State detection method and device for aerated concrete brick production line

Abstract

This application discloses a method and device for detecting the status of an aerated concrete brick production line, which is applied to the field of aerated concrete brick production technology. The method includes: obtaining an electricity consumption dataset, an equipment dataset, and a brick dataset for the aerated concrete brick production line; performing feature extraction on the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain an electricity consumption feature sequence, an equipment operation feature sequence, and a production quality feature sequence for the aerated concrete brick production line; performing image processing based on the electricity consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a status image of the aerated concrete brick production line; and detecting the status of the aerated concrete brick production line based on the status image to obtain a status detection result for the aerated concrete brick production line. This application can effectively improve the accuracy of status detection of an aerated concrete brick production line.

Description

Technical Field

[0001] The present application relates to the technical field of aerated concrete brick production, and in particular to a method and device for detecting the status of an aerated concrete brick production line. Background Art

[0002] Aerated concrete (ACC) bricks are a type of aerated concrete product produced using a high-temperature autoclaving process. The production process begins with a crushing and grinding process, where raw materials such as quicklime and gypsum are crushed using crushing equipment, or raw materials such as fly ash are milled using a ball mill. These two processes are collectively referred to as "crushing and grinding equipment." Next, the milled raw materials are fed into a mixing process according to a preset proportion, where they are mixed and then poured into a slurry. It is important to note that the slurry temperature must meet process requirements (approximately 45°C) before pouring. If this does not meet the requirements, steam heating can be used. Furthermore, an aluminum powder suspension should be added 0.5-1 minute before pouring. Finally, the molding process begins. The mold after pouring is sent to the initial curing room for aeration and initial setting. After the initial curing is completed, the green body is sent to the cutting equipment. The cutting equipment performs operations such as cross-cutting, longitudinal cutting, and breading according to actual production needs to obtain the cut green body. The cut green body is placed on the autoclave, and the autoclave is sent to the autoclave for curing. After the curing process, the finished brick product is obtained. Of course, in some cases, the aeration and initial setting process can be omitted.

[0003] For aerated concrete brick production, regular evaluation of the production line's status is crucial. This evaluation allows for timely identification of bottlenecks and issues within the line, enabling optimization and adjustments to streamline the production process and improve overall production efficiency. Related technologies typically use deep learning models combined with information such as brick weight to detect the status of the line. However, these technologies often only consider a single dimension of data and ignore other key factors, resulting in reduced accuracy in detecting the status of the line. Summary of the Invention

[0004] The embodiments of the present application provide a method and device for detecting the status of an aerated concrete brick production line, which are used to improve the accuracy of status detection of the aerated concrete brick production line.

[0005] In one aspect, an embodiment of the present application provides a method for detecting the status of an aerated concrete brick production line, comprising the following steps:

[0006] Obtaining an electricity consumption dataset, an equipment dataset, and a brick dataset for the aerated concrete brick production line;

[0007] Performing feature extraction on the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain an electricity consumption feature sequence, an equipment operation feature sequence, and a production quality feature sequence of the aerated concrete brick production line;

[0008] Performing image processing based on the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a state image of the aerated concrete brick production line;

[0009] The state of the aerated concrete brick production line is detected according to the state image to obtain a state detection result of the aerated concrete brick production line.

[0010] On the other hand, an embodiment of the present application provides a state detection device for an aerated concrete brick production line, comprising:

[0011] An acquisition module, configured to acquire an electricity consumption dataset, an equipment dataset, and a brick dataset of the aerated concrete brick production line;

[0012] a feature extraction module, configured to perform feature extraction on the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain an electricity consumption feature sequence, an equipment operation feature sequence, and a production quality feature sequence for the aerated concrete brick production line;

[0013] an image processing module, configured to perform image processing based on the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a state image of the aerated concrete brick production line;

[0014] A detection module is used to detect the state of the aerated concrete brick production line according to the state image to obtain a state detection result of the aerated concrete brick production line.

[0015] According to a state detection method and device for an aerated concrete brick production line of the present application, taking into account the correlation between the power consumption, equipment operation and brick quality of the aerated concrete brick production line and the actual state of the aerated concrete brick production line, feature extraction is performed on multi-dimensional data such as power consumption data, equipment data, brick data, etc., to effectively capture the key time series features associated with the actual state of the aerated concrete brick production line, and then image generation is performed based on the time series features of the three modes. The image generation can reconstruct the time series feature sequence associated with the actual state of the aerated concrete brick production line while retaining the time dependency of the data, thereby converting the one-dimensional time series feature sequence into a state image for depicting the actual state of the aerated concrete brick production line. In this way, the feature information hidden in the data of each dimension can be better mined, and more real state expressions can be captured. Finally, the state of the aerated concrete brick production line is detected based on the state image, thereby giving full play to the advantages of the deep learning model in image processing tasks and effectively improving the state detection accuracy of the aerated concrete brick production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of a method for detecting the status of an aerated concrete brick production line provided by the present application;

[0017] Figure 2 This is a schematic diagram of a state detection method for an aerated concrete brick production line provided by this application;

[0018] Figure 3 This is a flow chart of electricity consumption dimension feature extraction provided by this application;

[0019] Figure 4 This is a flow chart of device operation dimension feature extraction provided by this application;

[0020] Figure 5 This is a flowchart of the production quality dimension feature extraction provided by this application;

[0021] Figure 6 It is a flow chart of the feature interaction and visualization provided by this application;

[0022] Figure 7 This is a structural diagram of a status detection device for an aerated concrete brick production line provided in this application. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0024] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0026] In response to the problems and defects in the related technologies, the embodiments of the present application provide a method and device for detecting the status of an aerated concrete brick production line, aiming to improve the accuracy of status detection of the aerated concrete brick production line.

[0027] A method for detecting the status of an aerated concrete brick production line provided by an embodiment of the present application will be described in detail below with reference to the accompanying drawings.

[0028] The embodiment of the present application provides a method for detecting the status of an aerated concrete brick production line, which can be applied to a terminal, a server, or software running on a terminal or a server. The terminal can be a tablet computer, a laptop computer, a desktop computer, etc., but is not limited thereto. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. In addition, the server can also be a node server in a blockchain network, but is not limited thereto. Blockchain is a new application model for computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms.

[0029] Reference Figure 1 , Figure 1 4 is a flow chart of a method for detecting the status of an aerated concrete brick production line provided in an embodiment of the present application. The method for detecting the status of an aerated concrete brick production line may include the following steps S101-S104.

[0030] S101, obtaining an electricity consumption dataset, an equipment dataset, and a brick dataset of an aerated concrete brick production line.

[0031] It should be noted that the electricity consumption dataset is associated with the electricity consumption of the aerated concrete brick production line and may include electricity impact data and process electricity consumption data at several preset moments. The electricity impact data is used to indicate environmental data that affects the power load of the aerated concrete brick production line. The electricity impact data may include cloud cover information, air pressure information, and relative humidity information for the environment in which the aerated concrete brick production line is located. The process electricity consumption data is used to indicate the electricity consumption data of various processes in the aerated concrete brick production line. The process electricity consumption data may include electricity consumption data for the crushing, mixing, and molding processes in the aerated concrete brick production line. The process electricity consumption data refers to the electricity consumption data of the equipment used to execute the process, such as power consumption. Furthermore, the equipment dataset is associated with the equipment operation status of the aerated concrete brick production line and may include equipment operation data at several preset moments. The equipment operation data may include equipment operation data for the crushing, mixing, and molding processes in the aerated concrete brick production line. The process equipment operation data refers to the operation data of the equipment used to execute the process, such as work efficiency. In addition, the brick data set is associated with the production quality of the aerated concrete brick production line, which may include image data, weight information and thickness information of at least one brick at several preset moments. The number of bricks can be flexibly set according to actual conditions. For example, the number of bricks can be 100, but is not limited to this.

[0032] It can be understood that the power usage dataset, the device dataset, and the brick dataset may be datasets that have been detected and stored in advance.

[0033] In this step, the status of the aerated concrete brick production line can often be measured from two aspects: production frequency and production quality. The higher the production frequency and the better the production quality of the aerated concrete brick production line, the better the status of the aerated concrete brick production line. Regarding production frequency, based on prior knowledge, the more electricity the aerated concrete brick production line uses and the more frequently the equipment of the aerated concrete brick production line operates, the higher the production frequency of the aerated concrete brick production line. Regarding production quality, based on prior knowledge, the better the quality of the bricks of the aerated concrete brick production line, the better the production quality of the aerated concrete brick production line. Therefore, it can be seen that the electricity consumption, equipment operation, and brick quality of the aerated concrete brick production line are closely related to the actual status of the aerated concrete brick production line. Therefore, this step obtains the three-dimensional data sets of the aerated concrete brick production line: the electricity consumption data set, the equipment data set, and the brick data set as the data source for status detection.

[0034] Optionally, the preset moments can be flexibly set based on actual circumstances, and this embodiment of the present application does not limit this. For example, if the status of an aerated concrete brick production line is being monitored over the past hour, the past hour can be evenly divided into sixty preset moments, with each preset moment representing one minute. For another example, if the status of an aerated concrete brick production line is being monitored over the previous day, the previous day can be evenly divided into twenty-four preset moments, with each preset moment representing one hour.

[0035] S102 , feature extraction is performed on the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain an electricity consumption feature sequence, an equipment operation feature sequence, and a production quality feature sequence of the aerated concrete brick production line.

