A monitoring and alarm system and method for aluminum processing diverter plate aluminum liquid leakage

Abstract

The present invention provides a monitoring alarm system and method for aluminum processing shunt plate aluminum liquid leakage, the monitoring alarm system includes a liquid surface image acquisition device, used to obtain the aluminum liquid surface image on the shunt plate in real time; a liquid surface image analysis module, used to perform real-time image recognition and analysis on the aluminum liquid surface image, and send a liquid surface warning signal to the alarm device when abnormal fluctuations in the aluminum liquid surface are detected; an infrared thermal imaging device, used to obtain an infrared thermal image of the space below the shunt plate in real time; an infrared thermal image analysis module, used to convert the infrared thermal image into temperature data in real time, and send a temperature warning signal to the alarm device when the temperature in the infrared thermal image exceeds a set threshold; and an alarm device, used to send an alarm signal when receiving a liquid surface warning signal or a temperature warning signal. The present invention adopts a dual-linkage alarm technology based on liquid surface fluctuation detection and infrared thermal imaging detection, providing double insurance for production safety.

Description

Technical Field

[0001] The present invention relates to the field of industrial production safety, and in particular to a monitoring and alarming system and method for aluminum liquid leakage in an aluminum processing diverter plate. Background Art

[0002] In recent years, accidents have occurred frequently in the aluminum processing industry, exposing many problems in production safety. Among them, aluminum liquid leakage is a high-risk point in the deep well casting process. Once high-temperature aluminum liquid leaks and improperly contacts with cooling water, it will explode.

[0003] At present, the aluminum processing industry has a low level of informatization in production safety monitoring, and production safety warnings are mostly experience-oriented. According to the survey, since the safety hazards in the production process rely heavily on human monitoring and judgment, the main cause of aluminum liquid leakage accidents is the failure to timely discover aluminum liquid leakage due to unauthorized staff leaving their posts or misjudgment due to lack of experience.

[0004] Therefore, it is imperative to adopt more digital and intelligent means to upgrade the technology of production safety monitoring and accident warning in the aluminum processing industry. Summary of the invention

[0005] The purpose of the present invention is to provide a monitoring and alarm system and method for aluminum liquid leakage in aluminum processing diverter disks, so as to solve the problem of heavy reliance on manual discovery of potential safety hazards in the aluminum processing process and improve the level of production safety.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A monitoring and alarm system for aluminum liquid leakage in an aluminum processing diverter plate, comprising:

[0008] A liquid level image acquisition device is disposed above the diverter plate and toward the upper surface of the diverter plate, and is used to acquire a liquid level image of the aluminum liquid on the diverter plate in real time;

[0009] The liquid level image analysis module is used to perform real-time image recognition and analysis on the aluminum liquid level image, and send a liquid level warning signal to the alarm device when abnormal fluctuations in the aluminum liquid level are detected;

[0010] An infrared thermal imaging device is disposed below the diverter plate and toward the bottom surface of the diverter plate, and is used to obtain an infrared thermal image of the space below the diverter plate in real time;

[0011] Infrared thermal image analysis module, used to convert infrared thermal images into temperature data in real time, and send a temperature warning signal to the alarm device when the temperature in the infrared thermal image exceeds a set threshold;

[0012] The alarm device is used to send out an alarm signal when receiving a liquid level warning signal or a temperature warning signal.

[0013] Preferably, the alarm device sends out a first alarm signal when it only receives a liquid level warning signal; the alarm device sends out a second alarm signal when it only receives a temperature warning signal; and the alarm device sends out a third alarm signal when it receives both a liquid level warning signal and a temperature warning signal.

[0014] Preferably, the liquid surface image analysis module and the infrared thermal image analysis module are arranged in a cloud server.

[0015] Preferably, in the liquid level image analysis module, a trained liquid level abnormality recognition model is used to perform real-time image recognition analysis on the aluminum liquid level image, and when a vortex or bubble is detected on the aluminum liquid level, a liquid level warning signal is sent to the alarm device;

[0016] The liquid level abnormality recognition model is obtained by the following method:

[0017] Based on the YOLOv4 network model, the number of classifications of the model is set to 2, corresponding to vortexes and bubbles respectively; the feature extraction network CSPdarknet is replaced by the ShuffleNet network; the ordinary convolution in the enhanced feature network PAnet is replaced by the depthwise separable convolution.

