Production process data acquisition and analysis system and method

Abstract

The invention provides a production process data acquisition and analysis system and method, and relates to the field of intelligent production, and the method comprises the steps: carrying out the time sequence feature extraction of air temperature and air pressure data of a drying chamber at a plurality of time points, so as to obtain the time sequence features of porous brick drying control related parameters; meanwhile, feature extraction is carried out on the porous brick surface images at multiple time points to obtain porous brick drying quality dynamic change features, and then response features of porous brick drying quality and drying control related parameters are established in the same dimension space; and finally, a specific wind temperature regulation and control direction is obtained through a wind temperature guidance regulation and control result generator based on a classifier according to the response characteristics. In this way, accurate control over the air temperature parameters in the drying process of the porous bricks can be achieved, and then the production quality and efficiency of the bricks are improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent production, and more specifically, to a production process data collection and analysis system and method. Background Art

[0002] In the traditional brick production process, bricks contain a lot of water after they are formed. If this water is not properly removed, the bricks may crack during subsequent use due to evaporation or freezing expansion of the water, thereby reducing their strength and durability. Through the drying process, it can be ensured that the moisture content inside the bricks reaches a stable and appropriate level, thereby improving their overall strength and service life. The traditional drying process often relies on experience or fixed parameter settings, lacks accurate monitoring and dynamic adjustment of the real-time environment and material status, which can easily lead to unstable drying effects, and may cause problems such as over-drying, cracking or insufficient drying, which in turn affects the quality of the products produced.

[0003] Therefore, a production process data collection and analysis solution is needed. Summary of the invention

[0004] In order to solve the above technical problems, the present application is proposed. The embodiments of the present application provide a production process data collection and analysis system and method.

[0005] According to one aspect of the present application, a production process data acquisition and analysis system is provided, comprising:

[0006] A porous brick drying related data acquisition module is used to obtain the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at multiple predetermined time points within a predetermined time period;

[0007] A porous brick drying related data encoding module is used to perform response association encoding on the drying chamber wind temperature values, drying chamber wind pressure values ​​and porous brick surface images at the plurality of predetermined time points to obtain a porous brick drying quality response feature matrix;

[0008] The wind temperature control guidance result generation module is used to obtain the wind temperature control guidance result at the current time point based on the porous brick drying quality response characteristic matrix.

[0009] According to another aspect of the present application, a method for collecting and analyzing production process data is provided, comprising:

[0010] Acquire the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at a plurality of predetermined time points within a predetermined time period;

[0011] Performing response association coding on the drying chamber wind temperature values, the drying chamber wind pressure values ​​and the porous brick surface images at the plurality of predetermined time points to obtain a porous brick drying quality response feature matrix;

[0012] Based on the porous brick drying quality response characteristic matrix, the wind temperature control guidance result at the current time point is obtained.

[0013] Compared with the prior art, the production process data acquisition and analysis system and method provided by the present application extracts the time series features of the wind temperature and wind pressure data of the drying chamber at multiple time points to obtain the time series features of the parameters related to the drying control of porous bricks, and extracts the features of the surface images of porous bricks at multiple time points to obtain the dynamic change features of the drying quality of porous bricks, and then establishes the response features of the porous brick drying quality and the parameters related to the drying control in the same dimensional space, and finally passes the response features through the classifier-based wind temperature guidance control result generator to obtain the specific wind temperature control direction. In this way, the wind temperature parameters in the drying process of porous bricks can be accurately controlled, thereby improving the production quality and efficiency of bricks. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0015] Figure 1 4 is a system block diagram of a production process data acquisition and analysis system according to an embodiment of the present application.

[0016] Figure 2 It is a block diagram of the porous brick drying related data encoding module in the production process data acquisition and analysis system according to the embodiment of the present application.

[0017] Figure 3 It is a block diagram of a time series feature extraction unit for porous brick drying control parameters in a production process data acquisition and analysis system according to an embodiment of the present application.

[0018] Figure 4 It is a block diagram of a porous brick surface quality feature extraction unit in a production process data acquisition and analysis system according to an embodiment of the present application.

[0019] Figure 5 It is a block diagram of a wind temperature control guidance result generation module in a production process data acquisition and analysis system according to an embodiment of the present application.

