Aluminum processing risk early warning system and method

Abstract

The invention relates to an aluminum processing risk early warning system and method, and the method comprises a data collection step, a data analysis step, and a judgment step. A data acquisition step: acquiring video data and thermal imaging data of an aluminum processing area through a camera, and acquiring sensing data through a data acquisition gateway; a data analysis step: analyzing production behavior characteristics based on the video data; monitoring the real-time temperature of the mold and / or the molten aluminum launder based on the thermal imaging data; a judgment step: after judging that each feature in the sensing data is not matched, further judging whether a production behavior feature exists or not or whether the real-time temperature is greater than a preset temperature threshold after the mold acts, if so, generating an abnormality, and otherwise, not generating an abnormality; the video data and the thermal imaging data are analyzed to obtain the production behavior characteristics and the temperature state, and whether the working state is normal or not is judged crosswise based on the production behavior characteristics, the temperature state and the sensing data, so that the accuracy of risk early warning is improved.

Description

Technical Field

[0001] This application belongs to the technical field of aluminum processing supervision, and particularly relates to an aluminum processing risk early warning system and method. Background Art

[0002] The aluminum processing industry is an important part of the global manufacturing industry, covering multiple links such as aluminum smelting, processing, forming, and application. With the development of technology and the enhancement of environmental awareness, aluminum, as a lightweight, high-strength, and corrosion-resistant metal material, has gradually expanded its application fields.

[0003] Aluminum processing casting is a key process of injecting molten aluminum into a mold for forming, including steps of melting, refining, casting, cooling, and demolding; during the aluminum processing casting process, the molten aluminum is transported to the mold through a launder system and forms aluminum ingots or profiles after cooling and solidification. There is a risk of high-temperature molten aluminum leakage during the aluminum processing casting process. Aluminum processing enterprises vary in size, the brands of production equipment are different, and the production environments also vary greatly. Generally, medium and large-sized enterprises have a relatively high degree of automation integration and relatively good application of automated production technology.

[0004] However, in some small and medium-sized aluminum processing enterprises, due to the situation that the equipment casting system and the safety protection system are independent of each other or have a low degree of integration, significant safety management problems are exposed during the actual production process; moreover, based on the single alarm trigger condition of the safety protection system, it is easy to misjudge the production operation, or it is impossible to identify hidden potential risks and achieve an ideal risk early warning effect. Summary of the Invention

[0005] In view of the deficiencies in the related technologies, this application provides an aluminum processing risk early warning system and method, which analyzes video data and thermal imaging data to obtain production behavior characteristics and temperature states, and cross-judges whether the working state is normal based on the production behavior characteristics, temperature states, and perception data, thereby improving the accuracy of risk early warning.

[0006] On the one hand, this application provides an aluminum processing risk early warning method, including:

[0007] A data collection step, collecting video data and thermal imaging data of the aluminum processing area through a camera, and obtaining perception data through a data acquisition gateway, where the perception data includes the start / stop signal of the aluminum processing system, the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the water inlet pressure of the mold, and the water inlet flow rate of the mold;

[0008] A data analysis step, analyzing production behavior characteristics based on the video data, where the production behavior characteristics include personnel gathering, visible flames, and visible smoke; monitoring the real-time temperature of the mold and / or the aluminum liquid launder based on the thermal imaging data;

[0009] Judgment step: Judge whether the liquid level of the outlet launder or the inlet liquid level of the distribution launder exceeds the preset liquid level, and judge whether the water inlet pressure or the water inlet flow rate of the mold crystallizer exceeds the water inlet threshold. If both are yes and the start-stop signal of the aluminum processing system is off, then further judge whether there is the production behavior characteristic or the real-time temperature is greater than the preset temperature threshold after the mold action. If so, an abnormality occurs; otherwise, no abnormality is found.

[0010] In some embodiments, the aluminum processing risk warning method further includes:

[0011] First alarm step: When it is detected that the start-stop signal of the aluminum processing system is off, if the liquid level of the outlet launder or the inlet liquid level of the distribution launder exceeds the preset liquid level, and the water inlet pressure or the water inlet flow rate of the mold crystallizer exceeds the water inlet threshold, then send a first alarm signal to the control center.

[0012] In some embodiments, the aluminum processing risk warning method further includes:

[0013] Second alarm step: After detecting the first alarm signal, further judge whether there is the production behavior characteristic or the real-time temperature is greater than the preset temperature threshold after the mold action. If so, send a second alarm signal to the control center; otherwise, do not alarm.

[0014] In some embodiments, the judgment step further includes:

[0015] First judgment step: When it is monitored that the start-stop signal of the aluminum processing system stops being sent, judge whether the liquid level of the outlet launder or the inlet liquid level of the distribution launder exceeds the preset liquid level. If so, send a first judgment signal to continue the judgment; otherwise, stop the judgment.

[0016] In some embodiments, the judgment step further includes:

[0017] Second judgment step: When the first judgment signal is detected, judge whether the water inlet pressure or the water inlet flow rate of the mold crystallizer exceeds the water inlet threshold. If so, send a second judgment signal to continue the judgment; otherwise, stop the judgment.

