Power supply and distribution abnormal value disposal method for metallurgical iron and steel plant

Abstract

The invention discloses a power supply and distribution abnormal value disposal method for a metallurgical steel plant, and the method comprises the steps: continuously detecting the power supply and distribution data of the metallurgical steel plant in real time through an advanced data collection module, and carrying out the precise comparison with a preset standard fluctuation interval, thereby rapidly recognizing an abnormal value. When an abnormal value occurs, the data processing module is started immediately, the influence of external factors (such as weather conditions, manual operation and animal interference) and equipment faults on the power supply and distribution system is analyzed deeply, and corresponding treatment measures are generated accordingly. If an abnormal value is caused by external factors and does not cause long-term influence on equipment, the system automatically optimizes a standard fluctuation interval to adapt to actual operation conditions, and in addition, a feedback module verifies a disposal effect to ensure that the abnormity is properly handled. The method can significantly improve the stability and operation efficiency of the power supply and distribution system, reduces the production risk, and is an important auxiliary means for management of metallurgical steel plants.

Description

Technical Field

[0001] This invention relates to a method for handling abnormal values ​​in power supply and distribution in a metallurgical steel plant. Background Technology

[0002] A metallurgical steel plant is an industrial facility focused on the production and processing of steel. Its main task is to transform iron ore and other raw materials into high-quality steel through a series of complex metallurgical processes, such as mining, beneficiation, sintering, ironmaking, steelmaking, and rolling. This steel is widely used in various fields, including construction, transportation, machinery, electronics, and energy.

[0003] The power supply and distribution system of a metallurgical steel plant is a crucial component of its production process, responsible for providing a stable and reliable power supply to the entire plant. The main function of the power supply and distribution system is to convert electrical energy from the external power grid into energy suitable for the internal equipment and processes of the metallurgical steel plant, and then transmit this energy to various points of consumption through the distribution network.

[0004] Currently, to ensure the power quality of metallurgical and steel plants, the parameters of the plant's electrical equipment and transmission lines are often recorded. However, during this recording process, some deviations often occur, which are considered outliers in the data. When no errors are reported from the electrical equipment or transmission lines, these outliers are often simply handled by deletion, replacement, or interpolation. However, in some situations, these outliers can indeed cause changes in the parameters of outdoor electrical equipment and transmission lines without the equipment reporting errors, such as during severe convective weather or sudden temperature changes. Furthermore, this approach does not improve the power quality of metallurgical and steel plants; it merely serves as a data record and cannot optimize the power supply based on the causes of the outliers.

[0005] Therefore, in order to solve the above-mentioned technical problems, this application proposes a method for handling abnormal values ​​in power supply and distribution in metallurgical steel plants. Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a method for handling abnormal values ​​in power supply and distribution in metallurgical steel plants.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for handling abnormal power supply and distribution values ​​in a metallurgical steel plant, comprising the following steps:

[0008] Step 1: Data collection phase. Set up a data collection module to collect normal power supply and distribution data of the metallurgical steel plant over a period of time. Based on the power supply and distribution data for this period of time, set a standard fluctuation range for power supply and distribution for each device, and then store the standard fluctuation range for each device.

[0009] Step 2: When collecting power supply and distribution data from the metallurgical steel plant, compare the collected data with the standard fluctuation range of the corresponding equipment to check whether the newly collected power supply and distribution data belongs to the standard fluctuation range. If so, it means that the data is normal and the normal value is stored in the data collection module. If not, mark the data as an abnormal value and mark the timestamp, equipment and related parameters of the abnormal value.

[0010] Step 3: Configure a data processing module. This module detects external factors when anomalies occur. These external environmental factors include, but are not limited to, temperature, humidity, human operation, and animal factors. The detected external environmental factor data is stored in the data collection module. The data processing module includes a weather factor detection unit, a human factor detection unit, an animal factor detection unit, and a handling unit. When anomalies occur, the weather factor detection unit, human factor detection unit, and animal factor detection unit detect whether extreme changes in weather conditions have occurred; whether errors have occurred in human operation; and whether animals, such as birds or rats, have approached the power equipment or transmission lines. If so, the anomaly is caused by an external factor, and subsequent data recordings without anomalies indicate that the factor will not have a long-term impact on the equipment. If not, the anomaly is caused by equipment failure or data acquisition error. In this case, the handling unit generates a corresponding maintenance notification, performs maintenance on the equipment with the anomaly, and retains the anomaly for subsequent analysis.