[0020] Figure 6The present invention is a flowchart of a method for collecting and analyzing production process data according to an embodiment of the present application. DETAILED DESCRIPTION

[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] In the traditional brick production process, bricks usually contain a lot of water after they are formed. If this water is not properly removed, the bricks are likely to crack in the subsequent use stage due to the gradual evaporation of water or freezing and expansion in a low temperature environment. Once this happens, the strength of the bricks will be greatly reduced and the durability will be seriously affected. By implementing the drying process, it is possible to ensure that the moisture content inside the bricks is stable at an appropriate level. Only in this way can the overall strength of the bricks be effectively improved and their service life be extended. However, in reality, the traditional drying process often relies too much on past experience or fixed parameter settings. This method lacks accurate monitoring and dynamic adjustment of the real-time environment and material status. At the same time, there are differences in the material of the bricks themselves, the initial moisture content after forming, etc., but the fixed parameter settings cannot adapt to these changes. Because of this, the traditional drying process is prone to unstable drying effects, and there may be problems such as over-drying, making the bricks fragile and brittle, cracking, or insufficient drying, which causes too much moisture to remain inside the bricks, which in turn has a great negative impact on the quality of the products produced. Therefore, a production process data acquisition and analysis solution is expected, which can realize intelligent optimization of the drying process by real-time analysis of the drying quality and drying control parameters of porous bricks during the drying process.

[0023] In recent years, deep learning and neural networks have been widely used in computer vision, natural language processing, text signal processing and other fields. In addition, deep learning and neural networks have also demonstrated a level close to or even beyond that of humans in image classification, object detection, semantic segmentation, text translation and other fields. The development of deep learning and neural networks provides new solutions and solutions for production process data collection and analysis.

[0024] Figure 1 : is a system block diagram of a production process data acquisition and analysis system according to an embodiment of the present application. Figure 1As shown, in the production process data acquisition and analysis system 100, it includes: a porous brick drying related data acquisition module 110, which is used to obtain the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at multiple predetermined time points within a predetermined time period; a porous brick drying related data encoding module 120, which is used to respond to the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at the multiple predetermined time points. The response association encoding is performed to obtain the porous brick drying quality response feature matrix; the wind temperature control guidance result generation module 130 is used to obtain the wind temperature control guidance result at the current time point based on the porous brick drying quality response feature matrix.

[0025] Specifically, in the technical solution of the present application, it is first necessary to obtain the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface image at multiple predetermined time points within a predetermined time period. It should be understood that wind temperature is one of the most important parameters in the drying process, which directly affects the evaporation rate of water inside the brick. Appropriate wind temperature can promote uniform and rapid evaporation of water, thereby avoiding cracking or deformation of bricks due to excessive or uneven distribution of internal water. By real-time monitoring of the wind temperature value, the temperature condition of the current drying environment can be understood, and then adjusted as needed to ensure that the drying process is carried out within the optimal temperature range. The wind pressure value reflects the intensity and distribution of the airflow in the drying chamber. During the drying process, appropriate wind pressure can ensure that the airflow flows evenly in the drying chamber, thereby helping all parts of the brick to be evenly heated and accelerating the evaporation of water. The surface image of the porous brick provides an intuitive reflection of the drying quality of the brick. As the drying process proceeds, the color, texture, humidity and other characteristics of the brick surface will change, and these changes can serve as an important basis for evaluating the drying effect. By capturing the surface image of porous bricks and extracting features using image processing technology, the drying state of bricks can be understood in real time, including the degree of water evaporation, the formation of surface cracks, etc. In general, the wind temperature value, the wind pressure value of the drying chamber, and the surface image of porous bricks during drying are obtained in order to fully understand the real-time state of the drying process. These data together constitute the basis and foundation for wind temperature regulation. By real-time monitoring and analysis of these data, precise control and optimization of the brick drying process can be achieved, thereby ensuring high-quality production of bricks.

[0026] In an embodiment of the present application, a possible implementation method for obtaining the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at multiple predetermined time points within a predetermined time period may be: the drying chamber wind temperature data may be collected and obtained by a temperature sensor installed in the drying chamber; the drying chamber wind pressure data may be collected and obtained by a ventilation duct installed in the drying chamber; and the porous brick surface image data may be obtained by a high-definition camera installed in the drying chamber.