[0036] It should be noted that the power consumption feature sequence is a time feature sequence associated with the power consumption of the aerated concrete brick production line, which may include target power consumption features at several preset moments. Furthermore, the equipment operation feature sequence is a time feature sequence associated with the equipment operation of the aerated concrete brick production line, which may include target equipment operation features at several preset moments. In addition, the production quality feature sequence is a time feature sequence used to indicate the production quality of the aerated concrete brick production line. The production quality can be measured by the cracks, shape, and weight of the finished bricks. Therefore, the production quality feature sequence may include a first quality feature sequence, a second quality feature sequence, and a third quality feature sequence. The first quality feature sequence is a time feature sequence associated with the cracks of the bricks produced by the aerated concrete brick production line, and the first quality feature sequence may include target crack features at several preset moments; the second quality feature sequence is a time feature sequence associated with the shape of the bricks produced by the aerated concrete brick production line, and the second quality feature sequence may include target shape features at several preset moments; the third quality feature sequence is a time feature sequence associated with the weight of the bricks produced by the aerated concrete brick production line, and the third quality feature sequence may include target weight features at several preset moments.

[0037] In this step, after obtaining the data source used for status detection, the electricity consumption feature sequence of the aerated concrete brick production line is obtained by feature extraction of the electricity consumption dataset, the equipment operation feature sequence of the aerated concrete brick production line is obtained by feature extraction of the equipment dataset, and the production quality feature sequence of the aerated concrete brick production line is obtained by feature extraction of the brick dataset. In this way, the feature information of the electricity consumption dimension, the equipment operation dimension and the production quality dimension can be effectively captured, thereby realizing feature extraction.

[0038] S103 , performing image processing based on the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a state image of the aerated concrete brick production line.

[0039] In this step, images are generated based on the feature sequences of the electricity consumption dimension, the equipment operation dimension, and the production quality dimension to obtain a portrait indicating the actual status of the aerated concrete brick production line, namely, a status image, thereby upgrading the one-dimensional time series feature sequence to a two-dimensional image.

[0040] S104: Detect the status of the aerated concrete brick production line according to the status image to obtain a status detection result of the aerated concrete brick production line.

[0041] In this step, the state image is input into a pre-trained state detection model, and state detection is performed through the state detection model to obtain the state detection result of the aerated concrete brick production line, thereby giving full play to the advantages of the deep learning model in image processing and realizing the state detection of the aerated concrete brick production line.

[0042] Optionally, the state detection model is a deep learning model, and its type can be flexibly set according to actual conditions. For example, the state detection model can be a YOLOv8 model or an Adaboost model, but is not limited thereto. In addition, the data set used in training the state detection model may include multiple preset state sample images and label information corresponding to each state sample image, wherein the generation of the state sample image is the same as that of the above steps S101-S103, that is, the state sample image is generated based on multiple data in the electricity data sample set, the equipment data sample set, and the brick data sample set, and the label information corresponding to the state sample image is the same as the state detection result of the aerated concrete brick production line, which will not be repeated in the embodiments of the present application.

[0043] Optionally, the status detection result of the aerated concrete brick production line depends on the type of status detection task in the embodiment of the present application. For example, the status detection task in the embodiment of the present application may be a binary classification task, that is, the status detection model is used to detect whether the status of the aerated concrete brick production line is qualified, and the status detection result of the aerated concrete brick production line may include any one of qualified or unqualified. For another example, the status detection task in the embodiment of the present application may be a multi-classification task, that is, the status detection model is used to detect the type of status of the aerated concrete brick production line, and the status detection result of the aerated concrete brick production line may include any one of excellent, good, qualified or poor, but is not limited to this.

[0044] It can be seen that the embodiment of the present application takes into account the correlation between the power consumption, equipment operation and brick quality of the aerated concrete brick production line and the actual state of the aerated concrete brick production line. By performing feature extraction on multi-dimensional data such as power consumption data, equipment data, brick data, etc., the key time series features associated with the actual state of the aerated concrete brick production line are effectively captured, and then image generation is performed based on the time series features of the three modes. The image generation can reconstruct the time series feature sequence associated with the actual state of the aerated concrete brick production line while retaining the time dependency of the data, thereby converting the one-dimensional time series feature sequence into a state image for depicting the actual state of the aerated concrete brick production line. In this way, the feature information hidden in the data of each dimension can be better mined, and more real state expressions can be captured. Finally, the state of the aerated concrete brick production line is detected based on the state image, thereby giving full play to the advantages of the deep learning model in image processing and solving the limitations brought by single data prediction, which can effectively improve the state detection accuracy of the aerated concrete brick production line.

[0045] In some embodiments, reference Figure 2 and Figure 3 Regarding feature extraction for electricity consumption, this embodiment extracts feature information related to the electricity consumption of the aerated concrete brick production line from two dimensions: electricity consumption influencing factors and internal electricity consumption factors. This feature information is then integrated into a corresponding time feature sequence. The specific implementation of step S102 may include, but is not limited to, steps S201-S203.

[0046] S201 : Obtaining electricity consumption impact characteristics at each preset moment based on electricity consumption impact data at each preset moment.

[0047] It should be noted that the electricity consumption of the aerated concrete brick production line can be further divided into electricity influencing factors and internal electricity factors. The electricity influencing factors are used to indicate the degree of influence of the environment in which the aerated concrete brick production line is located on the electricity consumption of the aerated concrete brick production line. The electricity influencing factors are characterized by the electricity influence characteristics.

[0048] In this step, under normal circumstances, environmental factors will indirectly affect the power consumption of the aerated concrete brick production line, thereby affecting the status detection of the aerated concrete brick production line. For example, in hot weather, additional air conditioning equipment is often required to maintain a comfortable working environment, while in rainy weather, additional drainage equipment is often required to drain water. The operation of these devices will increase the power consumption of the aerated concrete brick production line. However, these devices are not related to the production of the aerated concrete brick production line. Therefore, it is possible that the power consumption of the aerated concrete brick production line is high but the state of the aerated concrete brick production line is poor, thereby affecting the status detection of the aerated concrete brick production line. For example, the relative humidity is high during the rainy season. In this environment, the corrosion and aging process of the aerated concrete brick production line equipment will be accelerated, further increasing the power consumption of the aerated concrete brick production line. Although this additional power consumption is related to the production of the aerated concrete brick production line, this additional power consumption does not play a substantial role in the production process of the aerated concrete brick production line. Therefore, it is still possible that the power consumption of the aerated concrete brick production line is high but the state of the aerated concrete brick production line is poor, thereby affecting the status detection of the aerated concrete brick production line.

[0049] To improve the accuracy of status detection for aerated concrete brick production lines, this step requires considering factors influencing electricity consumption when extracting features in the electricity consumption dimension. Specifically, based on prior knowledge, cloud cover information, air pressure information, and relative humidity information can collectively measure actual weather conditions, and the actual weather conditions can be used to measure the environment in which the aerated concrete brick production line is located. Therefore, as described in the aforementioned embodiment, electricity consumption impact data can include cloud cover information, air pressure information, and relative humidity information for the environment in which the aerated concrete brick production line is located. Cloud cover information refers to the area of ​​clouds and the percentage of the sky occupied by clouds. When the cloud cover is between 0% and 20%, the weather is likely to be sunny. When the cloud cover is between 20% and 50%, the weather is likely to be lightly cloudy. When the cloud cover is between 50% and 80%, the weather is likely to be cloudy. When the cloud cover is between 80% and 100%, the weather is likely to be overcast. In the above cases, if the air pressure drops, there is a possibility of rain. For sunny days and few clouds, the probability of rain is small and generally light rain; for cloudy and overcast days, the probability of rain is large and generally moderate rain or heavy rain. In addition, on rainy days outside the plum rain season or during the plum rain season, the relative humidity will rise. Based on this, this step constructs a comprehensive weather model, which can reflect the impact of cloud cover, air pressure and relative humidity on the electricity consumption of the aerated concrete brick production line. For each preset moment, this step inputs the cloud cover information, air pressure information and relative humidity information of the current preset moment into the comprehensive weather model. The electricity consumption impact characteristics of the current preset moment can be obtained by indexing the comprehensive weather model, as shown in the following formula (1):

[0050] (1);

[0051] In formula (1), Indicates the characteristics of electricity consumption impact; Indicates air pressure information; Indicates cloud cover information; Indicates relative humidity information; 、 、 and Respectively represent the influence coefficients of atmospheric pressure information on the electricity consumption of aerated concrete brick production line under sunny, partly cloudy, cloudy and overcast days; 、 、 and Respectively represent the influence coefficients of relative humidity information on the electricity consumption of aerated concrete brick production line under sunny, partly cloudy, cloudy and overcast days; 、 、 and Respectively represent the influence coefficients of cloud cover information on the electricity consumption of aerated concrete brick production line under sunny, partly cloudy, cloudy and overcast days.

[0052] Optionally, cloud cover information, air pressure information, and relative humidity information can all be obtained in real time by crawling local meteorological data websites. In addition, the specific values ​​of the influence coefficients of various types of information on the electricity consumption of the aerated concrete brick production line can be pre-set according to actual conditions, and this embodiment is not limited here. Among them, few clouds and cloudy days have a lower impact on the electricity consumption of the aerated concrete brick production line, while sunny and cloudy days have a greater impact on the electricity consumption of the aerated concrete brick production line. Therefore, the influence coefficient under few clouds and cloudy days can be smaller than the influence coefficient under sunny and cloudy days. For example, the value range of the influence coefficient under few clouds and cloudy days can be 0.1~0.5, while the value range of the influence coefficient under sunny and cloudy days can be 0.05~0.1, but it is not limited to this.

