input into the real-time temperature sequence encoder to obtain multiple steam real-time temperature time series feature vectors. Considering that the time series of each steam real-time temperature in the time series of multiple steam real-time temperatures has a mutual correlation and influence relationship between different time scales, and considering that the Bi-LSTM model can process time series data and capture the time series dynamic fluctuation characteristics and patterns of different time spans in the time series data. Therefore, in the technical solution of the present application, the time series of each steam real-time temperature in the time series of the multiple steam real-time temperatures is respectively input into the real-time temperature sequence encoder based on the Bi-LSTM model to obtain multiple steam real-time temperature time series feature vectors. It is worth mentioning that Bi-LSTM is a recurrent neural network that can process sequence data while considering the past and future contexts of the elements in the sequence. Unlike traditional RNNs, Bi-LSTM consists of two LSTM layers, one processing the forward direction of the sequence (from left to right) and the other processing the reverse direction of the sequence (from right to left). This enables Bi-LSTM to capture both the past and future context of elements in the sequence.

[0045] Specifically, the real-time temperature time series feature aggregation unit 333 is used to input the multiple steam real-time temperature time series feature vectors into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector. Considering that the multiple steam real-time temperature time series feature vectors are features captured by multiple temperature sensors and encoded by sequence encoders, each feature has different degrees of influence within the entire temperature spatial distribution range. Therefore, in order to focus on the distribution of each steam real-time temperature time series feature in the entire spatial domain, so as to more comprehensively and comprehensively represent the global distribution representation within the entire temperature space range, in the technical solution of the present application, the multiple steam real-time temperature time series feature vectors are input into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector. It is worth mentioning that the feature aggregation network based on feature energy local significance gating is a network that focuses on measuring and integrating local energy features to significantly enhance the model's insight into temperature analysis and data processing efficiency.

[0046] In an embodiment of the present application, the multiple steam real-time temperature time series feature vectors are input into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector, including: first, based on the maximum value, minimum value, mean and variance of each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors, the feature energy level coefficient of each steam real-time temperature time series feature vector is determined to obtain a set of steam real-time temperature time series feature energy level coefficients, so as to reflect the importance or significance of different steam temperature features; then, the scale of the local neighborhood is determined, which is to consider the position and importance of each feature vector in the context of multiple steam real-time temperature time series feature vectors. Then, the local neighborhood significance descriptor of each steam real-time temperature time series feature energy level coefficient in the set of the steam real-time temperature time series feature energy level coefficients is calculated to obtain a set of steam real-time temperature time series feature energy level significance descriptors, which is helpful to capture the energy distribution characteristics of the steam real-time temperature time series feature in the local neighborhood, so as to identify the temperature time series features that are significant in the local range; subsequently, the set of steam real-time temperature time series feature energy level significance descriptors is input into a mask module based on a gating function to obtain a set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors; that is, through nonlinear probabilistic transformation and adaptive feature screening, dynamic adjustment of the significance of steam real-time temperature features is achieved. In particular, the mask module based on the gating function first normalizes each feature energy level significance descriptor to eliminate the influence of different feature dimensions and numerical ranges to obtain a standardized coefficient, and then performs mask gating adaptive screening on each normalized coefficient to retain the most representative and informative features, and remove redundant or insignificant features to obtain a set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors; finally, using the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors as a sequence of weights, the position-weighted sum of the multiple steam real-time temperature time series feature vectors is calculated to obtain the temperature space domain time series distribution cluster representation vector; here, the masked features are used as weights to perform weighted summation on the multiple steam real-time temperature time series feature vectors to obtain the temperature space domain time series distribution cluster representation vector that focuses on the significant time series feature information of each local spatial domain.

[0047] Among them, based on the maximum value, minimum value, mean value and variance of each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors, the process of determining the characteristic energy level coefficient of each steam real-time temperature time series feature vector to obtain a set of steam real-time temperature time series characteristic energy level coefficients includes: respectively extracting the maximum value and the minimum value in the steam real-time temperature time series feature vector to obtain the maximum value of the steam real-time temperature time series feature and the minimum value of the steam real-time temperature time series feature; respectively calculating the mean and variance of the steam real-time temperature time series feature vector to obtain the mean value of the steam real-time temperature time series feature and the variance of the steam real-time temperature time series feature; adding the variance of the steam real-time temperature time series feature and the preset hyperparameter to obtain the steam real-time temperature time series feature. The first energy level coefficient of the temperature timing characteristic; calculating the square of the difference between the maximum value of the steam real-time temperature timing characteristic and the mean value of the steam real-time temperature timing characteristic to obtain the maximum difference value of the steam real-time temperature timing; calculating the square of the difference between the minimum value of the steam real-time temperature timing characteristic and the mean value of the steam real-time temperature timing to obtain the minimum difference value of the steam real-time temperature timing; adding the maximum difference value of the steam real-time temperature timing, the minimum difference value of the steam real-time temperature timing and the preset hyperparameter to obtain the second energy level coefficient of the steam real-time temperature timing characteristic; dividing the first energy level coefficient of the steam real-time temperature timing characteristic by the second energy level coefficient of the steam real-time temperature timing characteristic to obtain the energy level coefficient of the steam real-time temperature timing characteristic.