[0011] Step 4: Set up a feedback module. The feedback module interfaces with the detection module. When an abnormal value occurs, if the detection module does not find any abnormal external environmental factors, and the equipment and data acquisition are not faulty, it means that the value is normal and not an abnormal value. The feedback module will then generate feedback, mark the abnormal data as a normal value, and optimize the standard fluctuation range of the equipment.

[0012] 3. Preferably, the specific steps for detecting external environmental factors in step three using the weather factor detection unit, human factor detection unit, and animal factor detection unit are as follows:

[0013] Climate factor detection: The weather factor detection unit includes multiple sensors, including but not limited to temperature sensors and humidity sensors. The temperature and humidity sensors intermittently detect the external temperature and humidity and set detection intervals. When an abnormal value is found, the data collection module searches for the timestamp of the abnormal value and checks whether the difference between the temperature / humidity in the detection interval under that timestamp and the temperature / humidity in the previous detection interval is greater than the set threshold. If so, it indicates that the temperature / humidity is abnormal; if not, it indicates that the temperature / humidity is normal.

[0014] Human factor detection involves recording staff operation logs through a human factor detection unit. When an anomaly occurs, the human factor detection unit searches the operation logs under the timestamp of the anomaly and confirms whether the staff operation was abnormal.

[0015] Animal factor detection: The animal factor detection unit consists of multiple monitoring cameras. These cameras are used to collect images of electrical equipment and power transmission lines in the metallurgical and steel plant in real time. When an anomaly occurs, the animal factor detection unit searches for whether any animals have approached the electrical equipment and power transmission lines at that time, based on the timestamp of the anomaly.

[0016] Preferably, each detection interval is first labeled sequentially as 1, 2, ..., N, where N is a positive integer ≥ 2. The temperature data in the environmental factor data of the Mth time interval is labeled as X, and the temperature data in the (M+1)th time interval is labeled as Y, where 1 ≤ M ≤ N, and M is a positive integer. Then, the temperature difference value is calculated using the following formula:

[0017] Temperature difference value = |YX|;

[0018] Among them, the threshold for temperature difference is set to 15. When the temperature difference is greater than 15, the temperature data recorded by the environmental factor collection unit in that time interval is marked as abnormal.

[0019] Label the air humidity in the Mth time interval as A, and the air humidity in the (M+1)th time interval as B, where 1 ≤ M ≤ N, and M is a positive integer. Then, calculate the percentage difference in air humidity using the following formula:

[0020] Humidity difference percentage = |BA| / A × 100%;

[0021] The threshold for the percentage difference in air humidity is set at 20%. When the percentage difference in temperature is greater than 20%, the air humidity data recorded by the environmental factor collection unit in that time interval is marked as abnormal.

[0022] Preferably, the specific method for detecting animals at the locations of electrical equipment and power transmission lines using the animal factor detection unit is as follows:

[0023] A: Use multiple surveillance cameras to conduct all-weather, all-round image monitoring of the electrical equipment and transmission lines of the metallurgical and steel plant, and obtain clear images of the electrical equipment and transmission lines of the metallurgical and steel plant through image denoising, enhancement, and binarization.

[0024] B: Select images containing animals, and use image annotation tools, such as LabelImg, to annotate the animals in the images. The annotation method is to indicate the position of the animals in the image by using bounding boxes, and divide the images into training image set and test image set;

[0025] C: The convolutional neural network (CNN) is trained using a combination of cross-entropy loss and localization loss, along with a training image set, and the trained CNN is validated using a test image set.

[0026] D: Use a trained convolutional neural network (CNN) to detect animal images of electrical equipment and power transmission lines in the acquired images.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. In this invention, when an anomaly occurs, the time of its occurrence is recorded. Based on this time, it is checked whether there have been extreme changes in temperature and humidity at that time. If so, the cause of the anomaly is weather-related. Regarding human factors, an operation log is set up. Similarly, at the time the anomaly occurred, the employee operation log is checked to see if any human error occurred. If so, it indicates that the anomaly was caused by human error. The operation is then corrected, and the anomaly is recorded and associated with the erroneous operation for future reference. If not, it indicates that the anomaly may be caused by equipment malfunction or accidental data recording. In this case, the equipment exhibiting the anomaly can be inspected and repaired. If the equipment is functioning correctly and no abnormalities are observed due to weather or animal factors, the anomaly indicates an accident during data recording. In this case, appropriate corrective measures should be taken, and the anomaly should be retained for subsequent analysis. Animal factors are addressed by monitoring electrical equipment and checking for birds or other animals perched on power lines or equipment when an anomaly occurs. All three checks are performed simultaneously. This method allows for the investigation of the causes of anomalies, going beyond simple deletion or replacement. It effectively addresses problems in the power supply and distribution systems of metallurgical and steel plants by investigating the causes of anomalies, thereby enabling timely detection of issues with electrical equipment and power lines.