[0053] S202 , obtaining the power consumption characteristics of the process at each preset moment based on the power consumption data of the process at each preset moment.

[0054] It should be noted that, according to the above steps, the electricity consumption of the aerated concrete brick production line can be further subdivided into electricity influencing factors and internal electricity consumption factors. The internal electricity consumption factors are used to indicate the electricity consumption of the equipment of the aerated concrete brick production line. The internal electricity consumption factors are characterized by the electricity consumption characteristics of the process. The electricity consumption characteristics of the process can include the electricity consumption characteristics of the crushing and grinding process, the electricity consumption characteristics of the mixing process, and the electricity consumption characteristics of the forming process.

[0055] In this step, under normal circumstances, the electricity consumption of various processes in the aerated concrete brick production line is measured together. According to the production process of aerated concrete bricks, the aerated concrete brick production line can include three processes: grinding process, mixing process and forming process. Therefore, at each preset time, the following operations are performed:

[0056] The crushing and grinding process involves equipment including crushers and ball mills. Crusher processing is used to process large-particle raw materials such as quicklime and gypsum, while ball mill processing is used to process small-particle raw materials such as fly ash. When the equipment is not aged, its crushing and grinding capacity has not degraded, and it is fully capable of handling the maximum weight of raw materials it can process. However, as the equipment ages, its crushing and grinding capacity degrades, typically manifesting as a decrease in the maximum weight of raw materials it can handle. Therefore, equipment operating time directly affects its maximum weight of raw materials it can handle (crushing capacity). Due to this degraded crushing capacity, the equipment often takes longer to process the same weight of raw materials, resulting in an increase in electricity consumption. However, as previously shown, the higher the equipment's electricity consumption, the better the production line's performance. This clearly contradicts the aforementioned observation that equipment aging indicates that electricity consumption at this time does not truly reflect the health of the equipment or production line. Therefore, the equipment's maximum weight of raw materials it can handle (crushing capacity) directly affects its electricity consumption, leading to distorted (overstated) electricity consumption.

[0057] Based on this, the electricity consumption data of the crushing and grinding process at the current preset moment can include the electricity consumption of the crusher, the electricity consumption influence coefficient of the crusher, the electricity consumption of the ball mill, and the electricity consumption influence coefficient of the ball mill at the current preset moment. Among them, the electricity consumption of the equipment (i.e., the crusher or ball mill) at the current preset moment refers to the difference between the cumulative electricity consumption of the equipment at the current preset moment and the cumulative electricity consumption of the equipment at the previous preset moment, and the electricity consumption influence coefficient of the equipment at the current preset moment is used to correct the electricity consumption of the equipment at the current preset moment. In this step, based on the electricity consumption data of the crushing and grinding process at the current preset moment, combined with the following formula (2), the electricity consumption characteristics of the crushing and grinding process at the current preset moment can be obtained:

[0058] (2);

[0059] In formula (2), Indicates the electricity consumption characteristics of the crushing and grinding process, and is used to measure the electricity consumption of the crushing and grinding process; Indicates the electricity consumption influence coefficient of the crusher; Indicates the power consumption of the crusher; Indicates the electricity consumption influence coefficient of the ball mill; Indicates the power consumption of the ball mill; and Must be greater than zero and less than or equal to one.

[0060] It is worth noting that regarding the power consumption coefficient of the ball mill and crusher, as can be seen from the above content, the operating time of the equipment will directly affect the maximum weight of raw materials that the equipment can process (crushing capacity), and the maximum weight of raw materials that the equipment can process (crushing capacity) will directly affect the power consumption of the equipment and cause the power consumption of the equipment to be distorted (overestimated). In order to correct the power consumption of the equipment, this step first calculates the maximum weight of raw materials that the equipment can process at the current preset time, which can be expressed as , Indicates the maximum weight of raw materials that can be processed by the equipment, which is pre-calibrated at the factory and is always greater than zero. Indicates the operating time of the device. The operating time refers to the difference between the year of the current preset time and the year the device was manufactured. , Represents the raw material decay parameter, which represents the annual reduction in the weight of processable raw materials. It is a preset value and is greater than zero. Then, based on the maximum processable raw material weight of the equipment at the current preset moment, combined with the preset mapping table, the power consumption impact coefficient of the equipment at the current preset moment is dynamically determined. Among them, the preset mapping table pre-stores multiple preset maximum processable raw material weights and the power consumption impact coefficient corresponding to each preset maximum processable raw material weight. The table is pre-established based on historical data. The dynamic power consumption impact coefficient can be determined by looking up the table, thereby correcting the power consumption of the equipment. It should be noted that the power consumption impact coefficient is positively correlated with the maximum processable raw material weight, that is, the smaller the maximum processable raw material weight, the smaller the power consumption impact coefficient.

[0061] For the mixing process, its equipment includes a mixer, and the raw material ratio will have a certain impact on the mixing process. Generally speaking, a suitable raw material ratio can speed up the mixing process, thereby reducing the power consumption of the mixer, while an inappropriate raw material ratio slows down the mixing process, thereby increasing the power consumption of the mixer. Therefore, the power consumption data of the mixing process at the current preset moment may include the power consumption of the mixer at the current preset moment and the influence coefficient of the raw material ratio on the power consumption of the mixer, and these data can be collected in advance. Among them, the power consumption here is the same as the above-mentioned power consumption, and will not be repeated here. This step is based on the power consumption data of the mixing process at the current preset moment, combined with the following formula (3), to obtain the power consumption characteristics of the mixing process at the current preset moment:

[0062] (3);

[0063] In formula (3), Indicates the electricity consumption characteristics of the mixing process and is used to measure the electricity consumption of the mixing process; Indicates the influence coefficient of raw material ratio on the power consumption of the mixer; Indicates the power usage of the blender. Optionally, The specific value of can be preset according to the actual situation, and this embodiment does not limit it here, but it should be noted that Must be greater than zero and less than or equal to one.

[0064] For the molding process, considering that gasification and initial condensation are not required in some cases, the equipment for the molding process only considers the cutting machine and the autoclave. In addition, since the molding process is relatively simple and there are no additional factors that have a significant impact on its process, the power consumption data of the molding process at the current preset time can include the power consumption of the cutting machine and the power consumption of the autoclave at the current preset time, and these data can be collected in advance. Among them, the power consumption here is the same as the power consumption mentioned above, and will not be repeated here. This step is based on the power consumption data of the molding process at the current preset time, combined with the following formula (4), to obtain the power consumption characteristics of the molding process at the current preset time:

[0065] (4);

[0066] In formula (4), Indicates the electricity consumption characteristics of the molding process and is used to measure the electricity consumption of the molding process; Indicates the power consumption weight of the cutting machine, Indicates the power consumption of the cutting machine; Indicates the power consumption weight of the autoclave, Indicates the power consumption of the autoclave. Optionally, and The specific value of can be preset according to the actual situation, and this embodiment does not limit it here, but it should be noted that the sum of these two power weights is one, and and Must be greater than zero and less than or equal to one.

[0067] S203 , obtaining target power consumption characteristics at a plurality of preset moments according to the power consumption impact characteristics and process power consumption characteristics at a plurality of preset moments, and further obtaining a power consumption characteristic sequence.

[0068] It should be noted that the target electricity consumption characteristics at the preset time are used to measure the electricity consumption of the aerated concrete brick production line at the preset time.

[0069] In this step, for each preset moment, the power consumption impact characteristics and process power consumption characteristics of the current preset moment are fused to obtain the target power consumption characteristics of the current preset moment; the power consumption characteristic sequence can be constructed through the target power consumption characteristics of several preset moments. , For the The target electricity consumption characteristics at a preset time, is the number of preset moments, thereby capturing the key timing characteristics of the electricity consumption dimension.

[0070] For example, at each preset moment, a machine learning method is used to process the electricity consumption impact characteristics and process electricity consumption characteristics at the current preset moment to obtain the target electricity consumption characteristics for the current preset moment. The machine learning method can be set based on actual conditions. For example, the machine learning method can be a support vector machine or a random forest, but is not limited to these.

[0071] For example, the process power consumption characteristics at a preset time are used to measure the power consumption of various processes in the aerated concrete brick production line at that preset time. For each preset time, the comprehensive power consumption characteristics at the current preset time are derived based on the proportion of each process's time in the total production time of the aerated concrete brick production line and the power consumption characteristics of each process at the current preset time. The comprehensive power consumption characteristics can measure the power consumption of the aerated concrete brick production line itself at the current preset time. However, because the power consumption of the aerated concrete brick production line is affected by external environmental factors, the power consumption represented by the comprehensive power consumption characteristics contains false power consumption caused by environmental factors and is not the actual power consumption.