[0018] Preferably, the training process of the liquid level anomaly recognition model includes:

[0019] S101, establishing a data set: collecting aluminum liquid surface images containing vortices or bubbles in the liquid surface through a liquid surface image acquisition device, and preprocessing the collected images to establish an aluminum liquid surface fluctuation data set;

[0020] S102, model training: using the data set established in S101 to train a liquid level anomaly recognition model;

[0021] S103, detection effect evaluation: performing detection effect evaluation on the trained liquid level anomaly recognition model;

[0022] S104, model deployment: deploy the liquid level anomaly recognition model that has passed the detection effect evaluation into the liquid level image analysis module to achieve real-time detection of abnormal liquid level fluctuations.

[0023] Preferably, the S101 specifically includes:

[0024] S1011, collecting a liquid surface image of aluminum liquid containing vortices or bubbles on the liquid surface by a liquid surface image acquisition device;

[0025] S1012, performing grayscale conversion on the acquired image;

[0026] S1013, performing data expansion on the image by data enhancement, and forming an image set after the expansion;

[0027] S1014, annotate the positions of the vortices and bubbles in each image in the image set, record the coordinates of the vortices and bubbles in the image, and make a data set together with the image set.

[0028] Preferably, in S103, accuracy, recall rate, and mAP are used as evaluation indicators to evaluate the detection effect of the trained liquid level anomaly recognition model.

[0029] Preferably, in S103, FPS is also used as an evaluation index to evaluate the detection effect of the trained liquid level anomaly recognition model.

[0030] Preferably, it also includes a security management platform; when the alarm device sends out an alarm signal, it also sends alarm information to the security management platform.

[0031] A monitoring and alarm method for aluminum liquid leakage in an aluminum processing diverter plate, comprising:

[0032] A liquid surface image acquisition device is arranged above the diverter plate, and the liquid surface image acquisition device is arranged toward the upper surface of the diverter plate;

[0033] An infrared thermal imaging device is arranged below the diverter plate, and the infrared thermal imaging device is arranged toward the bottom surface of the diverter plate;

[0034] The liquid level image of the aluminum liquid in the diverter tray is acquired in real time by a liquid level image acquisition device, and the liquid level image analysis module performs real-time image recognition and analysis on the liquid level image of the aluminum liquid. When abnormal fluctuations in the aluminum liquid level are detected, a liquid level warning signal is sent to the alarm device;

[0035] The infrared thermal image of the space below the diverter plate is acquired in real time through the infrared thermal imaging device, and the infrared thermal image is converted into temperature data in real time through the infrared thermal image analysis module. When the temperature in the infrared thermal image exceeds the set threshold, a temperature warning signal is sent to the alarm device;

[0036] When the alarm device receives a liquid level warning signal or a temperature warning signal, it sends out an alarm signal.

[0037] The present invention provides a monitoring and alarm system and method for aluminum liquid leakage in aluminum processing diverter plates, which adopts a dual-linkage alarm technology based on liquid level fluctuation detection and infrared thermal imaging detection, providing double insurance for production safety. At the same time, the present invention uses artificial intelligence technology to monitor whether aluminum liquid leakage occurs during the casting process in a real-time video monitoring manner, thereby enhancing the perception, monitoring, and early warning capabilities of safe production, accelerating the transformation of safe production from static analysis to dynamic perception, and from post-emergency to pre-prevention, and improving the level of production safety. In addition, the present invention can upgrade and transform the monitoring and alarm system without changing the existing production site, and can even achieve transformation without stopping the machine, minimizing the impact of the technology upgrade process on existing production equipment and production processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a structural schematic diagram of a monitoring and alarm system for aluminum liquid leakage in an aluminum processing diverter plate provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0039] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0040] Embodiment 1

[0041] like Figure 1 As shown, an embodiment of the present invention provides a monitoring and alarm system for aluminum liquid leakage in an aluminum processing diverter plate, comprising:

[0042] A liquid level image acquisition device is disposed above the diverter plate and toward the upper surface of the diverter plate, and is used to acquire a liquid level image of the aluminum liquid on the diverter plate in real time;