[0018] In some embodiments, the judgment step further includes:

[0019] Third judgment step: When the second judgment signal is detected, judge whether there are characteristics of personnel gathering, visible flames, and visible smoke in the aluminum processing area within a period of time after the mold action. If so, an abnormality occurs; otherwise, no abnormality is found.

[0020] In some embodiments, the judgment step further includes:

[0021] Fourth judgment step: When the second judgment signal is detected, judge whether the real-time temperature of the mold and / or the molten aluminum chute is greater than a preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

[0022] In some embodiments, the data analysis step further includes:

[0023] First analysis step: Based on the video data, divide multiple monitoring screens, where the monitoring screens at least include a flame recognition area, a smoke recognition area, a mold erection action recognition area, and a personnel aggregation recognition area; and based on the multiple monitoring screens, monitor the appearance or disappearance of the production behavior characteristics.

[0024] In some embodiments, the data analysis step further includes:

[0025] Second analysis step: Analyze the temperature matrix through an image processing algorithm to extract the temperature distribution characteristics of the mold and / or the molten aluminum chute. The temperature distribution characteristics include identifying high-temperature areas, calculating the average temperature, and detecting the temperature gradient; and based on the temperature distribution characteristics, obtain the real-time temperature near the mold and / or the molten aluminum chute.

[0026] On the other hand, an aluminum processing risk warning system is also provided, including:

[0027] A data acquisition module, configured to collect video data and thermal imaging data of the aluminum processing area through a camera, and obtain perception data through a data acquisition gateway. The perception data includes an aluminum processing system start / stop signal, an outlet chute liquid level, an inlet liquid level of the distribution chute, a crystallizer water inlet pressure, and a crystallizer water inlet flow rate;

[0028] A data analysis module, configured to analyze production behavior characteristics based on the video data, where the production behavior characteristics include personnel aggregation, visible flames, and visible smoke; and monitor the real-time temperature of the mold and / or the molten aluminum chute based on the thermal imaging data;

[0029] A judgment module, configured to judge whether the outlet chute liquid level or the inlet liquid level of the distribution chute exceeds a preset liquid level, and judge whether the crystallizer water inlet pressure or the crystallizer water inlet flow rate exceeds a water inlet threshold. If both are yes and the aluminum processing system start / stop signal is off, further judge whether there are the production behavior characteristics after the mold action or whether the real-time temperature is greater than a preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

[0030] Based on the above technical solutions, the present application provides an aluminum processing risk warning system and method. By analyzing video data and thermal imaging data, production behavior characteristics and temperature states are obtained, and based on the production behavior characteristics, temperature states, and perception data, it is cross-judged whether the working state is normal, thereby improving the accuracy of risk warning. Through the first alarm step, a warning is given when the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the water inlet pressure of the mold, the water inlet flow rate of the mold, and the start-stop signal of the aluminum processing system do not match in the perception data. Through the second alarm step, an alarm is given after further judging that the production behavior characteristics or the real-time temperature is abnormal. Through the first judgment step and the second judgment step, it is judged whether to continue the judgment, thereby streamlining the judgment process. Through the third judgment step, it is judged whether production behavior characteristics appear, thereby judging whether production anomalies occur. Through the fourth judgment step, it is judged whether the real-time temperature is higher than the preset temperature threshold, thereby judging whether production anomalies occur. Through the first analysis step, the video data is analyzed to monitor the appearance or disappearance of production behavior characteristics. Through the second analysis step, the thermal imaging data is analyzed to obtain the real-time temperature near the mold and / or the aluminum liquid launder.

[0031] Other features and advantages of the present invention will be described in the following description, and some will become obvious from the description in the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures pointed out in the specification, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The schematic embodiments and descriptions thereof of the present application are used to explain the present application, and do not constitute an improper limitation to the present application. In the drawings:

[0033] Figure 1 is the general flow chart of the aluminum processing risk warning method of the present application;

[0034] Figure 2 is the flow chart of the data analysis step of the aluminum processing risk warning method of the present application;

[0035] Figure 3 is the flow chart of the judgment step of the aluminum processing risk warning method of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0037] In the description of the present application, it should be understood that the orientation or positional relationship indicated by the terms "center", "lateral", "longitudinal", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present application.

[0038] The terms "first", "second", "third" are only used for descriptive purposes and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first", "second", "third" may explicitly or implicitly include one or more of such features.

[0039] In the description of the present application, it should be noted that unless otherwise clearly specified and limited, the terms "installation", "connection", "coupling" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a direct connection or an indirect connection through an intermediate medium, and it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances. Specific Embodiment 1

[0041] Referring to the attached Figure 1 drawings, the present application provides an aluminum processing risk early warning method, including a data collection step S1, a data analysis step S2, and a judgment step S3.

[0042] In the data collection step S1, video data and thermal imaging data of the aluminum processing area are collected through a camera, and sensing data is obtained through a data acquisition gateway. The sensing data includes the start / stop signal of the aluminum processing system, the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the inlet pressure of the mold, and the inlet flow rate of the mold.