[0072] According to the above steps, the power consumption impact feature is used to indicate the impact of the environment in which the aerated concrete brick production line is located on the power consumption of the aerated concrete brick production line. Therefore, after obtaining the comprehensive power consumption feature, this example first performs maximum-minimum normalization on the power consumption impact features of several preset moments so that the power consumption impact features of each preset moment are all between 0 and 1. Then, for each preset moment, the product of the normalized power consumption impact feature at the current preset moment and the comprehensive power consumption feature at the current preset moment is obtained as the power consumption feature generated by the environmental impact at the current preset moment. Then, the difference between the comprehensive power consumption feature at the current preset moment and the power consumption feature generated by the environmental impact at the current preset moment is obtained as the target power consumption feature at the current preset moment. The target power consumption feature can indicate the actual power consumption of the aerated concrete brick production line at the current preset moment. In this way, the influence of false power consumption on the feature extraction process of the power consumption dimension can be reduced, and the feature accuracy associated with the power consumption of the aerated concrete brick production line can be improved, thereby effectively improving the state detection accuracy of the aerated concrete brick production line. Among them, the target power consumption feature at the preset moment satisfies the following formula (5):

[0073] (5);

[0074] In formula (5), Indicates the target electricity consumption characteristics at a preset time; Indicates the normalized electricity consumption impact characteristics at the current preset time; Indicates the proportion of the grinding process in the production time of the aerated concrete brick production line; Indicates the proportion of the mixing process in the production time of the aerated concrete brick production line; Indicates the proportion of the time spent on the forming process in the production time of the aerated concrete brick production line. Optionally, the proportion of the time spent on each process in the production time of the aerated concrete brick production line can be set according to actual conditions and is not specifically limited to this.

[0075] In some embodiments, reference Figure 2 and Figure 4 Regarding feature extraction from the equipment operation dimension, this embodiment extracts feature information related to the equipment operation status of the aerated concrete brick production line on a process-by-process basis and integrates this feature information into a corresponding time feature sequence. The specific implementation of step S102 may also include, but is not limited to, the following steps S301-S302.

[0076] S301, obtaining the process equipment operation characteristics at each preset moment based on the equipment operation data at each preset moment.

[0077] It should be noted that the process equipment operation characteristics may include the equipment operation characteristics of the crushing and grinding process, the equipment operation characteristics of the stirring process and the equipment operation characteristics of the forming process.

[0078] In this step, the equipment operating status of each process is used to measure the equipment operation of the aerated concrete brick production line. For each process, if the equipment is in operation for a long time, it will experience a certain degree of wear and tear, which will affect the equipment's operation. Therefore, the equipment's operating time directly affects its operating status. In addition, the equipment's operating efficiency and operating status can also indicate the equipment's operating status to a certain extent. Therefore, for each preset moment, the following operations are performed:

[0079] For the crushing and grinding process, its equipment includes a crusher and a ball mill. The equipment operation data of the crushing and grinding process at the current preset time may include the working efficiency, operating time influence coefficient and crushing rate of the crusher at the current preset time, and the working efficiency, operating time influence coefficient and ball mill rate of the ball mill at the current preset time. In this step, the equipment operation characteristics of the crushing and grinding process at the current preset time can be obtained based on the equipment operation data of the crushing and grinding process at the current preset time and combined with the following formula (6):

[0080] (6);

[0081] In formula (6), Indicates the equipment operating characteristics of the crushing and grinding process; Indicates the working efficiency of the crusher; Indicates the impact coefficient of the crusher's operating time; Indicates the crushing rate of the crusher; Indicates the working efficiency of the ball mill; Indicates the influence coefficient of ball mill operation time; Indicates the ball milling rate of the ball mill; and Must be greater than zero and less than or equal to one.

[0082] For the mixing process, the equipment includes a mixer. The equipment operation data of the mixing process at the current preset time may include the working efficiency, operation time influence coefficient and mixing rate of the mixer at the current preset time. In this step, the equipment operation characteristics of the mixing process at the current preset time can be obtained based on the equipment operation data of the mixing process at the current preset time and combined with the following formula (7):

[0083] (7);

[0084] In formula (7), Indicates the equipment operating characteristics of the mixing process; Indicates the working efficiency of the mixer; Indicates the influence coefficient of the mixer's running time; Indicates the stirring rate of the mixer; Must be greater than zero and less than or equal to one.

[0085] Optionally, the equipment operation data of the grinding process and the equipment operation data of the stirring process can be pre-calibrated according to actual conditions, and this embodiment does not specifically limit this.

[0086] Exemplarily, for the equipment used in the crushing and grinding process, the difference between the current preset moment and the previous preset moment is defined as the current time interval, and the difference between the discharge weight of the equipment at the current preset moment and the discharge weight of the equipment at the previous preset moment is defined as the discharge value of the equipment at the current preset moment. Then, the working efficiency of the equipment at the current preset moment can be the ratio of the discharge value of the equipment at the current preset moment to the current time interval; in addition, the difference between the weight of the raw materials inside the equipment at the current preset moment and the weight of the raw materials inside the equipment at the previous preset moment is defined as the raw material value of the equipment at the current preset moment. Then, the operation (i.e., crushing or ball milling) rate of the equipment at the current preset moment can be the ratio of the discharge value of the equipment at the current preset moment to the raw material value of the equipment at the current preset moment.

[0087] As another example, for a mixer in a mixing process, the difference between the current preset moment and the previous preset moment is defined as the current time interval. Then, the working efficiency of the mixer at the current preset moment can be the ratio of the mixing batches completed within the current time interval to the current time interval; in addition, the mixing rate of the mixer at the current preset moment can be the homogeneity of the concrete mixture at the current preset moment. The concept and measurement of the homogeneity of the concrete mixture can follow the prior knowledge based on industry standards, which belongs to the prior art, but is not limited to this.

[0088] For the molding process, considering that gasification and initial condensation are not required in some cases, the equipment of the molding process only considers the cutting machine and the autoclave. The equipment operation data of the molding process at the current preset time can include the working efficiency, operating time influence coefficient and cutting rate of the cutting machine, and the working efficiency, operating time influence coefficient and autoclave rate of the autoclave. In this step, the equipment operation characteristics of the molding process at the current preset time can be obtained by processing the equipment operation data of the molding process at the current preset time, as shown in the following formula (8):

[0089] (8);

[0090] In formula (8), Indicates the equipment operation characteristics of the molding process; Indicates the working efficiency of the cutting machine; Indicates the impact coefficient of the cutting machine's operating time; Indicates the cutting rate of the cutting machine; Indicates the working efficiency of the autoclave; Indicates the influence coefficient of the running time of the autoclave; Indicates the autoclave rate of the autoclave; and Must be greater than zero and less than or equal to one.

[0091] Optionally, the equipment operation data of the forming process can be pre-calibrated according to actual conditions, and this embodiment does not impose specific limitations on this. For example, the difference between the current preset moment and the previous preset moment is defined as the current time interval, the difference between the weight of the blank of the cutting machine at the current preset moment and the weight of the blank of the cutting machine at the previous preset moment is defined as the current cut blank difference, and the difference between the weight of the blank of the autoclave at the current preset moment and the weight of the blank of the autoclave at the previous preset moment is defined as the current autoclaved blank difference. Then, the working efficiency of the cutting machine at the current preset moment can be the ratio of the current cut blank difference to the current time interval, and the working efficiency of the autoclave at the current preset moment can be the ratio of the current autoclaved blank difference to the current time interval; in addition, the working efficiency of the cutting machine at the previous preset moment can be the ratio of the current autoclaved blank difference to the current time interval. The difference between the volume of the blank and the volume of the blank of the cutting machine at the current preset moment is determined as the cutting value of the blank at the current preset moment, then the cutting rate of the cutting machine at the current moment can be the ratio of the cutting value of the blank at the current preset moment to the volume of the blank of the cutting machine at the previous preset moment; the difference between the density of the blank of the autoclave at the previous moment and the density of the blank of the autoclave at the current preset moment is determined as the autoclaving value of the blank at the current preset moment, then the autoclaving rate of the autoclave at the current preset moment can be the ratio of the autoclaving value of the blank at the current preset moment to the density of the blank of the autoclave at the previous moment, and the blank density can be the ratio of the blank weight to the blank volume, but is not limited to this.

[0092] It is worth noting that regarding the equipment's operating time influence coefficient, the equipment's working efficiency and operating conditions are positively correlated with the equipment's operating conditions, i.e., the greater the working efficiency, the better the operating conditions. The operating conditions are mapped by parameters such as crushing rate and ball milling rate. The greater the operating conditions, the better the operating conditions. However, the equipment's operating time is negatively correlated with the equipment's operating conditions, i.e., the greater the operating time, the worse the operating conditions. If only the operating conditions reflected by the equipment's working efficiency and operating conditions are considered, then the operating conditions are difficult to reflect the actual operating conditions. In order to correct the operating conditions reflected by the equipment's working efficiency and operating conditions, this embodiment introduces the above-mentioned operating time influence coefficient, which is expressed as an exponential function: , is the device's operating time (defined as above), , is a coefficient attenuation parameter with a preset value (e.g., 2). Assuming the operating time starts at zero (the equipment is put into operation in its first year after leaving the factory), the impact of the operating time on the equipment's operating conditions is nearly zero. The operating conditions reflected by the equipment's work efficiency and operating conditions are the true conditions. Therefore, the operating time impact coefficient at this time is one, and the operating conditions reflected by the equipment's work efficiency and operating conditions are not corrected. However, as the operating time increases, the impact of the operating time on the equipment's operating conditions gradually increases, and the operating conditions reflected by the equipment's work efficiency and operating conditions become distorted (overstated). Since the operating time impact coefficient gradually increases exponentially, it gradually decreases from one and approaches zero (but will not be equal to zero), thereby correcting the operating conditions reflected by the equipment's work efficiency and operating conditions, reducing them and correcting the operating conditions.

[0093] S302 , obtaining target equipment operation characteristics at a plurality of preset moments based on process equipment operation characteristics at a plurality of preset moments, and further obtaining an equipment operation characteristic sequence.

[0094] It should be noted that the target equipment operation characteristics at the preset time are used to measure the equipment operation status of the aerated concrete brick production line at the preset time.