[0027] Figure 2 FIG. 1 is a block diagram of a porous brick drying-related data encoding module in a production process data acquisition and analysis system according to an embodiment of the present application. Figure 2 As shown, the porous brick drying related data encoding module 120 includes: a porous brick drying control parameter time series feature extraction unit 121, which is used to extract the correlation features of the drying chamber wind temperature value and the drying chamber wind pressure value at the plurality of predetermined time points to obtain the porous brick drying control parameter time series correlation feature vector; a porous brick surface quality feature extraction unit 122, which is used to extract the surface features of the porous brick surface images at the plurality of predetermined time points to obtain the porous brick surface quality dynamic change feature vector; a responsiveness estimation unit 123, which is used to calculate the responsiveness estimation of the porous brick surface quality dynamic change feature vector relative to the porous brick drying control parameter time series correlation feature vector to obtain the porous brick drying quality response feature matrix.

[0028] Figure 3 FIG. 1 is a block diagram of a time series feature extraction unit for porous brick drying control parameters in a production process data acquisition and analysis system according to an embodiment of the present application. Figure 3 As shown, the porous brick drying control parameter timing feature extraction unit 121 includes: a porous brick drying control parameter data structuring subunit 1211, used to arrange the drying chamber wind temperature values ​​and the drying chamber wind pressure values ​​of the multiple predetermined time points according to the time dimension to obtain a drying chamber wind temperature timing input vector and a drying chamber wind pressure timing input vector; a porous brick drying control parameter association subunit 1212, used to associate the drying chamber wind temperature timing input vector with the drying chamber wind pressure timing input vector to obtain a porous brick drying control parameter timing association matrix; a porous brick drying control parameter timing encoding subunit 1213, used to pass the porous brick drying control parameter timing association matrix through a porous brick drying control parameter timing encoder to obtain the porous brick drying control parameter timing association feature vector.

[0029] Considering that the drying of bricks is a dynamic process that changes over time, the drying environment will change in real time. In order to clearly understand the changing trends of the drying control parameters wind pressure and wind temperature over time, so as to capture the dynamic characteristics of the control parameters in the drying process, in the technical solution of the present application, the drying chamber wind temperature values ​​and drying chamber wind pressure values ​​at the multiple predetermined time points are arranged according to the time dimension to obtain the drying chamber wind temperature time series input vector and the drying chamber wind pressure time series input vector. By arranging the discrete drying control parameter data according to the time dimension, time series analysis can be used to capture the time dependence and correlation in the wind temperature and wind pressure data, so as to more accurately describe and control the drying process.

[0030] Considering that wind temperature and wind pressure together constitute the drying environment of the drying room, there is a certain interaction and mutual influence between them. Considering one of the parameters alone often cannot fully reflect the true state of the drying environment. In order to more comprehensively reflect the overall condition of the drying environment and provide more comprehensive and accurate information for subsequent drying process analysis and regulation, it is necessary to associate the drying chamber wind temperature timing input vector and the drying chamber wind pressure timing input vector to obtain the porous brick drying control parameter timing association matrix. It should be well understood that if the wind temperature is regulated only based on the wind temperature parameter timing data and the porous brick surface image data, this will ignore the influence of the wind pressure parameter on the wind temperature, resulting in poor regulation effect.

[0031] The porous brick drying control parameter temporal correlation matrix is ​​a matrix containing the correlation between the control parameters such as the drying chamber air temperature and air pressure at multiple time points. This matrix not only reflects the trend of each parameter over time, but also reveals the interaction and mutual influence between the parameters. Specifically, it may contain information such as the correlation coefficient, hysteresis effect, and trend consistency between the air temperature and air pressure. In order to extract key features from the original matrix data, so as to better characterize the dynamic characteristics of the control parameters of the drying process and the correlation between the parameters, it is necessary to pass the porous brick drying control parameter temporal correlation matrix through the porous brick drying control parameter temporal encoder to obtain the porous brick drying control parameter temporal correlation feature vector. In this application, the porous brick drying control parameter temporal encoder is a convolutional neural network model using mutually transposed convolution kernels in adjacent layers. The characteristic of this structure is that the convolution kernels of adjacent layers are mutually transposed, that is, the convolution kernel of the upper layer is used as a transposed convolution kernel in the next layer. This design helps to maintain the spatial structure and temporal relationship of the features during the encoding process, while enhancing the sensitivity and robustness of the network to the input data. That is, using the porous brick drying control parameter temporal encoder to process the matrix can reflect the spatiotemporal characteristics of the original data and learn accurate and rich feature representations.