[0048] More specifically, the process of inputting the set of steam real-time temperature time series feature energy level significance descriptors into a mask module based on a gating function to obtain a set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors includes: normalizing each steam real-time temperature time series feature energy level significance descriptor in the set of steam real-time temperature time series feature energy level significance descriptors to obtain a set of normalized steam real-time temperature time series feature energy level significance descriptors; inputting each normalized steam real-time temperature time series feature energy level significance descriptor in the set of normalized steam real-time temperature time series feature energy level significance descriptors into a gating function for masking processing to obtain the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors. Among them, each normalized steam real-time temperature time series feature energy level significance descriptor in the set of normalized steam real-time temperature time series feature energy level significance descriptors is input into the gating function for masking processing to obtain the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors, including: setting each normalized steam real-time temperature time series feature energy level significance descriptor greater than a predetermined threshold in the set of normalized steam real-time temperature time series feature energy level significance descriptors to the original value, and setting the rest to zero, so as to obtain the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors.

[0049] In summary, in the above embodiment, the multiple steam real-time temperature time series feature vectors are input into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector, including: the multiple steam real-time temperature time series feature vectors are input into a feature aggregation network based on feature energy local significance gating, and are processed with the following feature aggregation formula to obtain the temperature space domain time series distribution cluster representation vector; wherein the feature aggregation formula is: ; ; ; ; ; ; ;in, are the multiple steam real-time temperature time series feature vectors, , , , , respectively represent the first, second, and third of the multiple steam real-time temperature time series feature vectors , , The real-time steam temperature time series feature vector, represents the number of feature vectors in the multiple steam real-time temperature time series feature vectors, and Respectively The mean and variance of the steam real-time temperature time series feature vector, and Respectively The maximum and minimum values ​​in the real-time steam temperature time series feature vector, is the preset hyperparameter, It is The characteristic energy level coefficient of the steam real-time temperature time series characteristic vector, It is The characteristic energy level coefficient of the steam real-time temperature time series characteristic vector, Indicates The number of characteristic energy coefficients in the local neighborhood centered at is the first in the set of significant descriptors of the steam real-time temperature time series feature energy level The energy level significance descriptor of the steam real-time temperature time series feature, represents the natural exponential function, It is the first in the set of significant descriptors of the normalized steam real-time temperature time series feature energy level. Normalized steam real-time temperature time series feature energy level significance descriptor, For masking, is the first in the set of the mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors. A mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptor, is the predetermined threshold, The cluster representation vector is the temperature spatial domain time series distribution.

[0050] Specifically, the temperature distribution offset coefficient calculation unit 334 is used to calculate the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients as the temperature distribution topological representation feature. It should be understood that in order to further quantify the difference or deviation between each steam real-time temperature time series feature vector and the temperature space domain time series distribution cluster representation vector, so that the temperature distribution of the fiber cement board during steam curing can be uniform and the steam supply can be controlled more accurately, in the technical solution of the present application, the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector is calculated to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients. That is to say, by calculating the temperature time series distribution offset coefficient between the two, abnormal conditions in the temperature distribution can be identified. Specifically, if the offset coefficient of a certain eigenvector is significantly higher than that of other eigenvectors, it indicates that there may be problems or abnormalities in the temperature distribution during this time period. In this way, based on the result of the offset coefficient, the system can automatically adjust the steam supply to reduce the deviation of the temperature distribution and ensure uniform temperature distribution in the cylinder. In particular, the temperature time series distribution offset coefficient is calculated based on the difference vector and covariance matrix between the steam real-time temperature time series eigenvector and the temperature space domain time series distribution cluster representation vector.