[0043] In the data analysis step S2, the production behavior characteristics are analyzed based on the video data. The production behavior characteristics include personnel gathering, visible flame, and visible smoke. The real-time temperature of the mold and / or the aluminum liquid launder is monitored based on the thermal imaging data.

[0044] In the judgment step S3, it is judged whether the liquid level of the outlet launder or the liquid level at the inlet of the distribution launder exceeds the preset liquid level, and it is judged whether the inlet pressure of the mold or the inlet flow rate of the mold exceeds the inlet threshold. If both are yes and the start / stop signal of the aluminum processing system is off, then it is further judged whether there are production behavior characteristics after the mold action or whether the real-time temperature is greater than the preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

[0045] Among them, the aluminum processing risk early warning method includes a data collection step S1, which collects video data and thermal imaging data of the aluminum processing area through a camera, and obtains perception data through a data acquisition gateway. The perception data includes the start-stop signal of the aluminum processing system, the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the water inlet pressure of the mold, and the water inlet flow rate of the mold.

[0046] Specifically, deploy high-definition cameras at key positions in the aluminum processing area to ensure that key areas such as casting equipment, casting launders, and personnel operation areas can be covered.

[0047] Collect video data of the aluminum processing area through a camera. The monitoring screen in the video data includes at least a flame recognition area, a smoke recognition area, a mold erection action recognition area, and a personnel gathering recognition area.

[0048] Moreover, deploy infrared thermal imaging cameras at least on the aluminum liquid launder, the surface of the mold, and near the mold to obtain thermal imaging data of these areas.

[0049] Deploy a data acquisition gateway in the control system of the aluminum processing equipment as the center for data collection and transmission; the data acquisition gateway should have multi-protocol support capabilities and be able to be compatible with equipment of different brands and models.

[0050] Connect the data acquisition gateway to the PLC (programmable logic controller) through industrial communication protocols such as Modbus or OPC UA. The PLC is responsible for reading perception data from on-site sensors and actuators and transmitting it to the data acquisition gateway through the protocol.

[0051] After receiving the data transmitted by the PLC, the data acquisition gateway performs necessary data preprocessing tasks, such as data cleaning, format conversion, outlier processing, etc., to ensure the accuracy and availability of the data.

[0052] The data acquisition gateway pushes the preprocessed real-time data to the server platform for storage through a secure network channel (such as VPN) for subsequent analysis to obtain the operating status of the aluminum processing system.

[0053] Aluminum processing casting is a key process for injecting molten aluminum into a mold for shaping, including steps of melting, refining, casting, cooling, and demolding. Among them, the aluminum liquid flow process in the casting process is as follows: the molten aluminum liquid first enters the distribution launder and then flows into multiple molds (crystallizers) through the outlet launder.

[0054] The distribution launder is a transition launder for distributing aluminum liquid from the furnace to multiple molds. The liquid level at the inlet of the distribution launder refers to the initial height when the aluminum liquid enters the distribution launder.

[0055] The outlet launder is the last section of the channel for the aluminum liquid to flow from the distribution launder to the mold. The liquid level of the outlet launder refers to the height of the aluminum liquid in the launder. A sudden drop in the liquid level of the outlet launder may indicate aluminum leakage.

[0056] The mold is the core device for cooling and shaping molten aluminum. The water inlet pressure of the mold refers to the pipeline pressure when the cooling water enters the mold; the water inlet flow rate of the mold refers to the amount of water passing through the cooling water circulation system of the mold per unit time.

[0057] Among them, the sensed data includes start / stop signals of the aluminum processing system, the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the water inlet pressure of the mold, the water inlet flow rate of the mold, etc.

[0058] Refer to Appendix Figure 1 , the aluminum processing risk warning method includes a data analysis step S2, analyzing the production behavior characteristics based on video data analysis, and the production behavior characteristics include personnel gathering, visible flames, and visible smoke; monitoring the real-time temperature of the mold and / or the aluminum liquid launder based on thermal imaging data.

[0059] Specifically, use video AI technology to monitor the mold actions in the video data of the aluminum processing area, and extract the production behavior characteristics, which include visible flame characteristics, visible smoke characteristics, and personnel gathering characteristics.

[0060] Among them, the mold action characteristic is the preparation action of the casting equipment standing upright at the working position. By monitoring whether there is a mold action characteristic in the aluminum processing area, it is possible to analyze whether the aluminum processing area is in a normal working state.

[0061] The visible flame characteristic is the open fire generated during aluminum processing production. By monitoring whether there is a flame characteristic in the aluminum processing area, it is possible to analyze whether the aluminum processing area is in a normal working state.

[0062] The visible smoke characteristic is the smoke generated by the evaporation of water vapor during aluminum processing production. By monitoring whether there is a smoke characteristic in the aluminum processing area, it is possible to analyze whether the aluminum processing area is in a normal working state.

[0063] The personnel gathering characteristic is that aluminum processing workers gather and work in the aluminum processing area. By monitoring whether there is a personnel gathering characteristic in the aluminum processing area, it is possible to analyze whether the aluminum processing area is in a normal working state.