[0032] In an embodiment of the present application, a possible implementation method of passing the porous brick drying control parameter timing association matrix through a porous brick drying control parameter timing encoder to obtain a porous brick drying control parameter timing association feature vector may be: each layer of the porous brick drying control parameter timing encoder performs the following on the input data in the forward pass of the layer: performing convolution processing, pooling processing and nonlinear processing based on a first two-dimensional convolution kernel on the input data to obtain a first activation feature map; performing convolution processing, pooling processing and nonlinear activation processing based on a second two-dimensional convolution kernel on the first activation feature map to obtain a second activation feature map, wherein the first two-dimensional convolution kernel and the second two-dimensional convolution kernel are transposed of each other; wherein the input of the first layer of the porous brick drying control parameter timing encoder is the porous brick drying control parameter timing association matrix, and the output of the last layer of the porous brick drying control parameter timing encoder is the porous brick drying control parameter timing association feature vector.

[0033] Figure 4 FIG. 1 is a block diagram of a porous brick surface quality feature extraction unit in a production process data acquisition and analysis system according to an embodiment of the present application. Figure 4 As shown, the porous brick surface quality feature extraction unit 122 includes: a porous brick surface image preprocessing subunit 1221, used for performing image preprocessing on the porous brick surface images at the multiple predetermined time points to obtain multiple porous brick surface enhanced images; a porous brick surface quality dynamic change feature generation subunit 1222, used for passing the multiple porous brick surface enhanced images through a porous brick surface quality change dynamic capturer to obtain a porous brick surface quality dynamic change feature map; a porous brick surface quality feature map pooling subunit 1223, used for performing mean pooling on each feature matrix of the porous brick surface quality dynamic change feature map along the channel dimension to obtain the porous brick surface quality dynamic change feature vector.

[0034] Considering that the collected porous brick surface images may be affected by various factors, such as uneven lighting, noise interference, etc., these factors will reduce the clarity and accuracy of the image, thereby affecting the accuracy of the subsequent porous brick drying quality analysis. In order to improve the quality of the porous brick surface images, it is necessary to perform image preprocessing on the porous brick surface images at the multiple predetermined time points to obtain multiple porous brick surface enhanced images. The preprocessed porous brick surface enhanced image will have less noise, and some surface details will be more obvious in the enhanced image, thereby improving the quality of subsequent porous brick surface data processing. In particular, in one possible implementation of the present application, a bilateral filtering algorithm is used to preprocess the porous brick surface image.

[0035] Next, the plurality of porous brick surface enhanced images are passed through a porous brick surface quality change dynamic capturer to obtain a porous brick surface quality dynamic change feature map. In the present application, the porous brick surface quality change dynamic capturer is a convolutional neural network model using a three-dimensional convolution kernel. The three-dimensional convolution kernel can simultaneously process information in space (width and height of the image) and time (multiple time points), so as to effectively extract the dynamic characteristics of the porous brick surface quality changing over time. Compared with the two-dimensional convolution kernel, the three-dimensional convolution kernel can better capture the time dependency in the image sequence, which is conducive to analyzing the trend of the porous brick surface quality changing over time. It is worth mentioning that by sharing weights in the time and space dimensions, the three-dimensional convolution kernel can reduce the number of model parameters, thereby reducing the computational complexity, which makes the model more efficient when processing large-scale image sequences, and is conducive to realizing real-time porous brick surface quality monitoring. Through the processing of the porous brick surface quality change dynamic capturer, the defect changes, texture changes, and small shape changes on the porous brick surface can be captured. By analyzing the changing trends of the surface quality of porous bricks at different times, it is possible to predict possible quality problems in the future and correlate them with the responses of the drying control parameters to further optimize the air temperature regulation and take preventive measures in advance, thereby improving the drying quality of porous bricks.

[0036] In an embodiment of the present application, a possible implementation method of passing the multiple porous brick surface enhanced images through a porous brick surface quality change dynamic capturer to obtain a porous brick surface quality dynamic change feature map can be: each layer of the porous brick surface quality change dynamic capturer performs the following on the input data during the forward transmission process of the layer: performing convolution processing based on a three-dimensional convolution kernel on the input data to obtain a convolution feature map; performing mean pooling processing based on a local feature matrix on the convolution feature map to obtain a pooling feature map; performing nonlinear activation on the pooling feature map to obtain an activation feature map; wherein the output of the last layer of the porous brick surface quality change dynamic capturer is the porous brick surface quality dynamic change feature map, and the input of the first layer of the porous brick surface quality change dynamic capturer is the multiple porous brick surface enhanced images.