[0095] In this step, for each preset moment, the equipment operation characteristics of each process at the current preset moment are integrated to obtain the target equipment operation characteristics at the current preset moment. The equipment operation characteristic sequence can be constructed through the target equipment operation characteristics of several preset moments. , Indicates the The target equipment operation characteristics at a preset moment are obtained to obtain the key timing characteristics of the equipment operation dimension. By fully considering the equipment operation conditions of the crushing and grinding process, mixing process and forming process in the aerated concrete brick production line, the feature accuracy associated with the equipment operation conditions of the aerated concrete brick production line can be effectively improved, thereby improving the status detection accuracy of the aerated concrete brick production line.

[0096] For example, for each preset moment, a machine learning method can be used to process the equipment operating characteristics of the crushing and grinding process, the equipment operating characteristics of the stirring process, and the equipment operating characteristics of the forming process at the current preset moment to obtain the target equipment operating characteristics at the current preset moment. For another example, for each preset moment, the equipment operating characteristics of the crushing and grinding process, the equipment operating characteristics of the stirring process, and the equipment operating characteristics of the forming process at the current preset moment are weighted to obtain the target equipment operating characteristics at the current preset moment, as shown in the following formula (9):

[0097] (9);

[0098] In formula (9), Indicates the target device operating characteristics at a preset time; Indicates the operating weight of the crushing and grinding process; Indicates the operation weight of the mixing process; Represents the operating weight of the forming process. Optionally, the operating weights of each process can be pre-set based on actual conditions. For example, the operating weights of each process can be set based on the proportion of the time spent on the aerated concrete brick production line. For example, the operating weight of the grinding process can be the proportion of the time spent on the grinding process to the production time of the aerated concrete brick production line. The same applies to other processes, but is not limited to this.

[0099] In some embodiments, reference Figure 2 and Figure 5 In terms of feature extraction of production quality dimensions, the specific implementation process of step S102 may also include but is not limited to the following steps S401-S403:

[0100] S401, obtaining target crack features at a plurality of preset moments based on image data of each brick at a plurality of preset moments, thereby obtaining a first quality feature sequence;

[0101] S402, obtaining target morphological features at a plurality of preset moments based on the image data and thickness information of each brick at a plurality of preset moments, thereby obtaining a second quality feature sequence;

[0102] S403: According to the weight information of each brick at a plurality of preset moments, target weight characteristics at a plurality of preset moments are obtained, and then a third quality characteristic sequence is obtained.

[0103] It should be noted that the target crack characteristics, target morphological characteristics and target weight characteristics at the preset moment are all used to measure the production quality of the aerated concrete brick production line at the preset moment.

[0104] In this embodiment, in terms of feature extraction of the production quality dimension, this embodiment captures feature information associated with the production quality of the aerated concrete brick production line from three aspects: cracks, morphology, and weight of the finished bricks, and integrates these feature information into a corresponding time feature sequence. Specifically, for each preset moment, the target crack feature at the current preset moment is obtained through the image data of each brick at the current preset moment, and the target morphological feature at the current preset moment is obtained through the image data and thickness information of each brick at the current preset moment, and the target weight feature at the current preset moment is obtained through the weight information of each brick at the current preset moment. This embodiment constructs a first quality feature sequence through the target crack features at several preset moments. The first quality feature sequence can be expressed as , Indicates the The target crack features at a preset moment are used to construct a second quality feature sequence through the target morphological features at several preset moments. The second quality feature sequence can be expressed as , Indicates the The target shape features at a preset moment, and the third quality feature sequence are constructed by the target weight features at several preset moments. The third quality feature sequence can be expressed as , Indicates the The target weight feature at each preset moment is determined. In this way, the production quality of the aerated concrete brick production line is measured by fully considering the cracks, shape and weight of the finished bricks. This can effectively improve the feature accuracy associated with the production quality of the aerated concrete brick production line, thereby improving the status detection accuracy of the aerated concrete brick production line.

[0105] In some embodiments, in the above step S401, obtaining target crack features at a plurality of preset moments based on image data of each brick at a plurality of preset moments may include the following steps S01-S02:

[0106] S01, for each preset moment, performing crack detection on each brick based on the image data of each brick at the preset moment, obtaining the number and location of cracks in each brick at the preset moment, and obtaining crack characteristics of each brick based on the center point position, number and location of cracks in the brick at the preset moment;

[0107] S02, obtaining target crack characteristics at each preset moment based on the crack characteristics of each brick at each preset moment.

[0108] It should be noted that the "number of cracks" refers to the number of cracks in a brick, and the "crack position" includes the location of at least one crack within the brick. Furthermore, the "center point position" refers to the coordinate position of the center point of the brick's image data in the image coordinate system. The image coordinate system is a two-dimensional coordinate system constructed with the lower left corner of the image as the origin, the image width as the horizontal axis, and the image length as the vertical axis.

[0109] In this embodiment, regarding the cracks in the finished bricks, the following operations are performed at each preset moment:

[0110] In the first step, the bricks produced by the aerated concrete brick production line between the last preset time and the current preset time are defined as finished bricks at the current preset time. A camera positioned directly above the bricks is used to capture the finished bricks at the current preset time, obtaining image data of the finished bricks at the current preset time. The length of the bricks is approximately or completely parallel to the length of the image, and the width of the bricks is approximately or completely parallel to the width of the image. Optionally, preprocessing, such as denoising, is performed on the image data of each brick to ensure image quality.

[0111] In the second step, a pre-trained crack detection model is used to identify cracks in the image data of each brick at the current preset moment, and the crack position of each brick at the current preset moment is obtained. The crack position is the position of the candidate box of the crack in the image data in the image coordinate system. The number of cracks can be obtained by counting the crack positions. The crack detection model is trained using a plurality of preset brick sample images and label information corresponding to each brick sample image. The label information corresponding to the brick sample image includes the crack state and crack position of the brick sample image. The crack state includes either containing cracks or not containing cracks. Optionally, the crack detection model is a deep learning model, and its type can be set according to actual conditions. For example, the crack detection model can be a FasterRCNN model, but is not limited to this.

[0112] The third step is to determine the crack characteristics of each brick at the current preset moment by analyzing the center point, number of cracks, and crack location. Specifically, cracks located at the edge of the brick will affect the brick's appearance or local strength. However, cracks near the centerline will affect the brick's load-bearing capacity, seriously affecting the brick's quality. The quality of the bricks can indirectly reflect the production quality of the aerated concrete brick production line. Therefore, in order to further measure the quality of bricks and thus accurately measure the production quality of aerated concrete brick production lines, first, for each crack in the brick, the straight-line distance between the crack position and the center point of the crack is obtained as the crack distance of the crack; then, cracks with a crack distance less than a preset distance threshold are screened from all cracks in the brick as abnormal cracks, and the ratio of the number of abnormal cracks in the brick to the number of cracks in the brick is obtained as the abnormal crack distribution information of the brick. This distribution information can reflect the distribution of cracks that affect the bearing capacity of the brick; at the same time, according to prior knowledge, the more cracks a brick has, the worse the quality of the brick, and the worse the production quality of the aerated concrete brick production line, so the relative quantity error between the number of cracks in the brick and the maximum tolerable crack number threshold is obtained, which is specifically expressed as obtaining the difference between the number of cracks in the brick and the maximum tolerable crack number threshold, and then obtaining the ratio of this difference to the maximum tolerable crack number threshold as the relative quantity error; finally, the abnormal crack distribution information and the relative quantity error of the brick are weighted to obtain the crack characteristics of the brick. Thus, when determining that a brick contains cracks, this embodiment determines the final crack characteristics by considering the impact of the crack location and number on the quality of the brick. This can accurately measure the quality of the brick, and thus accurately measure the production quality of the aerated concrete brick production line, which is beneficial to the state detection accuracy of the aerated concrete brick production line. The crack characteristics of the brick are shown in the following formula (10):

[0113] (10);

[0114] In formula (10), Indicates the crack characteristics of bricks; The weight representing the abnormal crack distribution information; Indicates the number of abnormal cracks in the bricks; Indicates the number of cracks in the brick; Indicates the maximum tolerable crack number threshold; Represents the weight of the relative quantity error.

[0115] Optionally, the probability threshold, distance threshold, and maximum tolerable crack number threshold can be flexibly set based on actual circumstances. Furthermore, the weight of the abnormal crack distribution information and the weight of the relative number error can also be set based on actual circumstances, and this embodiment does not specifically limit this. For example, the weight of the abnormal crack distribution information and the weight of the relative number error can be 0.5, but this is not limited to this. It should be noted that the sum of the weights of the abnormal crack distribution information and the relative number error is one.

[0116] In the fourth step, since the crack characteristics of multiple bricks at the current preset moment are included, in order to obtain the final target crack characteristics, it is necessary to aggregate the crack characteristics of multiple bricks at the current preset moment. For example, a weighted calculation can be performed on the crack characteristics of multiple bricks at the current preset moment to obtain the target crack characteristics at the current preset moment. For another example, the average of the crack characteristics of multiple bricks at the current preset moment can be calculated as the target crack characteristics at the current preset moment.

[0117] By traversing several preset moments in the above manner, target crack features at several preset moments can be obtained, which can improve the feature accuracy associated with the production quality of the aerated concrete brick production line, thereby improving the state detection accuracy of the aerated concrete brick production line.