[0064] Refer to Appendix Figure 1 And Figure 2 , in some embodiments, the data analysis step S2 further includes a first analysis step S201. The first analysis step S201 divides multiple monitoring screens based on the video data. The monitoring screens at least include a flame recognition area, a smoke recognition area, a mold erection action recognition area, and a personnel gathering recognition area; and monitors the appearance or disappearance of production behavior characteristics based on the multiple monitoring screens.

[0065] Specifically, the process of extracting production behavior characteristics from video data in the aluminum processing area based on video AI technology is as follows:

[0066] Through RTSP (Real Time Streaming Protocol) or other video streaming protocols, the video data collected by the camera is transmitted to the data processing center in real time.

[0067] On the monitoring interface of the data processing center, design an interactive map or floor plan to display the layout of the aluminum processing area; on the monitoring interface, manually delimit the recognition area through the mouse or touch screen. The recognition area should cover key parts such as mold actions, personnel gatherings, flame and smoke detection, etc.; set a unique identifier and attributes for each recognition area, such as area name, type (mold, personnel, flame / smoke), warning threshold, etc.

[0068] Deploy the trained YOLO (You Only Look Once) algorithm and convolutional neural network (CNN) model on the server of the data processing center; the YOLO algorithm is responsible for identifying the erection or opening actions of the mold, and the CNN model is responsible for detecting abnormal areas such as the number of personnel gatherings, flames, and smoke.

[0069] After the analysis by the YOLO algorithm and the CNN model, the results are output in the form of structured data, and the structured data includes "mold erection", "excessive number of personnel gatherings", "appearance of visible smoke", "appearance of visible flame", etc.

[0070] The structured data is output in the form of JSON (JavaScript Object Notation) or other formats that are easy to parse and store.

[0071] Refer to Appendix Figure 1 And Figure 2 , in some embodiments, the data analysis step S2 further includes a second analysis step S202. In the second analysis step S202, the temperature matrix is analyzed through an image processing algorithm to extract the temperature distribution characteristics of the mold and / or the aluminum liquid flow tank. The temperature distribution characteristics include identifying high-temperature areas, calculating the average temperature, and detecting the temperature gradient; and based on the temperature distribution characteristics, the real-time temperature near the mold and / or the aluminum liquid flow tank is obtained.

[0072] Specifically, thermal imaging technology is used to identify and analyze the temperature data of the aluminum liquid flow tank, the mold surface, and the vicinity of the mold, so as to obtain the real-time temperature, and the temperature state can be analyzed based on the real-time temperature.

[0073] The process of using thermal imaging technology to identify and analyze temperature data to obtain real-time temperature is as follows:

[0074] The infrared thermal imaging camera outputs the collected temperature data in matrix form. Each pixel represents a tiny area in the monitoring region and contains the temperature value of that area.

[0075] Use image processing algorithms to analyze the temperature matrix, and extract the temperature distribution characteristics of the molten aluminum flow tank, the mold surface, and the area near the mold. The temperature distribution characteristics include identifying high-temperature areas, calculating the average temperature, detecting the temperature gradient, etc.; thus, obtain the real-time temperatures of the molten aluminum flow tank, the mold surface, and the area near the mold based on the temperature distribution characteristics.

[0076] In some embodiments, the data analysis step S2 further includes a first analysis step S201, which analyzes the real-time temperature of the aluminum processing area based on the thermal imaging data, and judges whether the real-time temperature of the aluminum processing area exceeds the preset temperature threshold. If so, output the temperature status as the high-temperature status; otherwise, output the temperature status as the safe temperature status.

[0077] Specifically, by setting a preset temperature threshold for the key monitoring area in the aluminum processing area, if the real-time temperature identified by the thermal imaging technology is greater than or equal to the preset temperature threshold, it is analyzed that the temperature status in the aluminum processing area is the high-temperature status; otherwise, the temperature status is the safe temperature status.

[0078] Refer to the appendix Figure 1 , the aluminum processing risk warning method includes a judgment step S3, which judges whether the liquid level of the outlet chute or the liquid level at the inlet of the distribution chute exceeds the preset liquid level, and judges whether the water inlet pressure or the water inlet flow rate of the mold exceeds the water inlet threshold. If both are yes and the start-stop signal of the aluminum processing system is off, then further judge whether there are production behavior characteristics after the mold action or whether the real-time temperature is greater than the preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

[0079] The judgment step is used to cross-judge whether the working state is normal according to the production behavior characteristics, temperature status, and perception data, so as to improve the accuracy of risk warning.

[0080] Refer to the appendix Figure 1 And Figure 3 , in some embodiments, the judgment step S3 further includes a first judgment step S301. After detecting that the start-stop signal of the aluminum processing system stops being sent, judge whether the liquid level of the outlet chute or the liquid level at the inlet of the distribution chute exceeds the preset liquid level. If so, send a first judgment signal to continue the judgment; otherwise, stop the judgment.