[0029] 2. In this invention, by setting a feedback module, when the detection module does not detect any abnormal external environmental factors, and the equipment does not malfunction and the data acquisition does not have any errors, the feedback module can mark the abnormal data as normal values ​​and optimize the standard fluctuation range of the equipment. This feedback mechanism helps to continuously improve and optimize the abnormal value handling method and increase the sensitivity of abnormal value judgment. Attached Figure Description

[0030] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0031] Figure 1 This is a logic block diagram of the subsequent use stage of a method for handling abnormal power supply and distribution values ​​in a metallurgical steel plant according to the present invention.

[0032] Figure 2 This is a logic block diagram of the data collection stage of a method for handling abnormal values ​​in power supply and distribution in a metallurgical steel plant according to the present invention.

[0033] Figure 3 This is a logic block diagram illustrating two scenarios where abnormal values ​​in a power supply and distribution anomaly handling method for a metallurgical steel plant are caused by external factors, according to the present invention. Detailed Implementation

[0034] Example: As shown in the system block diagram of the present invention, the present invention provides a method for handling abnormal values ​​in power supply and distribution in a metallurgical steel plant, including the following steps:

[0035] Step 1: Data collection phase. Set up a data collection module to collect normal power supply and distribution data of the metallurgical steel plant over a period of time. Based on the power supply and distribution data for this period, set a standard fluctuation range for power supply and distribution for each device. Then, store the standard fluctuation range for each device. In order to cover different weather conditions, the data collection time span is set to 1 year.

[0036] It should be noted that in some special cases, such as emergency shutdowns or restarts of metallurgical and steel plants, the power supply and distribution parameters cannot be included in the standard fluctuation range and must be marked as special cases separately.

[0037] Step 2: In the subsequent use phase, when collecting power supply and distribution data of the metallurgical steel plant, the collected data is compared with the standard fluctuation range of the corresponding equipment to check whether the newly collected power supply and distribution data belongs to the standard fluctuation range. If so, it means that the data is normal and the normal value is stored in the data collection module. If not, the data is marked as an abnormal value and the timestamp and equipment of the abnormal value are marked.

[0038] Step 3: Configure a data processing module. This module detects external factors when anomalies occur. These external environmental factors include, but are not limited to, temperature, humidity, human operation, and animal factors. The detected external environmental factor data is stored in the data collection module. The data processing module includes a weather factor detection unit, a human factor detection unit, an animal factor detection unit, and a handling unit. When an anomaly occurs, the weather factor detection unit, human factor detection unit, and animal factor detection unit detect whether extreme changes in weather conditions have occurred; whether errors have occurred in human operation; and whether animals, such as birds or rats, have approached the power equipment or transmission lines. If so, the anomaly is caused by an external factor; if not, the anomaly is caused by equipment failure or data acquisition error. In this case, the handling unit generates a corresponding maintenance notification, performs maintenance on the equipment exhibiting the anomaly, and retains the anomaly for subsequent analysis.

[0039] It should be noted that when an anomaly occurs and, after detection by the weather factor detection unit, human factor detection unit, and animal factor detection unit, it is found that the anomaly is caused by an external factor, there are two scenarios. First, if, after the anomaly occurs, as external factors such as the weather returning to normal, the human error being corrected, or animals near the electrical equipment and transmission lines being driven away, the subsequently collected values ​​return to normal, it indicates that the factor has not had a long-term impact on the equipment. In this case, the handling unit retains the value, since the anomaly was caused by a genuine external factor, and retaining the anomaly is beneficial for investigating the impact of changes in external factors on the electrical equipment and transmission lines. Secondly, regarding the impact on the line, if after an anomaly occurs, and external factors such as the weather returning to normal, correction of human error, or removal of animals that approach the electrical equipment and transmission lines fail to restore the collected values ​​to normal, it indicates that the factor has had a long-term impact on the electrical equipment or transmission lines. In this case, the equipment needs to be inspected and the anomaly needs to be deleted through the processing unit. Since the anomaly was caused by the equipment malfunction, retaining it would not have any practical research significance. The decision to retain the anomaly can be made based on the cause of its occurrence, to prevent the deletion of anomalies caused by the actual situation from resulting in incomplete data.