[0043] The liquid level image analysis module is used to perform real-time image recognition and analysis on the aluminum liquid level image, and send a liquid level warning signal to the alarm device when abnormal fluctuations in the aluminum liquid level are detected; the specific manifestation of abnormal fluctuations in the aluminum liquid level is the appearance of vortices or bubbles on the aluminum liquid level;

[0044] An infrared thermal imaging device is disposed below the diverter plate and toward the bottom surface of the diverter plate, and is used to obtain an infrared thermal image of the space below the diverter plate in real time;

[0045] The infrared thermal image analysis module is used to convert the infrared thermal image into temperature data in real time, and send a temperature warning signal to the alarm device when the temperature in the infrared thermal image exceeds a set threshold; wherein the set threshold is preferably 600°C;

[0046] The alarm device is used to send out an alarm signal when receiving a liquid level warning signal or a temperature warning signal. The alarm device sends out different alarm signals according to the different warning signals received. Specifically, the alarm device sends out a first alarm signal when only receiving a liquid level warning signal; the alarm device sends out a second alarm signal when only receiving a temperature warning signal; and the alarm device sends out a third alarm signal when receiving both a liquid level warning signal and a temperature warning signal.

[0047] In this embodiment, the liquid surface image acquisition device is a high-definition AI camera, the infrared thermal imaging device is a binocular infrared camera; and the alarm device is an audible and visual alarm. In addition, the number of the liquid surface image acquisition device, the infrared thermal imaging device, and the alarm device can be one or more, respectively. By setting a plurality of liquid surface image acquisition devices and infrared thermal imaging devices, and distributing them in different orientations, the aluminum liquid surface image and infrared thermal image can be acquired from different angles, thereby avoiding the problem of untimely alarm caused by a single camera being blocked or misidentified. Among them, in order to ensure timely monitoring of the temperature changes in the space below the diverter plate, the infrared thermal imaging device is distributed at the diagonal below the diverter plate, and its monitoring area includes the space below the diverter plate and the aluminum rods arranged below the diverter plate.

[0048] In the actual casting process, because the aluminum liquid surface on the diverter plate is wide, the aluminum liquid temperature at the far end of the diverter plate is lower than that at the near end, making the aluminum liquid crystallization speed (falling speed) at the far end faster than that at the near end, resulting in the risk of aluminum liquid leakage at the far end, which is characterized by vortices on the aluminum liquid surface. In addition, once aluminum liquid leaks from the crystallization hole, bubbles will appear on the aluminum liquid surface in the crystallization hole.

[0049] At the same time, when the aluminum liquid leaks to the bottom of the diverter plate, the space below the diverter plate will experience a sudden rise in temperature. Generally speaking, the temperature of the aluminum liquid is around 650°C, while the temperature of the aluminum rod below the diverter plate is below 500°C. Therefore, when the temperature of the space below the diverter plate rises rapidly to above 600°C, it can be determined that aluminum liquid has leaked.

[0050] Based on the above situation, the present invention adopts a double linkage alarm mechanism of the upper and lower diverter disc, identifies whether there is a vortex or bubble on the liquid surface above the diverter disc through image recognition technology, and monitors whether there is an abnormality in the temperature field of the space below the diverter disc through infrared temperature measurement technology, so as to determine whether aluminum liquid leakage occurs. No matter which abnormal situation occurs, the alarm device will be triggered to send an alarm signal. Specifically, when a vortex or bubble is identified on the liquid surface above the diverter disc, the alarm device sends a first alarm signal; when the temperature of the space below the diverter disc is abnormal, the alarm device sends a second alarm signal; when the aluminum liquid above the diverter disc and the temperature of the space below the diverter disc are detected at the same time, the alarm device sends a third alarm signal. The urgency and danger procedures of the first alarm signal, the second alarm signal and the third alarm signal increase in sequence, reminding the staff to adopt different emergency treatment plans to prevent production accidents before they happen. At the same time, different indicator light signals or sound signals can be used to distinguish the first alarm signal, the second alarm signal and the third alarm signal, so that the on-site staff can quickly know the danger level of the current hidden danger.