[0081] When the data acquisition gateway monitors that the start-stop signal of the aluminum processing system is "1", it means that the start-stop signal of the aluminum processing system is on; when the data acquisition gateway monitors that the start-stop signal of the aluminum processing system is "0", it means that the start-stop signal of the aluminum processing system is off.

[0082] The preset liquid levels include a first preset liquid level and a second preset liquid level. If the data acquisition gateway monitors through a liquid level sensor that the liquid level of the outlet launder exceeds the first preset liquid level, it sends a first judgment signal to continue the judgment; if the data acquisition gateway monitors through a liquid level sensor that the liquid level of the outlet launder does not exceed the first preset liquid level, the judgment is stopped.

[0083] If the data acquisition gateway monitors through a liquid level sensor that the liquid level at the inlet of the distribution launder exceeds the second preset liquid level, it sends a first judgment signal to continue the judgment; if the data acquisition gateway monitors through a liquid level sensor that the liquid level at the inlet of the distribution launder does not exceed the second preset liquid level, the judgment is stopped.

[0084] Reference appendix Figure 1 and Figure 3 , in some embodiments, the judgment step S3 further includes a second judgment step S302. When the first judgment signal is detected, it is judged whether the water inlet pressure or the water inlet flow rate of the mold exceeds the water inlet threshold. If so, a second judgment signal is sent to continue the judgment, otherwise the judgment is stopped.

[0085] The water inlet threshold includes a water inlet pressure threshold and a water inlet flow rate threshold.

[0086] When the first judgment signal is detected, it is judged whether the water inlet pressure of the mold exceeds the water inlet pressure threshold. If the data acquisition gateway obtains through a pressure sensor that the water inlet pressure of the mold exceeds the water inlet pressure threshold, a second judgment signal is sent to continue the judgment; if the data acquisition gateway obtains through a pressure sensor that the water inlet pressure of the mold does not exceed the water inlet pressure threshold, the judgment is stopped.

[0087] When the first judgment signal is detected, it is judged whether the water inlet flow rate of the mold exceeds the water inlet flow rate threshold. If the data acquisition gateway obtains through a water meter that the water inlet flow rate of the mold exceeds the water inlet flow rate threshold, a second judgment signal is sent to continue the judgment; if the data acquisition gateway obtains through a pressure sensor that the water inlet flow rate of the mold does not exceed the water inlet flow rate threshold, the judgment is stopped.

[0088] Reference appendix Figure 1 and Figure 3 , in some embodiments, the judgment step S3 further includes a third judgment step S303. When the second judgment signal is detected, it is judged whether there are characteristics of personnel gathering, visible flames, and visible smoke in the aluminum processing area within a period of time after the mold action. If so, an abnormality occurs, otherwise no abnormality is seen.

[0089] Specifically, when the second judgment signal is detected, first, the actual position of the casting mold in the aluminum processing area is monitored and analyzed through video AI technology. If the casting mold moves to the erected working position, it continues to judge whether production behavior characteristics appear within a preset period of time after the mold action, otherwise the judgment is stopped.

[0090] Then, a preset threshold is set for the number of staff members gathering in the aluminum processing area. The actual number of staff members gathering in the aluminum processing area is monitored and analyzed through video AI technology. If the actual number of staff members is greater than or equal to the preset threshold, it is determined that a production behavior characteristic appears, and an abnormality is output. Otherwise, no abnormality is detected.

[0091] Alternatively, through video AI technology, it is monitored and analyzed whether visible smoke appears in the aluminum processing area. If visible smoke appears in the aluminum processing area, it is determined that a production behavior characteristic appears, and an abnormality is output. Otherwise, no abnormality is detected.

[0092] Alternatively, through video AI technology, it is monitored and analyzed whether visible flames appear in the aluminum processing area. If visible flames appear in the aluminum processing area, it is determined that a production behavior characteristic appears, and an abnormality is output. Otherwise, no abnormality is detected.

[0093] It should be noted that within a preset time period after the mold action, through video AI technology, it is determined whether any of the characteristics of visible flames, visible smoke, and personnel gathering appear in the aluminum processing area. The appearance of any characteristic represents the appearance of a production behavior characteristic, and an abnormality is output.

[0094] Reference appendix Figure 1 And Figure 3 , in some embodiments, the determination step S3 further includes a fourth determination step S304. When the second determination signal is detected, it is determined whether the real-time temperature of the mold and / or the aluminum liquid flow tank is greater than the preset temperature threshold. If so, an abnormality occurs. Otherwise, no abnormality is detected.

[0095] It should be noted that when the second determination signal is detected, when performing the third determination step S303, it is simultaneously determined whether the real-time temperatures of the aluminum liquid flow tank, the mold surface, and the vicinity of the mold are greater than a preset temperature threshold. If so, an abnormality occurs. Otherwise, no abnormality is detected.

[0096] If any output result of the fourth determination step S304 and the third determination step S303 is an abnormality, the output result of the determination step S3 is determined to be an abnormality. Otherwise, no abnormality is detected.