[0051] In an embodiment of the present application, the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector is calculated to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients as the temperature distribution topological representation feature, including: calculating the positional difference between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a set of steam real-time temperature difference feature vectors; calculating each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors The covariance matrix between the temperature time series feature vector and the temperature space domain time series distribution cluster representation vector is obtained, and the inverse of the covariance matrix is ​​calculated to obtain a set of steam real-time temperature inverse covariance matrices; based on the offset value between each group of corresponding steam real-time temperature difference feature vectors and the steam real-time temperature inverse covariance matrix in the set of the steam real-time temperature difference feature vectors and the set of the steam real-time temperature inverse covariance matrix, a set of steam real-time temperature offset values ​​is obtained; the square root of each steam real-time temperature offset value in the set of steam real-time temperature offset values ​​is calculated to obtain the temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients.

[0052] Among them, based on the offset value between each group of corresponding steam real-time temperature difference eigenvectors and the steam real-time temperature inverse covariance matrix in the set of the steam real-time temperature difference eigenvectors and the set of the steam real-time temperature inverse covariance matrix, the process of obtaining a set of steam real-time temperature offset values ​​includes: calculating the transposed vector of the steam real-time temperature difference eigenvector and the product of the steam real-time temperature inverse covariance matrix and the steam real-time temperature difference eigenvector to obtain the steam real-time temperature offset value.

[0053] In summary, in the above embodiment, the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector is calculated to obtain a temperature distribution pseudo-topological representation matrix composed of multiple temperature time series distribution offset coefficients as the temperature distribution pseudo-topological representation feature, including: calculating the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector using the following semantic difference calculation formula to obtain a plurality of temperature time series distribution offset coefficients; wherein, the semantic difference calculation formula is: ;in, is the first of the multiple steam real-time temperature time series feature vectors The real-time steam temperature time series feature vector, is the temperature spatial domain time series distribution cluster representation vector, is the transpose of a vector, For the said The covariance matrix between the steam real-time temperature time series feature vector and the temperature space domain time series distribution cluster representation vector, It is The temperature time series distribution offset coefficient between the steam real-time temperature time series feature vector and the temperature space domain time series distribution cluster representation vector.

[0054] Specifically, the steam control unit 335 is used to obtain a control instruction based on the temperature distribution topological representation feature, and the control instruction is used to indicate whether steam needs to be introduced. In a specific example of the present application, the temperature distribution topological representation matrix is ​​input into a steam control module based on a classifier to obtain the control instruction, and the control instruction is used to indicate whether steam needs to be introduced. That is, the temperature distribution topological representation matrix obtained by calculating the temperature offset between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector is used for classification processing to automatically determine whether steam needs to be introduced. In this way, through the automated temperature monitoring system, the dependence on manual experience is reduced, and the accuracy and response speed of temperature control are improved. And the system can automatically determine the need for steam introduction, so as to optimize the temperature distribution and ensure the temperature uniformity of each area in the curing room, thereby improving the intelligent level of cement board curing production control.

[0055] Preferably, inputting the temperature distribution topological representation matrix into a classifier-based steam control module to obtain control instructions comprises:

[0056] Expanding the temperature distribution topological representation matrix into a temperature distribution topological representation vector;

[0057] Subtract the zero norm of the temperature distribution topologically simulated representation vector from the length of the temperature distribution topologically simulated representation vector to obtain a temperature distribution topologically simulated isolated representation value;

[0058] Calculate the logarithm of the square of the temperature distribution simulated topological isolated representation value and the sum of the temperature distribution simulated topological isolated representation value with base 2 to obtain the temperature distribution simulated topological information order value;

[0059] Calculate a power function with each eigenvalue of the temperature distribution topology-simulated representation vector as the base, the difference of the temperature distribution topology-simulated isolated representation value minus one as the exponent, and perform point multiplication with the temperature distribution topology-simulated information order value to obtain the temperature distribution topology-simulated leading vector;

[0060] After performing a dot multiplication on the temperature distribution topological representation vector and the difference between the temperature distribution topological isolated representation value minus one, the dot multiplication is then performed on the reciprocal of the temperature distribution topological isolated representation value to obtain a temperature distribution topological field constraint vector;

[0061] Calculate an exponential function with a natural constant as the base and each eigenvalue of the temperature distribution topological field constraint vector as the exponent to obtain a temperature distribution topological field bias vector;

[0062] Performing point addition of the temperature distribution topology-simulating pilot vector and the temperature distribution topology-simulating field bias vector to obtain an optimized temperature distribution topology-simulating representation vector;

[0063] The optimized temperature distribution topology representation vector is input into a classifier-based steam control module to obtain a control instruction.