[0051] As an improvement, the liquid level image analysis module and the infrared thermal image analysis module can be set in the cloud server. The liquid level image acquisition device and the infrared thermal imaging device upload the aluminum liquid level image and the infrared thermal image acquired from the production site to the cloud server in real time. The video image content can be read at any time in the cloud server, and the liquid level image analysis and the infrared thermal image analysis can be performed by the liquid level image analysis module and the infrared thermal image analysis module respectively. The liquid level warning signal and the temperature warning signal issued by the liquid level image analysis module and the infrared thermal image analysis module are transmitted from the cloud server back to the alarm device at the production site to realize the alarm process.

[0052] In the liquid level image analysis module of the embodiment of the present invention, a trained liquid level anomaly recognition model is used to perform real-time image recognition analysis on the aluminum liquid level image, and when a vortex or bubble is detected on the aluminum liquid level, a liquid level warning signal is sent to the alarm device.

[0053] Among them, the liquid level anomaly recognition model is obtained by the following method: based on the YOLOv4 network model, the number of classifications of the model is set to 2, corresponding to vortices and bubbles respectively; the feature extraction network CSPdarknet is replaced by the ShuffleNet network; the ordinary convolution in the enhanced feature network PAnet is replaced by a depth-separable convolution.

[0054] Since the YOLOv4 network model in the prior art is relatively large, complex in structure, and has a large amount of calculation, it has high requirements on the hardware of the computing device and cannot meet the requirements of real-time detection; therefore, the present invention forms a liquid level anomaly recognition model after lightweight improvement of the YOLOv4 network model, so that the detection speed is improved while maintaining the detection accuracy of the original model, so as to ensure the real-time performance of safety detection at the production site, so that it can detect abnormal situations in time so as to issue an alarm in time.

[0055] Specifically, the YOLOv4 network model adds a residual block to the feature extraction network CSPdarknet to improve the detection accuracy, but this also increases the computational burden of the model; the Sufflenet network is a model with a small number of parameters, and compared with the feature extraction network CSPdarknet, it does not reduce too much recognition accuracy; therefore, the present invention uses the Sufflenet network instead of CSPdarknet for feature extraction. In addition, the YOLOv4 network model uses ordinary convolution, which has a huge number of parameters due to the complexity of the network; therefore, the present invention uses deep separable convolution to replace ordinary convolution to reduce the number of parameters of the original network and further improve the detection speed of the network.

[0056] Furthermore, the training process of the liquid level abnormality recognition model includes:

[0057] S101, establishing a data set: collecting aluminum liquid surface images containing vortices or bubbles in the liquid surface through a liquid surface image acquisition device, and preprocessing the collected images to establish an aluminum liquid surface fluctuation data set;

[0058] S102, model training: using the data set established in S101 to train a liquid level anomaly recognition model;

[0059] S103, detection effect evaluation: performing detection effect evaluation on the trained liquid level anomaly recognition model;

[0060] S104, model deployment: deploy the liquid level anomaly recognition model that has passed the detection effect evaluation into the liquid level image analysis module to achieve real-time detection of abnormal liquid level fluctuations.

[0061] Furthermore, the S101 specifically includes:

[0062] S1011, collecting a liquid surface image of aluminum liquid containing vortices or bubbles on the liquid surface by a liquid surface image acquisition device;

[0063] S1012, performing grayscale conversion on the acquired image, so as to obtain a bubble or vortex image with clear details and high contrast;

[0064] S1013, performing data augmentation on the image to form an image set after the augmentation; wherein the specific operation of the data augmentation is to flip the image left-right, up-down, and symmetrically, so as to increase the number of limited image samples in the image set;

[0065] S1014, annotate the positions of the vortices and bubbles in each image in the image set, record the coordinates of the vortices and bubbles in the image, and make a data set together with the image set.

[0066] Furthermore, in S103, the accuracy, recall, mAP, and FPS are used as evaluation indicators to evaluate the detection effect of the trained liquid level anomaly recognition model.

[0067] Among them, the accuracy is used to measure the accuracy of model prediction, and its formula is: Precision = TP / (TP+FP);

[0068] Recall is used to measure the ability of the model to detect positive samples, and its formula is: Recall = TP / (TP+FN);

[0069] In the formula, TP represents the number of positive samples detected as positive samples by the model, FP represents the number of negative samples detected as positive samples by the model, and FN represents the number of positive samples detected as negative samples by the model.