[0097] Reference appendix Figure 1 , in some embodiments, the aluminum processing risk warning method further includes a first alarm step S4. When it is detected that the start / stop signal of the aluminum processing system is off, if the liquid level of the outlet flow tank or the inlet liquid level of the distribution flow tank exceeds the preset liquid level, and the water inlet pressure or water inlet flow rate of the mold exceeds the water inlet threshold, a first alarm signal is sent to the control center.

[0098] Specifically, when the output result of the second judgment step S302 is yes, while sending the second judgment signal, a first alarm signal is sent to the control center for early warning, reminding the staff of the potential risk of illegal operation, reminding manual review, and helping to eliminate potential hazards.

[0099] Reference appendix Figure 1 , in some embodiments, the aluminum processing risk early warning method further includes a second alarm step S5. After detecting the first alarm signal, it further determines whether there are production behavior characteristics after the mold action or whether the real-time temperature is greater than the preset temperature threshold. If so, a second alarm signal is sent to the control center; otherwise, no alarm is issued.

[0100] Specifically, when the output result of the third judgment step S303 or the fourth judgment step S304 is yes, a second alarm signal is sent to the control center for formal alarm, reminding the staff of the major risk of illegal operation and the need for immediate rectification to avoid risks.

[0101] An aluminum processing risk early warning method, which combines video data, thermal imaging data, and perception data to realize the implementation process of the aluminum processing risk early warning method as follows:

[0102] First, three judgment conditions are set. Condition 1 includes:

[0103] Judge whether the liquid level of the outlet launder in the perception data exceeds the first preset liquid level, or judge whether the liquid level at the inlet of the distribution launder exceeds the second preset liquid level, and the judgment result is n1;

[0104] If the liquid level of the outlet launder exceeds the first preset liquid level, or if the liquid level at the inlet of the distribution launder exceeds the second preset liquid level, n1 is true; otherwise, n1 is false.

[0105] Judge whether the water inlet flow rate of the mold in the perception data exceeds the water inlet flow rate threshold, or judge whether the water inlet pressure of the mold exceeds the water inlet pressure threshold, and the judgment result is n2;

[0106] If the water inlet flow rate of the mold exceeds the water inlet flow rate threshold, or if the water inlet pressure of the mold exceeds the water inlet pressure threshold, n2 is true; otherwise, n2 is false.

[0107] Judge whether the start-stop signal of the aluminum processing system in the perception data is off or on, and the judgment result is n3;

[0108] If the start-stop signal of the aluminum processing system is "1", representing on, n3 is true; if the start-stop signal of the aluminum processing system is "0", representing off, n3 is false;

[0109] First, take the logical "AND" relationship of n1 and n2 to obtain z1;

[0110] If either n1 or n2 is false, then z1 is false; otherwise, z1 is true.

[0111] On the premise that n3 is false, then judge the truth value of z1 as the result value z2 of Condition 1.

[0112] If z1 is false, then z2 is false; otherwise, if z1 is true, then z2 is true.

[0113] Condition 2: Starting from the detection of the on-site mold movement through video data analysis as the time starting point, within t minutes, detect whether there is visible smoke, visible flame, or the feature of more than 2 people gathering in the video, and the judgment result is used as the result value z3 of Condition 2.

[0114] Within t minutes after the mold movement, if visible smoke appears, or visible flame appears, or there is a gathering of more than 2 people, then z3 is true; otherwise, z3 is false.

[0115] Condition 3: Through thermal imaging, detect whether the temperature in the area on the mold surface or in the aluminum liquid chute reaches or exceeds the preset temperature threshold T; the judgment result is used as the result value z4 of Condition 3.

[0116] If the temperature in the area on the mold surface or in the aluminum liquid chute reaches or exceeds the preset temperature threshold T, then z4 is true; otherwise, z4 is false.

[0117] The model algorithm is based on the following:

[0118] Take the logical "OR" relationship between z3 and z4 to get the result z5.

[0119] If either z3 or z4 is true, then z5 is true; otherwise, z5 is false.

[0120] On the premise that z2 is true, send the first alarm signal to the control center and judge whether Z5 is true to get the result z6.

[0121] If z5 is true, then z6 is true and send the second alarm signal to the control center; otherwise, z6 is false and no alarm is required.

[0122] The steps of the algorithm model are as follows:

[0123] Step 1: Process Condition 1 to obtain z2.

[0124] Step 2: Process Condition 2 to detect visible smoke, visible flame, and the feature of more than 2 people gathering in the video within t minutes to obtain z3.

[0125] Step 3: Process Condition 3 to detect the temperature in real time to obtain z4.

[0126] Step 4, calculate z5 = z3 OR z4.

[0127] Step 5, based on z2 being false, send a first alarm signal to the control center, and determine whether Z5 is true to obtain a result of z6; if z5 is true, then z6 is true, and send a second alarm signal to the control center, otherwise do not alarm.