[0118] In some embodiments, in step S402, obtaining target morphological features at a plurality of preset moments based on the image data and thickness information of each brick at a plurality of preset moments may include the following steps S11-S12:

[0119] S11, for each preset moment, obtaining a binary segmentation mask for each brick at the preset moment based on the image data of each brick at the preset moment, performing morphological detection on the binary segmentation mask for each brick at the preset moment, obtaining width information, length information, and angle information of the brick as morphological information of the brick, and obtaining morphological features of the brick based on the morphological information and thickness information;

[0120] S12, obtaining target morphological features at each preset moment according to the morphological features of each brick at each preset moment.

[0121] In this embodiment, in terms of the shape of the finished bricks, the following operations are performed at each preset moment:

[0122] The first step is to obtain image data and thickness information for each brick at the current preset moment. The thickness information is pre-determined. The acquisition of the image data for the bricks at the current preset moment can be found in the previous embodiment and will not be further described. Optionally, the image data for each brick is pre-processed, such as by denoising, to ensure image data quality.

[0123] In the second step, a pre-trained brick segmentation model is used to perform image segmentation on the image data of each brick at the current preset moment to obtain a binary segmentation mask for each brick at the current preset moment. The brick segmentation model is trained using a plurality of preset brick sample images and the brick regions corresponding to each brick sample image. Optionally, the brick segmentation model is a deep learning model, the type of which can be set based on actual circumstances. For example, the brick segmentation model can be a U-Net model or an ECA-Net model, but is not limited thereto.

[0124] In the third step, for each brick at the current preset moment, first perform morphological operations such as closing operations on the binary segmentation mask of the brick to clean the segmentation result, and then use edge detection operators such as Sobel operator and Canny operator to perform edge detection on the binary segmentation mask of the brick and extract the edge of the brick. Then, based on the edge of the brick, the width information, length information and angle information of the brick are determined as the morphological information of the brick. Finally, if the width information, length information, angle information and thickness information of the brick are all within the corresponding preset range, it means that the shape of the brick is qualified, and the morphological feature of the brick is determined to be one. Otherwise, it means that the shape of the brick is unqualified, and the morphological feature of the brick is determined to be zero. In this way, the morphological condition of the brick can be accurately measured, thereby accurately measuring the production quality of the aerated concrete brick production line, which is beneficial to the status detection accuracy of the aerated concrete brick production line.

[0125] Specifically, in terms of determining morphological information, according to the aforementioned embodiments, the image coordinate system refers to a two-dimensional coordinate system constructed with the lower left corner of the image as the coordinate origin and the two sides of the image as the horizontal and vertical axes. After edge detection is completed, the maximum and minimum values ​​of the brick edge on the horizontal axis in the image coordinate system can be obtained as the horizontal axis extreme values ​​of the brick, and the maximum and minimum values ​​of the brick edge on the vertical axis can be obtained as the vertical axis extreme values ​​of the brick. The absolute value of the difference between the maximum and minimum values ​​of the brick edge on the horizontal axis in the image coordinate system is obtained as the width information of the brick, and the absolute value of the difference between the maximum and minimum values ​​of the brick edge on the vertical axis is obtained as the length information of the brick. At the same time, the vectors of the four sides of the brick can be determined based on the horizontal and vertical extreme values ​​of the brick. For two adjacent sides of the brick, the angle information of the two sides can be obtained by the dot product and cross product of the vectors. In this way, the four angle information of the brick can be obtained. This belongs to the prior art and will not be described in detail. In this way, the width, length, and angle of a brick can be determined solely by its edges, thereby effectively improving the efficiency of obtaining brick shape information. Optionally, the preset ranges corresponding to the width information, length information, angle information, and thickness information can be pre-set based on actual conditions, and this embodiment does not specifically limit this.

[0126] Fourth, since the current preset moment contains morphological features of multiple bricks, to obtain the final target morphological features, these morphological features of the multiple bricks at the current preset moment need to be aggregated. Exemplarily, a weighted summation can be performed on the morphological features of the multiple bricks at the current preset moment to obtain the target morphological features at the current preset moment. Another exemplary embodiment can be the mean of the morphological features of the multiple bricks at the current preset moment, which is then used as the target morphological features at the current preset moment.

[0127] By traversing several preset moments in the above manner, the target morphological features at several preset moments can be obtained, which can improve the feature extraction accuracy associated with the production quality of the aerated concrete brick production line, thereby improving the state detection accuracy of the aerated concrete brick production line.

[0128] In some embodiments, in the above step S403, obtaining target weight characteristics at a plurality of preset moments based on the weight information of each brick at a plurality of preset moments may include S21-S22:

[0129] S21, for each brick at each preset time, if the weight information of the brick is within a preset weight range, the weight feature of the brick is determined to be one; otherwise, the weight feature of the brick is determined to be zero;

[0130] S22, obtaining target weight characteristics at each preset moment according to the weight characteristics of each brick at each preset moment.

[0131] In this embodiment, regarding the weight of finished bricks, the following operations are performed for each preset moment: First, the weight information of each brick at the current preset moment is obtained. The weight information is pre-measured information. Second, for each brick at the current moment, the weight information of the brick is compared with the preset weight range. If the weight information is within the preset weight range, the brick's weight is qualified, and the brick's weight characteristic is determined to be one. Otherwise, the brick's weight is unqualified, and the brick's weight characteristic is determined to be zero. Third, since the current preset moment includes weight characteristics of multiple bricks, in order to obtain the final target weight characteristic, the weight characteristics of multiple bricks at the current preset moment need to be aggregated. For example, the weight characteristics of multiple bricks at the current preset moment can be weighted to obtain the target weight characteristic at the current preset moment. For another example, the average of the weight characteristics of multiple bricks at the current preset moment can be calculated as the target weight characteristic at the current preset moment. By traversing several preset moments in the above manner, the target weight features at several preset moments can be obtained, which can improve the feature extraction accuracy associated with the production quality of the aerated concrete brick production line, thereby improving the status detection accuracy of the aerated concrete brick production line.

[0132] In some embodiments, reference Figure 2 and Figure 6 In the above step S103, image processing is performed based on the power consumption feature sequence, the equipment operation feature sequence and the production quality feature sequence to obtain the state image of the aerated concrete brick production line, which may include the following steps S501-S505.

[0133] S501 , normalizing the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a normalized power consumption feature sequence, a normalized equipment operation feature sequence, and a normalized production quality feature sequence.

[0134] In this step, since the internal eigenvalues ​​of each feature sequence are different, the internal eigenvalues ​​of each feature sequence need to be normalized before feature fusion so that the internal eigenvalues ​​of each feature sequence are between zero and one.

[0135] S502: Fusing the normalized electricity consumption feature sequence and the normalized production quality feature sequence to obtain a first fused feature sequence.

[0136] In this step, among the three dimensions of electricity consumption, production quality, and equipment operation, the production quality dimension can usually directly map the state of the aerated concrete brick production line, which is a direct dimension, while the electricity consumption dimension and the equipment operation dimension often indirectly map the state of the aerated concrete brick production line, which are indirect dimensions. In order to promote the feature fusion between direct and indirect dimensions, according to the above embodiment, the production quality dimension can be subdivided into three aspects: cracks, shape, and weight of the finished bricks. Based on this, this step uses the normalized first quality feature sequence as the query and the normalized electricity consumption feature sequence as the key and value, and uses the self-attention mechanism to fuse the above two feature sequences to obtain a first fused feature sequence, aiming to promote the feature fusion between the electricity consumption dimension and the production quality dimension, so that the electricity consumption feature sequence and the first quality feature sequence can perceive each other's feature information related to the state of the aerated concrete brick production line, thereby obtaining a more comprehensive and accurate time feature sequence. In this way, it is possible to better mine the feature information hidden in the data of each dimension, capture more real state expressions, and help improve the state detection accuracy of the aerated concrete brick production line.

[0137] S503: Fusing the normalized equipment operation feature sequence and the normalized production quality feature sequence to obtain a second fused feature sequence.

[0138] In this step, in order to promote the feature fusion between the direct dimension and the indirect dimension, according to the above embodiment, the production quality dimension can be subdivided into three aspects: cracks, shape and weight of finished bricks. Based on this, this step uses the normalized second quality feature sequence as the query and the normalized equipment operation feature sequence as the key and value, and uses the self-attention mechanism to fuse the above two feature sequences to obtain a second fused feature sequence, aiming to promote the feature interaction and fusion between the electricity consumption dimension and the production quality dimension, so that the equipment operation feature sequence and the second quality feature sequence can perceive each other's feature information related to the status of the aerated concrete brick production line, thereby obtaining a more comprehensive and accurate time feature sequence, which can better mine the feature information hidden in the data of each dimension and capture more real state expressions, which is conducive to improving the state detection accuracy of the aerated concrete brick production line.

[0139] S504 , fusing the normalized electricity usage feature sequence, the normalized equipment operation feature sequence, and the normalized production quality feature sequence to obtain a third fused feature sequence.

[0140] In this step, in order to promote the feature fusion between the direct dimension and the indirect dimension, in addition to fusing the direct dimension with the two indirect dimensions separately, it is also necessary to fuse the direct dimension with the two indirect dimensions simultaneously. Specifically, this step first jump-connects the normalized power consumption feature sequence and the normalized equipment operation feature sequence to obtain a fourth fused feature sequence. The fourth fused feature sequence has the same size as the normalized third quality feature sequence. This is intended to promote the feature interaction and fusion between the power consumption dimension and the equipment operation dimension, so that the power consumption feature sequence and the equipment operation feature sequence can perceive each other's feature information related to the state of the aerated concrete brick production line. Then, using the normalized third quality feature sequence as the query and the fourth fused feature sequence as the key and value, the self-attention mechanism is used to fuse the above two feature sequences to obtain the third fused feature sequence. This is intended to achieve feature interaction and fusion between the power consumption dimension, the equipment operation dimension, and the production quality dimension, thereby obtaining a more comprehensive and accurate time feature sequence. This can better mine the state feature information hidden in the data of each dimension, capture more real state expressions, and help improve the state detection accuracy of the aerated concrete brick production line.