[0070] mAP (Mean Average Percision) is used to evaluate the performance of the model in all categories. In this embodiment, mAP is equal to the average AP of the vortex and bubble categories. AP (Average Precision) is used to evaluate the performance of the model in a single category (vortex or bubble), and represents the area enclosed by the PR curve and the coordinate axis; the PR curve is a curve showing the relationship between recall and accuracy in a coordinate system with recall as the horizontal axis and accuracy as the vertical axis.

[0071] FPS is used to evaluate the recognition rate of the model, which indicates the number of image frames that the model can recognize per second.

[0072] In S103 of this embodiment, when evaluating the detection effect of the trained liquid level anomaly recognition model, the accuracy and recall rate should be greater than or equal to 85%, and the mAP should be greater than or equal to 80%; when the accuracy and recall rate of the trained liquid level anomaly recognition model are less than 85%, or the mAP is less than 80%, the hyperparameters of the liquid level anomaly recognition model are adjusted and the method returns to S102 for retraining.

[0073] In contrast, FPS does not have a clear evaluation value, and the larger the value, the better. FPS can reflect the real-time performance of the model in image recognition and is used to compare the image recognition speed of different training results. When choosing from multiple different training results, the model with a larger FPS is generally selected.

[0074] As an improvement, the embodiment of the present invention further includes a safety management platform; when the alarm device sends an alarm signal, it also sends alarm information to the safety management platform. The safety management platform is used to record the time and specific circumstances of the alarm, and at the same time transmit the alarm information to the person in charge of the relevant production department by remote communication.

[0075] Embodiment 2

[0076] The embodiment of the present invention provides a monitoring and alarm method for aluminum liquid leakage in an aluminum processing diverter plate, comprising:

[0077] A liquid surface image acquisition device is arranged above the diverter plate, and the liquid surface image acquisition device is arranged toward the upper surface of the diverter plate;

[0078] An infrared thermal imaging device is arranged below the diverter plate, and the infrared thermal imaging device is arranged toward the bottom surface of the diverter plate;

[0079] The liquid level image of the aluminum liquid in the diverter tray is acquired in real time by a liquid level image acquisition device, and the liquid level image analysis module performs real-time image recognition and analysis on the liquid level image of the aluminum liquid. When abnormal fluctuations in the aluminum liquid level are detected, a liquid level warning signal is sent to the alarm device;

[0080] The infrared thermal image of the space below the diverter plate is acquired in real time through the infrared thermal imaging device, and the infrared thermal image is converted into temperature data in real time through the infrared thermal image analysis module. When the temperature in the infrared thermal image exceeds the set threshold, a temperature warning signal is sent to the alarm device;

[0081] When the alarm device receives a liquid level warning signal or a temperature warning signal, it sends out an alarm signal.

[0082] The monitoring and alarm method provided in this embodiment corresponds to the monitoring and alarm system in the first embodiment, and the working process of each functional module used therein is also the same as that in the first embodiment, which will not be repeated here.