[0064] Specifically, the optimization process of the temperature distribution topological representation vector is expressed as: ; ;in, is the temperature distribution topological representation vector, To calculate each eigenvalue of the temperature distribution topological representation vector as the base, the difference between the temperature distribution topological isolated representation value minus one is a power function with exponent, is the length of the temperature distribution topological representation vector, represents the zero norm of the temperature distribution topological representation vector, is the temperature distribution topologically isolated representation value, represents the natural exponential function, represents dot product, Indicates point addition, is the topological representation vector of the optimized temperature distribution.

[0065] Here, the multiple steam real-time temperature time series feature vectors respectively represent the short-range and long-range bidirectional time series correlation characteristics of the steam real-time temperature in the local time domain, and the temperature space domain time series distribution cluster representation vector represents the global time domain aggregation characteristics of the steam real-time temperature based on the local time domain time series distribution energy significance. Therefore, the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector and the temperature space domain time series distribution cluster representation vector will also lead to inconsistent distribution offset response due to the uneven local time domain time series distribution significance in the global time domain. It is expected to further improve the full time domain offset response aggregation regression comprehensibility of the temperature distribution topological representation matrix, thereby improving the accuracy of the control instructions obtained by the steam control module based on the classifier input.

[0066] Therefore, the high-dimensional characteristic manifold of the temperature distribution topological representation vector expanded by the temperature distribution topological representation matrix is ​​represented by the eigenvalue of the feature set as the vector field of the aggregation dimension, and the superposition value of the vector field of the temperature distribution topological representation vector at the isolated zero position is used as the order information to fix the local position of the eigenvalue of its feature set, and add the bias of the reversibility of the characteristic regression distribution field of the temperature distribution topological representation vector as a reward to achieve the mapping target tracking of the regression distribution of the temperature distribution topological representation vector for the eigenvalue position, so that the feature set of the temperature distribution topological representation vector perceives the mapping migration to the aggregate distribution, thereby improving the accuracy of the control instructions obtained by the steam control module based on the classifier input of the temperature distribution topological feature matrix by improving the full time domain offset response aggregation regression comprehensibility of the temperature distribution topological representation vector. In this way, through the automated temperature monitoring system, the dependence on manual experience is reduced, and the accuracy and response speed of temperature control are improved. In addition, the system can automatically determine the steam entry requirements, thereby optimizing the temperature distribution, ensuring the temperature uniformity of each area in the curing room, and thus improving the intelligent level of fiber cement board curing production control.

[0067] In particular, the appearance inspection module 340 is used to perform an appearance inspection on the steam-cured fiber cement board to obtain a test result. It should be understood that the appearance inspection includes whether there are cracks, holes, bubbles, peeling on the board surface, whether the color and texture of the board are uniform, whether the edges and corners of the board are damaged, notched or deformed, etc. The appearance inspection can further improve the accuracy of board quality inspection and optimize product quality.

[0068] As described above, the online monitoring system 300 for the fiber cement board production line according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with an online monitoring algorithm for the fiber cement board production line. In a possible implementation, the online monitoring system 300 for the fiber cement board production line according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the online monitoring system 300 for the fiber cement board production line can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the online monitoring system 300 for the fiber cement board production line can also be one of the many hardware modules of the wireless terminal.

[0069] Alternatively, in another example, the online monitoring system 300 of the fiber cement board production line and the wireless terminal may be separate devices, and the online monitoring system 300 of the fiber cement board production line may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0070] Furthermore, an online monitoring method for a fiber cement board production line is also provided.

[0071] Figure 4 FIG. 1 is a flow chart of an online monitoring method for a fiber cement board production line according to an embodiment of the present application. Figure 4 As shown, according to the embodiment of the present application, the online monitoring method of the fiber cement board production line includes the steps of: S1, obtaining a plurality of time series of real-time steam temperatures collected by a plurality of temperature sensors deployed inside the cylinder; S2, respectively inputting the time series of each of the plurality of time series of real-time steam temperatures into a real-time temperature sequence encoder to obtain a plurality of real-time steam temperature time series feature vectors; S3, inputting the plurality of real-time steam temperature time series feature vectors into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector; S4, calculating the temperature time series distribution offset coefficient between each of the plurality of real-time steam temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a temperature distribution topological representation matrix composed of a plurality of temperature time series distribution offset coefficients as a temperature distribution topological representation feature; S5, obtaining a control instruction based on the temperature distribution topological representation feature, the control instruction being used to indicate whether steam needs to be introduced.