[0141] S505 , performing image processing based on the first fused feature sequence, the second fused feature sequence, and the third fused feature sequence to obtain a state image.

[0142] In this step, each fused feature sequence is processed by dimensionality increase through imaging, and the obtained two-dimensional image is used as data information reflecting the real state of the aerated concrete brick production line.

[0143] In some embodiments, reference Figure 2 In the above step S504, image processing is performed based on the first fusion feature sequence, the second fusion feature sequence and the third fusion feature sequence to obtain a state image, which may include the following steps S601-S604.

[0144] S601: Obtain a red channel image according to the first fusion feature sequence.

[0145] In this step, according to the first fusion feature sequence, the Gram sum angular field matrix is ​​constructed, and the red channel map is generated by the Gram sum angular field matrix, which is specifically implemented as prior art and will not be described in detail. The Gram sum angular field matrix can ensure time correlation, and as time increases, the Gram sum angular field matrix moves from the upper left to the lower right, and it represents relative correlation by the superposition of time direction. In addition, the Gram sum angular field matrix represents the correlation between each pair of points by the superposition of nonlinear cosine function in the polar coordinate system, and different time feature sequences have different features. After feature reconstruction, the feature information of the time feature sequence is enhanced, and then the difference between the feature information can be more highlighted.

[0146] S602: Obtain a green channel image according to the second fusion feature sequence.

[0147] In this step, a relative position matrix is ​​constructed based on the second fused feature sequence, and a green channel image is generated using the relative position matrix. The specific implementation of this is conventional technology and will not be described in detail here. The relative position matrix can incorporate redundant features from the original fused feature sequence, making it easier to capture inter-class and intra-class similarity information in the converted channel image, thereby providing richer feature information.

[0148] S603: Obtain a blue channel image according to the third fusion feature sequence.

[0149] In this step, a recursive graph matrix is ​​constructed based on the third fused feature sequence, and a blue channel map is generated using the recursive graph matrix. The specific implementation of this is conventional technology and will not be further described. The recursive graph matrix can intuitively capture the characteristic information of periodicity, chaos, or other complex dynamic behaviors in the fused feature sequence, thereby providing richer feature information.

[0150] S604 , obtaining a state image according to the red channel image, the green channel image, and the blue channel image.

[0151] In this step, the red channel image, green channel image, and blue channel image obtained by different visualization methods are superimposed into a three-channel RGB image. The RGB image is the state image. In this way, through three different types of visualization methods, the feature information in each fused feature sequence can be maximized while converting the image, ensuring the accuracy of the state image, which is conducive to giving full play to the advantages of the deep learning model in image processing and effectively improving the state detection accuracy of the aerated concrete brick production line.

[0152] In addition, refer to Figure 7 , the embodiment of the present application further provides a state detection device for an aerated concrete brick production line, comprising:

[0153] Acquisition module 701, used to acquire the electricity consumption dataset, equipment dataset and brick dataset of the aerated concrete brick production line;

[0154] Feature extraction module 702, used to extract features from the electricity consumption dataset, equipment dataset, and brick dataset to obtain the electricity consumption feature sequence, equipment operation feature sequence, and production quality feature sequence of the aerated concrete brick production line;

[0155] An image processing module 703 is used to perform image processing based on the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a state image of the aerated concrete brick production line;

[0156] The detection module 704 is used to detect the status of the aerated concrete brick production line according to the status image and obtain a status detection result of the aerated concrete brick production line.

[0157] The contents of the above method embodiments are all applicable to the present device embodiments. The functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0158] To facilitate understanding of the state detection method and device for the aerated concrete brick production line described above in the present application, the actual application scenario of the state detection method and device for the aerated concrete brick production line described above in the present application is illustrated herein by way of example.

[0159] In this application scenario, the status of the aerated concrete brick production line needs to be detected the previous day. Therefore, the previous day is evenly divided into 24 moments, each moment representing one hour. The specific process of detecting the status of the aerated concrete brick production line in this application scenario is as follows:

[0160] S801, data acquisition:

[0161] The number of bricks produced between two adjacent moments is 10,000. Due to the large number of bricks, 1,000 bricks are selected from the 10,000 bricks by random sampling as finished bricks, that is, each moment corresponds to 1,000 finished bricks. The finished bricks at each moment are photographed by a camera set directly above the bricks, and the image data of the finished bricks at each moment can be obtained; at the same time, the weight and thickness information of the finished bricks at each moment are collected. This information can be obtained by measuring with automated equipment, thus obtaining a brick data set for the aerated concrete brick production line.

[0162] Furthermore, by measuring data on the equipment in the crushing, mixing, and molding processes of the aerated concrete brick production line, we can obtain electricity consumption data and equipment operation data for each process at 24 moments in time. In addition, by measuring data on the environment in which the aerated concrete brick production line is located, we can obtain cloud cover information, air pressure information, and relative humidity information at 24 moments in time, thereby obtaining the electricity consumption data set and equipment data set of the aerated concrete brick production line.

[0163] Among them, the electricity consumption data of the crushing and grinding process includes the electricity consumption of the crusher, the electricity consumption impact coefficient of the crusher, the electricity consumption of the ball mill and the electricity consumption impact coefficient of the ball mill; the equipment operation data of the crushing and grinding process includes the working efficiency, operating time impact coefficient and crushing rate of the crusher, as well as the working efficiency, operating time impact coefficient and ball milling rate of the ball mill; the equipment operation data of the mixing process includes the working efficiency, operating time impact coefficient and mixing rate of the mixer at the current preset moment; the electricity consumption data of the mixing process may include the electricity consumption of the mixer and the electricity consumption impact coefficient of the raw material ratio on the mixer; the electricity consumption data of the molding process includes the electricity consumption of the cutting machine and the electricity consumption of the autoclave; the equipment operation data of the molding process includes the working efficiency, operating time impact coefficient and cutting rate of the cutting machine, as well as the working efficiency, operating time impact coefficient and autoclave rate of the autoclave.

[0164] S802, feature extraction:

[0165] At each moment, the electricity consumption impact data at the current moment is combined with the above formula (1) to obtain the electricity consumption impact characteristics at the current moment and perform maximum-minimum normalization; and the electricity consumption data of the process at the current moment is combined with the above formulas (2)-(4) to obtain the electricity consumption characteristics of the grinding process, mixing process, and molding process at the current moment. Based on the proportion of the time of each process in the production time of the aerated concrete brick production line and the electricity consumption characteristics of each process at the current moment, the comprehensive electricity consumption characteristics at the current moment are obtained; then, based on the normalized electricity consumption impact characteristics and the comprehensive electricity consumption characteristics, combined with the above formula (5), the target electricity consumption characteristics at the current moment are obtained. The electricity consumption characteristic sequence of the aerated concrete brick production line is constructed through the target electricity consumption characteristics of the twenty-four moments. This sequence is a time characteristic sequence.

[0166] Secondly, at each moment, the equipment operation characteristics of the grinding process are obtained based on the equipment operation data of the grinding process combined with the above formula (6), the equipment operation characteristics of the mixing process are obtained based on the equipment operation data of the mixing process combined with the above formula (7), and the equipment operation characteristics of the molding process are obtained based on the equipment operation data of the molding process combined with the above formula (8). Then, based on the proportion of the time taken by each process in the production time of the aerated concrete brick production line and the equipment operation characteristics of each process at the current moment, the target equipment operation characteristics at the current moment are obtained, as shown in the above formula (9). The equipment operation characteristic sequence of the aerated concrete brick production line is constructed through the target equipment operation characteristics of the twenty-four moments, which is a time characteristic sequence.

[0167] Furthermore, at each moment, the image data of each finished brick at the current moment is detected based on the trained FasterRCNN model to obtain the number and location of cracks in each finished brick at the current moment. For each brick, for each crack contained in the finished brick, the straight-line distance between the crack location and the center point of the finished brick is obtained as the crack distance of the crack. From all the cracks in the finished brick, cracks with a crack distance less than the distance threshold are screened as abnormal cracks, and the ratio of the number of abnormal cracks in the finished brick to the number of cracks in the finished brick is obtained as the abnormal crack distribution information of the finished brick. At the same time, the difference between the number of cracks in the finished brick and the maximum tolerable crack number threshold is obtained, and the ratio of this difference to the maximum tolerable crack number threshold is obtained as the relative number error. Finally, the abnormal crack distribution information and the relative number error of the finished brick are weighted to obtain the crack feature of the finished brick, as shown in the above formula (10). The average of the crack features of multiple finished bricks at the current moment is taken as the target crack feature at the current moment. The first quality feature sequence of the aerated concrete brick production line is constructed through the target crack features at twenty-four moments. This sequence is a time feature sequence.

[0168] Also, for each moment, the image data of each finished brick at the current moment is segmented based on the trained ECA-Net model to obtain the binary segmentation mask of each finished brick at the current moment. For each finished brick, morphological operations and edge detection are performed on the binary segmentation mask of the finished brick to obtain the edge of the finished brick. The length, width and angle of the finished brick are calculated based on the edge of the finished brick. If the width information, length information, angle information and thickness information of the finished brick are all within the corresponding preset range, the morphological feature of the finished brick is one, otherwise the morphological feature of the finished brick is zero. The average of the morphological features of multiple finished bricks at the current moment is used as the target morphological feature at the current moment. The second quality feature sequence of the aerated concrete brick production line is constructed through the target morphological features of twenty-four moments, and this sequence is a time feature sequence.