[0128] During the algorithm model establishment process, the variable definitions are as follows:

[0129] last_mold_action_time: Record the time of the most recent mold action, initially null;

[0130] hazard_detected: A boolean value, initially False, indicating whether a hazard is detected within the time window;

[0131] z3: A boolean value, the result of condition two;

[0132] z4: A boolean value, the result of condition three;

[0133] z2: The result of condition one;

[0134] The algorithm process is as follows:

[0135] When the video analysis detects a mold action, update last_mold_action_time to the current time and reset hazard_detected to False.

[0136] Process condition one, obtain the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the inlet flow rate of the mold, and the inlet pressure of the mold through the data acquisition gateway;

[0137] n1 = (outlet launder liquid level ≥ set value) OR (distribution launder inlet liquid level ≥ set value);

[0138] n2 = (mold inlet flow rate ≥ set value) OR (mold inlet pressure ≥ set value);

[0139] z1 = n1 AND n2;

[0140] z2 = z1 AND (NOT n3);

[0141] Process condition 2. If last_mold_action_time is not null and the current time ≤ last_mold_action_time + t minutes; when within the time window after the mold action, check whether there are visual smoke, visual flame, and the feature of more than 2 people gathering in the video data; if so, hazard_detected = True, z3 = hazard_detected, otherwise: z3 = False.

[0142] Process condition 3. Obtain the temperature data of the mold surface and the molten aluminum flow trough area, z4 = (mold surface temperature ≥ T) OR (molten aluminum flow trough temperature ≥ T).

[0143] Calculate z5 = z3 OR z4; calculate z6 = z5 AND z2.

[0144] The alarm logic is as follows:

[0145] If z2 is True, output the first alarm signal, and if z5 is also True, output the second alarm signal; otherwise, do not alarm.

[0146] This application provides an aluminum processing risk warning method. By analyzing video data and thermal imaging data to obtain production behavior characteristics and temperature states, and based on production behavior characteristics, temperature states, and perception data to cross-judge whether the working state is normal, thereby improving the accuracy of risk warning; through the first alarm step to give a warning when the outlet trough liquid level, distribution trough inlet liquid level, crystallizer inlet water pressure, crystallizer inlet water flow in the perception data do not match the aluminum processing system start-stop signal; through the second alarm step to give an alarm after further judging that the production behavior characteristics or the real-time temperature is abnormal; through the first judgment step and the second judgment step to judge whether to continue the judgment, thereby streamlining the judgment process; through the third judgment step to judge whether production behavior characteristics appear, thereby judging whether production anomalies occur; through the fourth judgment step to judge whether the real-time temperature is higher than the preset temperature threshold, thereby judging whether production anomalies occur; through the first analysis step to analyze the video data, thereby monitoring the appearance or disappearance of production behavior characteristics; through the second analysis step to analyze the thermal imaging data, thereby obtaining the real-time temperature near the mold and / or the molten aluminum flow trough. Specific Embodiment 2

[0148] This application also provides an aluminum processing risk warning system for correspondingly implementing the aluminum processing risk warning method in the above Specific Embodiment 1. The aluminum processing risk warning system includes a data acquisition module, a data analysis module, and a judgment module.

[0149] Among them, the data acquisition module is used to collect video data and thermal imaging data of the aluminum processing area through a camera, and obtain perception data through a data acquisition gateway. The perception data includes the start / stop signal of the aluminum processing system, the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the water inlet pressure of the mold, and the water inlet flow rate of the mold.

[0150] The data analysis module is used to analyze the production behavior characteristics based on the video data. The production behavior characteristics include personnel aggregation, visible flame, and visible smoke; and monitor the real-time temperature of the mold and / or the aluminum liquid launder based on the thermal imaging data.

[0151] The judgment module is used to judge whether the liquid level of the outlet launder or the liquid level at the inlet of the distribution launder exceeds the preset liquid level, and judge whether the water inlet pressure or the water inlet flow rate of the mold exceeds the water inlet threshold. If both are yes and the start / stop signal of the aluminum processing system is off, then further judge whether there are production behavior characteristics after the mold action or whether the real-time temperature is greater than the preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

[0152] In some embodiments, the aluminum processing risk warning system further includes a first alarm module, which is used to send a first alarm signal to the control center when it is detected that the start / stop signal of the aluminum processing system is off, if the liquid level of the outlet launder or the liquid level at the inlet of the distribution launder exceeds the preset liquid level, and the water inlet pressure or the water inlet flow rate of the mold exceeds the water inlet threshold.

[0153] In some embodiments, the aluminum processing risk warning system further includes a second alarm module, which is used to further judge whether there are production behavior characteristics after the mold action or whether the real-time temperature is greater than the preset temperature threshold after detecting the first alarm signal. If so, a second alarm signal is sent to the control center; otherwise, no alarm is issued.

[0154] Finally, it should be noted that: the various embodiments in this specification are described in a progressive manner, and the key points of each embodiment are the differences from other embodiments. The same and similar parts among the various embodiments can be referred to each other.

[0155] The above embodiments are only used to illustrate the technical solutions of the present application and not to limit them; although the present application has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that: still can modify the specific implementation manners of the present application or perform equivalent replacements on some technical features; without departing from the spirit of the technical solutions of the present application, they should all be covered within the scope of the technical solutions claimed in the present application.