[0072] In summary, the online monitoring method of the fiber cement board production line according to the embodiment of the present application is explained, which uses the time series of multiple real-time steam temperatures collected by multiple temperature sensors deployed inside the cylinder, and adopts data processing and analysis technology based on deep learning to perform time series encoding and spatial aggregation of the real-time temperatures of each steam, so as to automatically determine whether steam needs to be introduced according to the temperature distribution representation characteristics calculated between the time series characteristics of the real-time steam temperatures and the temperature space aggregation representation characteristics. In this way, through the automated temperature monitoring system, the dependence on manual experience is reduced, and the accuracy and response speed of temperature control are improved. In addition, the system can automatically determine the need for steam introduction, so as to optimize the temperature distribution and ensure the temperature uniformity of each area in the curing room, thereby improving the level of intelligence of cement board curing production control.

[0073] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. An online monitoring system for a fiber cement board production line, comprising: A fiber cement board transport loading module is used to place the fiber cement board into the cylinder of the production equipment; A steam curing module is used to seal the cylinder and pass steam to steam-cure the fiber cement board; a steam control module is used to control the steam through a compression device during the steam curing process to ensure uniform heat transfer inside the cylinder to obtain a steam-cured fiber cement board; An appearance inspection module, used for performing an appearance inspection on the steam-cured fiber cement board to obtain a test result; Wherein, the steam control module is characterized by comprising: A steam real-time temperature acquisition unit, used to acquire a time series of multiple steam real-time temperatures acquired by multiple temperature sensors deployed inside the cylinder; A real-time temperature time series encoding unit, used for inputting each steam real-time temperature time series of the multiple steam real-time temperature time series into a real-time temperature sequence encoder to obtain multiple steam real-time temperature time series feature vectors; A real-time temperature time series feature aggregation unit, used for inputting the plurality of steam real-time temperature time series feature vectors into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector; A temperature distribution offset coefficient calculation unit is used to calculate the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients as a temperature distribution topological representation feature; A steam control unit, used to obtain a control instruction based on the temperature distribution topological representation characteristics, wherein the control instruction is used to indicate whether steam needs to be introduced; The real-time temperature time series encoding unit is used to: input each steam real-time temperature time series in the multiple steam real-time temperature time series into a real-time temperature sequence encoder based on a Bi-LSTM model to obtain multiple steam real-time temperature time series feature vectors; Wherein, the real-time temperature time series feature aggregation unit includes: A steam real-time temperature time series characteristic energy level coefficient calculation subunit, used to determine the characteristic energy level coefficient of each steam real-time temperature time series characteristic vector in the multiple steam real-time temperature time series characteristic vectors based on the maximum value, minimum value, mean value and variance of each steam real-time temperature time series characteristic vector to obtain a set of steam real-time temperature time series characteristic energy level coefficients; A steam real-time temperature time series characteristic energy level significance descriptor calculation subunit is used to determine the scale of the local neighborhood, calculate the local neighborhood significance descriptor of each steam real-time temperature time series characteristic energy level coefficient in the set of steam real-time temperature time series characteristic energy level coefficients to obtain a set of steam real-time temperature time series characteristic energy level significance descriptors; A steam real-time temperature time series feature energy level significance descriptor mask subunit, used for inputting the set of steam real-time temperature time series feature energy level significance descriptors into a mask module based on a gating function to obtain a set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors; The temperature space domain time series distribution clustering subunit is used to calculate the position-weighted sum of the multiple steam real-time temperature time series feature vectors using the set of the mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors as a sequence of weights to obtain the temperature space domain time series distribution clustering representation vector; Wherein, the temperature distribution offset coefficient calculation unit includes: A steam real-time temperature difference calculation subunit, used for calculating the position difference between each steam real-time temperature time series feature vector of the plurality of steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a set of steam real-time temperature difference feature vectors; A steam real-time temperature inverse covariance calculation subunit is used to calculate the covariance matrix between each steam real-time temperature time series feature vector of the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector, and calculate the inverse of the covariance matrix to obtain a set of steam real-time temperature inverse covariance matrices; A steam real-time temperature offset value calculation subunit, used to obtain a set of steam real-time temperature offset values ​​based on the offset value between each corresponding set of steam real-time temperature differential eigenvectors and the steam real-time temperature inverse covariance matrix in the set of the steam real-time temperature differential eigenvectors and the steam real-time temperature inverse covariance matrix; The temperature distribution topology simulation calculation subunit is used to calculate the square root of each steam real-time temperature offset value in the set of steam real-time temperature offset values ​​to obtain the temperature distribution topology simulation representation matrix composed of multiple temperature time series distribution offset coefficients.