[0037] In the porous brick surface quality dynamic change feature map, each channel represents the surface quality characteristics at different time points. In order to reduce the dimension of the porous brick surface quality dynamic change feature map, retain the important feature information in the original feature map, and remove redundant information, so that the porous brick surface quality feature representation is more compact and effective, the technical solution of the present application needs to perform mean pooling on each feature matrix of the porous brick surface quality dynamic change feature map along the channel dimension to obtain the porous brick surface quality dynamic change feature vector. The feature vector after dimensionality reduction has fewer elements, so less computing resources are required in subsequent processing and analysis, which helps to improve the computing efficiency of the entire data processing flow.

[0038] Responsiveness estimation is a quantitative analysis technique used to evaluate the sensitivity and dependence of a system or process output on changes in its input. Precise regulation of drying control parameters (wind temperature, wind pressure) is essential to ensure the surface quality of porous bricks. However, the relationship between these control parameters and the surface quality of porous bricks is often complex and nonlinear. In order to deeply understand the role and influence of these parameters on the surface quality of porous bricks, it is necessary to calculate the responsiveness estimation of the dynamic change characteristic vector of the porous brick surface quality relative to the time series correlation characteristic vector of the porous brick drying control parameters to obtain the porous brick drying quality response characteristic matrix. Based on the calculated response characteristic matrix, the specific impact of wind temperature changes on the surface quality of porous bricks can be revealed, thereby guiding the setting and adjustment of wind temperature. For example, if it is found that excessively high wind temperature will cause cracks on the surface of porous bricks, this problem can be avoided by lowering the wind temperature. Through precise wind temperature regulation, the scrap rate and defective product rate in the production process can be reduced, and the qualified rate and production efficiency of porous bricks can be improved.

[0039] Figure 5 FIG. 1 is a block diagram of a wind temperature control guidance result generation module in a production process data acquisition and analysis system according to an embodiment of the present application. Figure 5 As shown, the wind temperature control guidance result generation module 130 includes: a response characteristic matrix expansion unit 131, used to expand the porous brick drying quality response characteristic matrix into a porous brick drying quality response characteristic vector; a response characteristic vector optimization unit 132, used to perform local disturbance adaptive optimization based on shift compensation on the porous brick drying quality response characteristic vector to obtain an optimized porous brick drying quality response characteristic vector; an optimized characteristic vector parsing unit 133, used to pass the optimized porous brick drying quality response characteristic vector through a classifier-based wind temperature control guidance result generator to obtain a classification result, and the classification result is used to indicate whether the drying chamber wind temperature at the current time point should increase, decrease or remain unchanged.

[0040] Taking into account that the porous brick drying quality response characteristic matrix contains row and column dimensions, in the data processing process, matrix operations are often more complex and time-consuming than vector operations. In order to reduce the amount of calculation and improve the speed and efficiency of porous brick drying quality response characteristic data processing, the technical solution of the present application needs to expand the porous brick drying quality response characteristic matrix into a porous brick drying quality response characteristic vector. By expanding the characteristic matrix in the row direction or the column direction, the key features in the original matrix data can be retained, that is, it will not affect the results of subsequent wind temperature control.

[0041] In particular, considering that in the actual brick production environment, the sensor will be affected by various noises such as electromagnetic interference, these noises will lead to the inaccuracy of the collected wind temperature and wind pressure data, and the accuracy of the sensor is limited. Even a high-precision sensor may cause disturbances to the input data due to its inherent measurement error. And in the process of porous brick surface image acquisition and preprocessing, factors such as lighting conditions, camera stability, image compression, etc. will cause distortion of image data, resulting in local feature disturbances in the porous brick dry quality response feature vector finally encoded. In order to reduce the prediction error of the classifier caused by local input disturbances, so that in actual production, the classifier can also better adapt to the frequent changes in environmental conditions and operating conditions, thereby improving the control accuracy of the drying process and product quality, therefore, in the technical solution of the present application, it is necessary to perform local disturbance adaptive optimization based on shift compensation on the porous brick dry quality response feature vector to obtain an optimized porous brick dry quality response feature vector.