[0169] Furthermore, for each finished brick at each moment, if the weight information of the finished brick is within the preset weight range, the weight characteristic of the finished brick is one; otherwise, the weight characteristic of the finished brick is zero. The average of the weight characteristics of the multiple finished bricks at the current moment is used as the target weight characteristic at that moment. The third quality characteristic sequence of the aerated concrete brick production line is constructed from the target weight characteristics at twenty-four moments. This sequence is a time characteristic sequence.

[0170] S803, Feature Fusion:

[0171] The electricity usage feature sequence, equipment operation feature sequence, and production quality feature sequence are normalized so that the internal eigenvalue of each feature sequence is between zero and one. Using the normalized first quality feature sequence as the query and the normalized electricity usage feature sequence as the key and value, the two feature sequences are fused using a self-attention mechanism to obtain a first fused feature sequence. Furthermore, using the normalized second quality feature sequence as the query and the normalized equipment operation feature sequence as the key and value, the two feature sequences are fused using a self-attention mechanism to obtain a second fused feature sequence. Furthermore, the normalized electricity usage feature sequence and the normalized equipment operation feature sequence are first skip-connected to obtain a fourth fused feature sequence. Then, using the normalized third quality feature sequence as the query and the fourth fused feature sequence as the key and value, the two feature sequences are fused using a self-attention mechanism to obtain a third fused feature sequence.

[0172] S804, visualization:

[0173] A Gram sum angular field matrix is ​​constructed according to the first fused feature sequence, and the red channel is generated through the Gram sum angular field matrix. A relative position matrix is ​​constructed according to the second fused feature sequence, and the green channel map is generated through the relative position matrix. A recursive graph matrix is ​​constructed according to the third fused feature sequence, and the blue channel map is generated through the recursive graph matrix. Finally, the red channel map, green channel map, and blue channel map are superimposed into a three-channel RGB image to obtain the status image of the aerated concrete brick production line.

[0174] S805, status detection:

[0175] The state image of the aerated concrete brick production line is input into the pre-trained AdaBoost model to obtain the detection results that characterize whether the aerated concrete brick production line is qualified, thereby realizing the state detection of the aerated concrete brick production line.

[0176] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and intent of the present application, and that the scope of the present application is defined by the claims and their equivalents.

[0177] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present application.

Claims

1. A method for detecting the status of an aerated concrete brick production line, characterized in that: The following steps are involved: Obtaining an electricity consumption dataset, an equipment dataset, and a brick dataset for the aerated concrete brick production line; Performing feature extraction on the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain an electricity consumption feature sequence, an equipment operation feature sequence, and a production quality feature sequence of the aerated concrete brick production line; Performing image processing based on the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a state image of the aerated concrete brick production line; Detecting the state of the aerated concrete brick production line according to the state image to obtain a state detection result of the aerated concrete brick production line; The production quality feature sequence includes a first quality feature sequence, a second quality feature sequence, and a third quality feature sequence. The image processing based on the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a state image of the aerated concrete brick production line includes: Normalizing the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a normalized power consumption feature sequence, a normalized equipment operation feature sequence, and a normalized production quality feature sequence; Using the normalized electricity usage feature sequence as a key and a value and the normalized first quality feature sequence as a query, the normalized electricity usage feature sequence and the normalized first quality feature sequence are fused using a self-attention mechanism to obtain a first fused feature sequence; Using the normalized device operation feature sequence as a key and a value and the normalized second quality feature sequence as a query, the normalized device operation feature sequence and the normalized second quality feature sequence are fused using a self-attention mechanism to obtain a second fused feature sequence; Performing a jump connection on the normalized power usage feature sequence and the normalized device operation feature sequence to obtain a fourth fused feature sequence, using the fourth fused feature sequence as a key and value and the normalized third quality feature sequence as a query, and fusing the fourth fused feature sequence and the normalized third quality feature sequence using a self-attention mechanism to obtain a third fused feature sequence; The state image is obtained by performing image processing based on the first fusion feature sequence, the second fusion feature sequence and the third fusion feature sequence.

2. The method according to claim 1, characterized in that The power consumption data set includes power consumption impact data and process power consumption data at a plurality of preset moments, wherein the power consumption impact data is used to indicate environmental data that affects the power load of the aerated concrete brick production line, and the process power consumption data includes power consumption data of the grinding process, the mixing process, and the forming process in the aerated concrete brick production line; The feature extraction of the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain the electricity consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence of the aerated concrete brick production line includes: Obtaining the electricity consumption impact characteristics of each preset moment according to the electricity consumption impact data at each preset moment; According to the power consumption data of each process at the preset time, the power consumption characteristics of each process at the preset time are obtained; wherein the power consumption characteristics of the process include the power consumption characteristics of the crushing and grinding process, the power consumption characteristics of the stirring process, and the power consumption characteristics of the forming process; According to the power consumption impact characteristics and process power consumption characteristics of the plurality of preset moments, target power consumption characteristics of the plurality of preset moments are obtained, and then the power consumption characteristic sequence is obtained.

3. The method according to claim 1, characterized in that The equipment data set includes equipment operation data at a plurality of preset moments, and the equipment operation data includes equipment operation data of a grinding process, a mixing process, and a forming process in the aerated concrete brick production line; The feature extraction of the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain the electricity consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence of the aerated concrete brick production line includes: According to the equipment operation data at each of the preset moments, the process equipment operation characteristics at each of the preset moments are obtained; wherein the process equipment operation characteristics include the equipment operation characteristics of the crushing and grinding process, the equipment operation characteristics of the stirring process, and the equipment operation characteristics of the forming process; According to the process equipment operation characteristics at the plurality of preset moments, target equipment operation characteristics at the plurality of preset moments are obtained, and then the equipment operation characteristic sequence is obtained.

4. The method according to claim 1, wherein The brick dataset includes image data, weight information and thickness information of at least one brick at a plurality of preset moments; The feature extraction of the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain the electricity consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence of the aerated concrete brick production line includes: Obtaining target crack features at the plurality of preset moments based on the image data of each brick at the plurality of preset moments, and thereby obtaining the first quality feature sequence; Obtaining target morphological features at the plurality of preset moments based on the image data and thickness information of each brick at the plurality of preset moments, and thereby obtaining the second quality feature sequence; According to the weight information of each brick at several preset moments, the target weight characteristics at several preset moments are obtained, and then the third quality characteristic sequence is obtained.

5. The method according to claim 4, characterized in that The step of obtaining target crack features at a plurality of preset moments based on the image data of each brick at the plurality of preset moments includes: At each of the preset moments, crack detection is performed on each of the bricks based on the image data of each of the bricks at the preset moments to obtain the number of cracks and the location of the cracks in each of the bricks at the preset moments. For each of the bricks at the preset moments, crack characteristics of the bricks are obtained based on the center point position of the brick and the number and location of the cracks; According to the crack characteristics of each brick at each preset moment, the target crack characteristics at each preset moment are obtained.

6. The method according to claim 4, characterized in that The target morphological features at the preset moments are obtained based on the image data and thickness information of each brick at the preset moments, including: For each of the preset moments, a binary segmentation mask of each of the bricks at the preset moment is obtained based on the image data of each of the bricks at the preset moment; for each of the bricks at the preset moment, morphological detection is performed on the binary segmentation mask of the brick to obtain width information, length information, and angle information of the brick as morphological information of the brick; and morphological features of the brick are obtained based on the morphological information and the thickness information; According to the morphological features of each brick at each preset moment, the target morphological features at each preset moment are obtained.

7. The method according to claim 4, characterized in that The step of obtaining target weight characteristics at a plurality of preset moments based on the weight information of each brick at the plurality of preset moments includes: For each brick at each preset time, if the weight information of the brick is within a preset weight range, the weight feature of the brick is determined to be one; otherwise, the weight feature of the brick is determined to be zero; According to the weight characteristics of each brick at each preset moment, the target weight characteristics at each preset moment are obtained.

8. The method according to claim 1, characterized in that The performing image processing based on the first fusion feature sequence, the second fusion feature sequence, and the third fusion feature sequence to obtain the state image includes: Obtaining a red channel image according to the first fusion feature sequence; Obtaining a green channel image according to the second fusion feature sequence; Obtaining a blue channel image according to the third fusion feature sequence; The state image is obtained according to the red channel image, the green channel image, and the blue channel image.

9. A state detection device for an aerated concrete brick production line, characterized in that: A method for detecting the state of an aerated concrete brick production line according to any one of claims 1 to 8, wherein the device comprises: An acquisition module, configured to acquire an electricity consumption dataset, an equipment dataset, and a brick dataset of the aerated concrete brick production line; a feature extraction module, configured to perform feature extraction on the electricity consumption dataset, the equipment dataset, and the brick dataset to obtain an electricity consumption feature sequence, an equipment operation feature sequence, and a production quality feature sequence for the aerated concrete brick production line; an image processing module, configured to perform image processing based on the power consumption feature sequence, the equipment operation feature sequence, and the production quality feature sequence to obtain a state image of the aerated concrete brick production line; A detection module is used to detect the state of the aerated concrete brick production line according to the state image to obtain a state detection result of the aerated concrete brick production line.

Images