2. The online monitoring system for fiber cement board production line according to claim 1, characterized in that: The steam real-time temperature time series characteristic energy level coefficient calculation subunit is used to: Respectively extracting the maximum value and the minimum value in the steam real-time temperature time series feature vector to obtain the maximum value of the steam real-time temperature time series feature and the minimum value of the steam real-time temperature time series feature; Calculating the mean and variance of the steam real-time temperature time series feature vector respectively to obtain the steam real-time temperature time series feature mean and the steam real-time temperature time series feature variance; Adding the steam real-time temperature time series characteristic variance and a preset hyperparameter to obtain a first energy level coefficient of the steam real-time temperature time series characteristic; Calculate the square of the difference between the maximum value of the steam real-time temperature time series characteristic and the mean value of the steam real-time temperature time series characteristic to obtain the maximum difference value of the steam real-time temperature time series; Calculate the square of the difference between the minimum value of the steam real-time temperature time series characteristic and the mean value of the steam real-time temperature time series characteristic to obtain the minimum difference value of the steam real-time temperature time series; The maximum difference value of the steam real-time temperature time series is added to the minimum difference value of the steam real-time temperature time series and the preset hyperparameter to obtain the second energy level coefficient of the steam real-time temperature time series characteristic; The first energy level coefficient of the steam real-time temperature time series characteristic is divided by the second energy level coefficient of the steam real-time temperature time series characteristic to obtain the energy level coefficient of the steam real-time temperature time series characteristic.

3. The online monitoring system for fiber cement board production line according to claim 2, characterized in that: The steam real-time temperature time series feature energy level significance descriptor mask subunit includes: A normalization secondary subunit, used for normalizing each steam real-time temperature time series feature energy level significance descriptor in the set of steam real-time temperature time series feature energy level significance descriptors to obtain a set of normalized steam real-time temperature time series feature energy level significance descriptors; The steam real-time temperature time series feature energy level significance descriptor mask probabilistic secondary sub-unit is used to input each normalized steam real-time temperature time series feature energy level significance descriptor in the set of normalized steam real-time temperature time series feature energy level significance descriptors into a gating function for masking processing to obtain the set of masked gated probabilistic steam real-time temperature time series feature energy level significance descriptors.

4. The online monitoring system for fiber cement board production line according to claim 3, characterized in that: The steam real-time temperature time series feature energy level significance descriptor mask probabilistic secondary sub-unit is used to: set each normalized steam real-time temperature time series feature energy level significance descriptor greater than a predetermined threshold in the set of normalized steam real-time temperature time series feature energy level significance descriptors to the original value, and set the rest to zero, so as to obtain the set of mask-gated probabilistic steam real-time temperature time series feature energy level significance descriptors.

5. The online monitoring system for fiber cement board production line according to claim 4, characterized in that: The steam real-time temperature offset value calculation subunit is used to calculate the transposed vector of the steam real-time temperature difference eigenvector and the product of the steam real-time temperature inverse covariance matrix and the steam real-time temperature difference eigenvector to obtain the steam real-time temperature offset value.

6. The online monitoring system for fiber cement board production line according to claim 5, characterized in that: The steam control unit is used to: input the temperature distribution topological representation matrix into a classifier-based steam control module to obtain the control instruction, and the control instruction is used to indicate whether steam needs to be introduced.

7. An online monitoring method for a fiber cement board production line, using the online monitoring system for a fiber cement board production line according to claim 1, characterized in that: include: Acquire a time series of multiple real-time steam temperatures collected by multiple temperature sensors deployed inside the cylinder; Inputting each steam real-time temperature time series of the plurality of steam real-time temperature time series into a real-time temperature sequence encoder to obtain a plurality of steam real-time temperature time series feature vectors; Inputting the plurality of steam real-time temperature time series feature vectors into a feature aggregation network based on feature energy local significance gating to obtain a temperature space domain time series distribution cluster representation vector; Calculating the temperature time series distribution offset coefficient between each steam real-time temperature time series feature vector in the multiple steam real-time temperature time series feature vectors and the temperature space domain time series distribution cluster representation vector to obtain a temperature distribution topological representation matrix composed of multiple temperature time series distribution offset coefficients as a temperature distribution topological representation feature; Based on the temperature distribution topological representation characteristics, a control instruction is obtained, and the control instruction is used to indicate whether steam needs to be introduced.

Images