[0042] Specifically, in an embodiment of the present application, the porous brick dry quality response feature vector is subjected to local perturbation adaptive optimization based on shift compensation to obtain an optimized porous brick dry quality response feature vector, including: determining the first weight matrix and the second weight matrix of the classifier before and after each iterative update; matrix multiplying the first weight matrix and the second weight matrix with the porous brick dry quality response feature vector respectively to obtain a first porous brick dry quality response modulation feature vector and a second porous brick dry quality response modulation feature vector; calculating the position difference between the first porous brick dry quality response modulation feature vector and the second porous brick dry quality response modulation feature vector to obtain a porous brick dry quality response shift information representation vector; calculating the F norm of the position mean vector between the first porous brick dry quality response modulation feature vector and the second porous brick dry quality response modulation feature vector as a shift compensation scaling factor; linearly scaling the porous brick dry quality response shift information representation vector with the shift compensation scaling factor to obtain a scaled porous brick dry quality response shift information representation vector, and inputting the scaled porous brick dry quality response shift information representation vector into Si gmoid activation function is used to obtain the backward propagation shift compensation representation vector of the porous brick dry quality response; the point-by-point multiplication of the backward propagation shift compensation representation vector of the porous brick dry quality response and the porous brick dry quality response feature vector is calculated to obtain the optimized porous brick dry quality response feature vector.

[0043] In an embodiment of the present application, another possible implementation method of performing local disturbance adaptive optimization based on shift compensation on the porous brick dry quality response characteristic vector to obtain an optimized porous brick dry quality response characteristic vector may be: processing the porous brick dry quality response characteristic vector according to the following formula to obtain the optimized porous brick dry quality response characteristic vector; wherein the formula is:

[0044]

[0045]

[0046] Among them, M 1 is the first weight matrix, M 2 is the second weight matrix, V is the porous brick drying quality response characteristic vector, represents matrix multiplication, Indicates adding by position point, ‖·‖ F represents the F norm of the feature vector, S represents the shift compensation scaling factor, ⊙ represents the point multiplication by position, represents point subtraction, si gmoi d represents activation function, and V' represents the optimized porous brick drying quality response feature vector.

[0047] In order to enhance the robustness of the classifier to local disturbances in the input feature vector, the porous brick dry quality response feature vector is subjected to displacement compensation based on back propagation representation. The key to the displacement compensation based on back propagation representation of the porous brick dry quality response feature vector is to capture the information of the weight matrix changing with the training process, and to use this information to generate the porous brick dry quality response displacement information representation vector, and then to generate the porous brick dry quality response back propagation displacement compensation representation vector by scaling and nonlinear transformation (Sigmoid function). In this way, the porous brick dry quality response feature vector is supported by the distribution difference of different classification scenarios in the training process of the classifier to describe the local displacement of the feature vector, so that not only the displacement of the feature vector caused by the change of the weight matrix can be quantified, but also a compensation vector can be generated according to the displacement information. The role of the compensation vector is to weight and adjust each element in the feature vector according to the degree of influence of the disturbance, thereby reducing the influence of the disturbance on the classification result. In this way, the robustness of the classifier to local disturbances in the input data can be significantly improved, the stability and generalization ability of the classifier can be enhanced, and by reducing the impact of local disturbances on the classification results, the classifier can maintain a high classification accuracy when facing noisy or variant data.

[0048] Finally, the optimized porous brick drying quality response feature vector is passed through the classifier-based wind temperature control guidance result generator to obtain a classification result, and the classification result is used to indicate whether the drying chamber wind temperature at the current time point should increase, decrease, or remain unchanged. It should be understood that the classifier-based wind temperature control guidance result generator is a machine learning classification algorithm, which is used to classify according to the information in the optimized porous brick drying quality response feature vector. It is a mapping relationship learned through a large amount of training data, and can more accurately predict the appropriate control direction of the wind temperature under the current porous brick drying quality state. Compared with the traditional empirical control or fixed rule control method, it has higher accuracy and reliability. The classification result directly guides the automatic control system to adjust the wind temperature parameters of the drying chamber to ensure that the porous bricks obtain the best quality effect during the drying process. In this way, intelligent production is realized and the drying efficiency and product quality of bricks are improved.

[0049] In summary, the production process data acquisition and analysis system 100 based on the embodiment of the present application is explained, which extracts the time series features of the wind temperature and wind pressure data of the drying chamber at multiple time points to obtain the time series features of the parameters related to the drying control of porous bricks, and extracts the features of the surface images of the porous bricks at multiple time points to obtain the dynamic change characteristics of the drying quality of the porous bricks, and then establishes the response characteristics of the porous brick drying quality and the parameters related to the drying control in the same dimensional space, and finally passes the response characteristics through the classifier-based wind temperature guidance control result generator to obtain the specific wind temperature control direction. In this way, the precise control of the wind temperature parameters in the porous brick drying process can be achieved, thereby improving the production quality and efficiency of the bricks.