Claims

1. An aluminum processing risk warning method, characterized in that, Including: A data collection step, where video data and thermal imaging data of the aluminum processing area are collected through a camera, and perception data is obtained through a data acquisition gateway. The perception data includes the start / stop signal of the aluminum processing system, the liquid level of the outlet launder, the liquid level at the inlet of the distribution launder, the water inlet pressure of the mold, and the water inlet flow rate of the mold. A data analysis step, where production behavior characteristics are analyzed based on the video data. The production behavior characteristics include personnel gathering, visible flames, and visible smoke. Monitoring the real-time temperature of the mold and / or the aluminum liquid launder based on the thermal imaging data. A judgment step, judging whether the liquid level of the outlet launder or the liquid level at the inlet of the distribution launder exceeds a preset liquid level, and judging whether the water inlet pressure of the mold or the water inlet flow rate of the mold exceeds a water inlet threshold. If both are yes and the start / stop signal of the aluminum processing system is off, then further judge whether there are the production behavior characteristics after the mold action or whether the real-time temperature is greater than a preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

2. The aluminum processing risk warning method according to claim 1, wherein Also including: A first alarm step. When it is detected that the start / stop signal of the aluminum processing system is off, if the liquid level of the outlet launder or the liquid level at the inlet of the distribution launder exceeds the preset liquid level, and the water inlet pressure of the mold or the water inlet flow rate of the mold exceeds the water inlet threshold, a first alarm signal is sent to the control center.

3. The aluminum processing risk warning method according to claim 2, wherein Also including: A second alarm step. After detecting the first alarm signal, further judge whether there are the production behavior characteristics after the mold action or whether the real-time temperature is greater than a preset temperature threshold. If so, a second alarm signal is sent to the control center; otherwise, no alarm is issued.

4. The aluminum processing risk warning method according to claim 1, wherein The judgment step further includes: A first judgment step. When it is monitored that the start / stop signal of the aluminum processing system stops being sent, judge whether the liquid level of the outlet launder or the liquid level at the inlet of the distribution launder exceeds the preset liquid level. If so, send a first judgment signal to continue the judgment; otherwise, stop the judgment.

5. The aluminum processing risk warning method according to claim 4, characterized in that, The judgment step further includes: A second judgment step. When the first judgment signal is detected, judge whether the water inlet pressure of the mold or the water inlet flow rate of the mold exceeds the water inlet threshold. If so, send a second judgment signal to continue the judgment; otherwise, stop the judgment.

6. The aluminum processing risk warning method according to claim 5, wherein, The judgment step further includes: A third judgment step. When the second judgment signal is detected, judge whether there are characteristics of personnel gathering, visible flames, and visible smoke in the aluminum processing area within a period of time after the mold action. If so, an abnormality occurs; otherwise, no abnormality is found.

7. The aluminum processing risk warning method according to claim 5, wherein, The judgment step further includes: A fourth judgment step. When the second judgment signal is detected, judge whether the real-time temperature of the mold and / or the aluminum liquid launder is greater than a preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

8. The aluminum processing risk warning method according to claim 1, characterized in that, The data analysis step further includes: A first analysis step. Based on the video data, multiple monitoring screens are divided. The monitoring screens at least include a flame recognition area, a smoke recognition area, a mold erection action recognition area, and a personnel gathering recognition area; and based on the multiple monitoring screens, the appearance or disappearance of the production behavior characteristics is monitored.

9. The aluminum processing risk warning method according to claim 1, wherein, The data analysis step further includes: The second analysis step is to analyze the temperature matrix through an image processing algorithm to extract the temperature distribution characteristics of the mold and / or the aluminum liquid chute. The temperature distribution characteristics include identifying high-temperature areas, calculating the average temperature, and detecting the temperature gradient; and obtaining the real-time temperature near the mold and / or the aluminum liquid chute based on the temperature distribution characteristics.

10. An aluminum processing risk warning system, characterized in that, It includes: A data acquisition module, which is used to collect video data and thermal imaging data of the aluminum processing area through a camera, and obtain sensing data through a data acquisition gateway. The sensing data includes the start-stop signal of the aluminum processing system, the liquid level of the outlet chute, the liquid level at the inlet of the distribution chute, the water inlet pressure of the mold, and the water inlet flow rate of the mold. A data analysis module, which is used to analyze the production behavior characteristics based on the video data. The production behavior characteristics include personnel gathering, visible flames, and visible smoke. Monitoring the real-time temperature of the mold and / or the aluminum liquid chute based on the thermal imaging data; A judgment module, which is used to judge whether the liquid level of the outlet chute or the liquid level at the inlet of the distribution chute exceeds the preset liquid level, and judge whether the water inlet pressure of the mold or the water inlet flow rate of the mold exceeds the water inlet threshold. If both are yes and the start-stop signal of the aluminum processing system is off, then further judge whether there are the production behavior characteristics after the mold action or whether the real-time temperature is greater than the preset temperature threshold. If so, an abnormality occurs; otherwise, no abnormality is found.

Images