[0083] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A monitoring and alarm system for aluminum liquid leakage in aluminum processing diverter plate, characterized in that: include: A liquid level image acquisition device is disposed above the diverter plate and toward the upper surface of the diverter plate, and is used to acquire a liquid level image of the aluminum liquid on the diverter plate in real time; The liquid level image analysis module is used to perform real-time image recognition and analysis on the aluminum liquid level image, and send a liquid level warning signal to the alarm device when abnormal fluctuations in the aluminum liquid level are detected; An infrared thermal imaging device is disposed below the diverter plate and toward the bottom surface of the diverter plate, and is used to obtain an infrared thermal image of the space below the diverter plate in real time; Infrared thermal image analysis module, used to convert infrared thermal images into temperature data in real time, and send a temperature warning signal to the alarm device when the temperature in the infrared thermal image exceeds a set threshold; An alarm device, used to send out an alarm signal when receiving a liquid level warning signal or a temperature warning signal; In the liquid level image analysis module, a trained liquid level abnormality recognition model is used to perform real-time image recognition and analysis on the aluminum liquid level image, and when a vortex or bubble is detected on the aluminum liquid level, a liquid level warning signal is sent to the alarm device; The liquid level abnormality recognition model is obtained by the following method: Based on the YOLOv4 network model, the number of model classifications is set to 2, corresponding to vortexes and bubbles respectively; the feature extraction network CSPdarknet is replaced by the ShuffleNet network; the ordinary convolution in the enhanced feature network PAnet is replaced by the depthwise separable convolution; The training process of the liquid level anomaly recognition model includes: S101, establishing a data set: collecting aluminum liquid surface images containing vortices or bubbles in the liquid surface through a liquid surface image acquisition device, and preprocessing the collected images to establish an aluminum liquid surface fluctuation data set; S102, model training: using the data set established in S101 to train a liquid level anomaly recognition model; S103, detection effect evaluation: performing detection effect evaluation on the trained liquid level anomaly recognition model; S104, model deployment: deploy the liquid level anomaly recognition model that has passed the detection effect evaluation into the liquid level image analysis module to achieve real-time detection of abnormal liquid level fluctuations.

2. The monitoring and alarm system for aluminum liquid leakage in aluminum processing diverter disk according to claim 1 is characterized in that: When the alarm device receives only a liquid level warning signal, it sends out a first alarm signal; when the alarm device receives only a temperature warning signal, it sends out a second alarm signal; when the alarm device receives both a liquid level warning signal and a temperature warning signal, it sends out a third alarm signal.

3. The monitoring and alarm system for aluminum liquid leakage in aluminum processing diverter disk according to claim 1 is characterized in that: The liquid surface image analysis module and the infrared thermal image analysis module are arranged in a cloud server.

4. The monitoring and alarm system for aluminum liquid leakage in aluminum processing diverter disk according to claim 1 is characterized in that: The S101 specifically includes: S1011, collecting a liquid surface image of aluminum liquid containing vortices or bubbles on the liquid surface by a liquid surface image acquisition device; S1012, performing grayscale conversion on the acquired image; S1013, performing data expansion on the image by data enhancement, and forming an image set after the expansion; S1014, annotate the positions of the vortices and bubbles in each image in the image set, record the coordinates of the vortices and bubbles in the image, and make a data set together with the image set.

5. The monitoring and alarm system for aluminum liquid leakage in aluminum processing diverter disk according to claim 1 is characterized in that: In S103, the accuracy, recall rate, and mAP are used as evaluation indicators to evaluate the detection effect of the trained liquid level anomaly recognition model.

6. The monitoring and alarm system for aluminum liquid leakage in aluminum processing diverter disk according to claim 5 is characterized in that: In S103, FPS is also used as an evaluation index to evaluate the detection effect of the trained liquid level anomaly recognition model.

7. The monitoring and alarm system for aluminum liquid leakage in aluminum processing diverter disk according to claim 1 is characterized in that: It also includes a security management platform; when the alarm device sends out an alarm signal, it also sends alarm information to the security management platform.

8. A monitoring and alarm method for aluminum liquid leakage in aluminum processing diverter plate, based on the monitoring and alarm system according to any one of claims 1 to 7, characterized in that: include: A liquid surface image acquisition device is arranged above the diverter plate, and the liquid surface image acquisition device is arranged toward the upper surface of the diverter plate; An infrared thermal imaging device is arranged below the diverter plate, and the infrared thermal imaging device is arranged toward the bottom surface of the diverter plate; The liquid level image of the aluminum liquid in the diverter tray is acquired in real time by a liquid level image acquisition device, and the liquid level image analysis module performs real-time image recognition and analysis on the liquid level image of the aluminum liquid. When abnormal fluctuations in the aluminum liquid level are detected, a liquid level warning signal is sent to the alarm device; The infrared thermal image of the space below the diverter plate is acquired in real time through the infrared thermal imaging device, and the infrared thermal image is converted into temperature data in real time through the infrared thermal image analysis module. When the temperature in the infrared thermal image exceeds the set threshold, a temperature warning signal is sent to the alarm device; When the alarm device receives a liquid level warning signal or a temperature warning signal, it sends out an alarm signal.

Images