[0050] As described above, the production process data acquisition and analysis system 100 according to the embodiment of the present application can be implemented in various terminal devices, such as a server for production process data acquisition and analysis. In one example, the production process data acquisition and analysis system 100 according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the production process data acquisition and analysis system 100 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the production process data acquisition and analysis system 100 can also be one of the many hardware modules of the terminal device.

[0051] Alternatively, in another example, the production process data acquisition and analysis system 100 and the terminal device may also be separate devices, and the production process data acquisition and analysis system 100 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0052] Figure 6Flow chart of the production process data collection and analysis method according to an embodiment of the present application. Figure 6 As shown, in the production process data collection and analysis method, it includes: S110, obtaining the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at multiple predetermined time points within a predetermined time period; S120, performing response association coding on the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at the multiple predetermined time points to obtain a porous brick drying quality response feature matrix; S130, based on the porous brick drying quality response feature matrix, obtaining the wind temperature control guidance result at the current time point.

[0053] Here, those skilled in the art will appreciate that the specific operations of each step in the above production process data collection and analysis method have been described in detail above. Figures 1 to 5 The description of the production process data acquisition and analysis system has been introduced in detail, and therefore, its repeated description will be omitted.

[0054] In summary, the production process data collection and analysis method based on the embodiment of the present application is explained, which extracts the time series features of the wind temperature and wind pressure data of the drying chamber at multiple time points to obtain the time series features of the parameters related to the drying control of porous bricks, and extracts the features of the surface images of the porous bricks at multiple time points to obtain the dynamic change characteristics of the drying quality of the porous bricks, and then establishes the response characteristics of the porous brick drying quality and the parameters related to the drying control in the same dimensional space, and finally passes the response characteristics through the classifier-based wind temperature guidance control result generator to obtain the specific wind temperature control direction. In this way, the precise control of the wind temperature parameters in the porous brick drying process can be achieved, thereby improving the production quality and efficiency of bricks.

Claims

1. A production process data acquisition and analysis system, characterized in that: include: A porous brick drying related data acquisition module is used to obtain the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at multiple predetermined time points within a predetermined time period; A porous brick drying related data encoding module is used to perform response association encoding on the drying chamber wind temperature values, drying chamber wind pressure values ​​and porous brick surface images at the plurality of predetermined time points to obtain a porous brick drying quality response feature matrix; The wind temperature control guidance result generation module is used to obtain the wind temperature control guidance result at the current time point based on the porous brick drying quality response characteristic matrix.

2. The production process data acquisition and analysis system according to claim 1, characterized in that: The porous brick drying related data encoding module includes: A porous brick drying control parameter time series feature extraction unit is used to extract correlation features of the drying chamber wind temperature values ​​and the drying chamber wind pressure values ​​at the plurality of predetermined time points to obtain a porous brick drying control parameter time series correlation feature vector; A porous brick surface quality feature extraction unit is used to extract surface features of the porous brick surface images at the plurality of predetermined time points to obtain a porous brick surface quality dynamic change feature vector; The responsiveness estimation unit is used to calculate the responsiveness estimation of the porous brick surface quality dynamic change characteristic vector relative to the porous brick drying control parameter time series correlation characteristic vector to obtain the porous brick drying quality response characteristic matrix.

3. The production process data acquisition and analysis system according to claim 2, characterized in that: The porous brick drying control parameter time series feature extraction unit comprises: The porous brick drying control parameter data structuring subunit is used to arrange the drying chamber wind temperature values ​​and the drying chamber wind pressure values ​​at the plurality of predetermined time points according to the time dimension to obtain a drying chamber wind temperature time series input vector and a drying chamber wind pressure time series input vector; A porous brick drying control parameter association subunit is used to associate the drying chamber wind temperature time series input vector with the drying chamber wind pressure time series input vector to obtain a porous brick drying control parameter time series association matrix; The porous brick drying control parameter timing encoding subunit is used to pass the porous brick drying control parameter timing association matrix through the porous brick drying control parameter timing encoder to obtain the porous brick drying control parameter timing association feature vector.

4. The production process data acquisition and analysis system according to claim 3, characterized in that: The porous brick surface quality feature extraction unit comprises: A porous brick surface image preprocessing subunit, used for performing image preprocessing on the porous brick surface images at the plurality of predetermined time points to obtain a plurality of porous brick surface enhanced images; A porous brick surface quality dynamic change feature generation subunit is used to pass the plurality of porous brick surface enhancement images through a porous brick surface quality dynamic change capturer to obtain a porous brick surface quality dynamic change feature map; The porous brick surface quality feature map pooling subunit is used to perform mean pooling on each feature matrix of the porous brick surface quality dynamic change feature map along the channel dimension to obtain the porous brick surface quality dynamic change feature vector.

5. The production process data acquisition and analysis system according to claim 4, characterized in that: The porous brick drying control parameter timing encoder is a convolutional neural network model in which adjacent layers use mutually transposed convolution kernels, and the porous brick surface quality change dynamic capturer is a convolutional neural network model using a three-dimensional convolution kernel.

6. The production process data acquisition and analysis system according to claim 5, characterized in that: The air temperature control guidance result generation module includes: A response characteristic matrix expansion unit, used for expanding the porous brick drying quality response characteristic matrix into a porous brick drying quality response characteristic vector; A response characteristic vector optimization unit, used for performing local disturbance adaptive optimization based on displacement compensation on the porous brick drying quality response characteristic vector to obtain an optimized porous brick drying quality response characteristic vector; The optimized feature vector parsing unit is used to pass the optimized porous brick drying quality response feature vector through the classifier-based wind temperature control guidance result generator to obtain a classification result, and the classification result is used to indicate whether the drying chamber wind temperature at the current time point should increase, decrease or remain unchanged.

7. The production process data acquisition and analysis system according to claim 6, characterized in that: The response feature vector optimization unit is used for: Determining a first weight matrix and a second weight matrix of the classifier before and after each iterative update; Performing matrix multiplication of the first weight matrix and the second weight matrix with the porous brick dry mass response characteristic vector respectively to obtain a first porous brick dry mass response modulation characteristic vector and a second porous brick dry mass response modulation characteristic vector; Calculating the position difference between the first porous brick dry mass response modulation feature vector and the second porous brick dry mass response modulation feature vector to obtain a porous brick dry mass response shift information representation vector; Calculate the F norm of the position mean vector between the first porous brick dry mass response modulation feature vector and the second porous brick dry mass response modulation feature vector as a shift compensation scaling factor; Linearly scaling the porous brick dry mass response shift information representation vector with the shift compensation scaling factor to obtain a scaled porous brick dry mass response shift information representation vector, and inputting the scaled porous brick dry mass response shift information representation vector into a Sigmoid activation function to obtain a porous brick dry mass response back-propagation shift compensation representation vector; The optimized porous brick drying mass response characteristic vector is obtained by calculating the point-by-point multiplication of the backward propagation shift compensation representation vector of the porous brick drying mass response and the characteristic vector of the porous brick drying mass response.

8. A method for collecting and analyzing production process data, characterized in that: include: Acquire the drying chamber wind temperature value, the drying chamber wind pressure value and the porous brick surface images at a plurality of predetermined time points within a predetermined time period; Performing response association coding on the drying chamber wind temperature values, the drying chamber wind pressure values ​​and the porous brick surface images at the plurality of predetermined time points to obtain a porous brick drying quality response feature matrix; Based on the porous brick drying quality response characteristic matrix, the wind temperature control guidance result at the current time point is obtained.

9. The production process data collection and analysis method according to claim 8, characterized in that: The drying chamber wind temperature values, the drying chamber wind pressure values ​​and the porous brick surface images at the plurality of predetermined time points are subjected to response association coding to obtain a porous brick drying quality response feature matrix, including: Extracting correlation features of the drying chamber wind temperature values ​​and the drying chamber wind pressure values ​​at the plurality of predetermined time points to obtain a time series correlation feature vector of the porous brick drying control parameters; Extracting surface features of the porous brick surface images at the plurality of predetermined time points to obtain a dynamic change feature vector of the porous brick surface quality; The responsiveness estimation of the porous brick surface quality dynamic change characteristic vector relative to the porous brick drying control parameter time series correlation characteristic vector is calculated to obtain the porous brick drying quality response characteristic matrix.

10. The production process data collection and analysis method according to claim 9, characterized in that: Based on the porous brick drying quality response characteristic matrix, the wind temperature control guidance result at the current time point is obtained, including: Expanding the porous brick drying quality response characteristic matrix into a porous brick drying quality response characteristic vector; Performing local disturbance adaptive optimization based on displacement compensation on the porous brick drying quality response characteristic vector to obtain an optimized porous brick drying quality response characteristic vector; The optimized porous brick drying quality response feature vector is passed through a classifier-based wind temperature control guidance result generator to obtain a classification result, and the classification result is used to indicate whether the drying chamber wind temperature at the current time point should increase, decrease or remain unchanged.

Images