[0040] The climate factor detection unit includes multiple sensors, including but not limited to temperature and humidity sensors. These sensors intermittently detect external temperature and humidity, with defined detection intervals. When an anomaly occurs, the data collection module searches for the timestamp of the anomaly and checks if the temperature / humidity difference between the detection interval at that timestamp and the previous detection interval exceeds a set threshold. If so, the temperature / humidity is abnormal; otherwise, it is normal. The specific method for determining whether the temperature / humidity is abnormal is as follows: First, each detection interval is sequentially labeled as 1, 2, ..., N, where N is a positive integer ≥ 2. The temperature data in the environmental factor data of the Mth time interval is labeled as X, and the temperature data in the (M+1)th time interval is labeled as Y, where 1 ≤ M ≤ N, and M is a positive integer. Then, the temperature difference value is calculated using the following formula:

[0041] Temperature difference value = |YX|;

[0042] Among them, the threshold for temperature difference is set to 15. When the temperature difference is greater than 15, the temperature data recorded by the environmental factor collection unit in that time interval is marked as abnormal.

[0043] Label the air humidity in the Mth time interval as A, and the air humidity in the (M+1)th time interval as B, where 1 ≤ M ≤ N, and M is a positive integer. Then, calculate the percentage difference in air humidity using the following formula:

[0044] Humidity difference percentage = |BA| / A × 100%;

[0045] The threshold for the percentage difference in air humidity is set at 20%. When the percentage difference in temperature exceeds 20%, the air humidity data recorded by the environmental factor collection unit in that time interval is marked as abnormal. If the temperature / humidity is abnormal at the time stamp of the abnormal value, the cause of the abnormal value is considered to be extreme weather. If the data recorded later does not show abnormal values, it means that the factor will not have a long-term impact on the equipment. At this time, the processing unit retains the value to explore the impact of changes in external factors on electrical equipment and transmission lines. If abnormal values ​​still appear later, it means that extreme weather has caused damage to the electrical equipment and transmission lines of the metallurgical steel plant, and the electrical equipment and transmission lines need to be repaired in time. If, in the process of exploring abnormal values ​​multiple times, it is found that weather factors have a significant impact on the power supply and distribution of the metallurgical steel plant, the environmental adaptability of the system can be optimized, such as replacing equipment with more suitable equipment to improve the stability of power supply and distribution.

[0046] Human factor detection involves recording staff operation logs through a human factor detection unit. When an anomaly occurs, the human factor detection unit searches the operation logs under the timestamp of the anomaly to confirm whether the staff's operation was abnormal. If the recorded operation log shows that the staff's operation was abnormal when the anomaly occurred, it indicates that the anomaly was caused by human error. In this case, the improper operation is corrected, and the anomaly is correlated with the operation for subsequent analysis. In such cases, the occurrence of human error can be reduced through training and system constraints.

[0047] Animal factor detection: The animal factor detection unit consists of multiple monitoring cameras. These cameras are used to collect images of electrical equipment and power transmission lines in the metallurgical and steel plant in real time. When an anomaly occurs, the animal factor detection unit searches for whether any animals have approached the electrical equipment and power transmission lines at that time, based on the timestamp of the anomaly.

[0048] Among them, the weather factor detection unit, human factor detection unit, and animal factor detection unit are performed simultaneously because the occurrence of outliers may be caused by multiple factors, rather than just one of weather factors, human factors, or animal factors.

[0049] It should be noted that the specific method for detecting animals at the locations of electrical equipment and power transmission lines using the animal factor detection unit is as follows:

[0050] A: Multiple surveillance cameras are used to conduct 24 / 7, all-around image monitoring of the electrical equipment and power transmission lines in the metallurgical and steel plant. Clear images of the electrical equipment and power transmission lines in the metallurgical and steel plant are obtained through image denoising, enhancement, and binarization.

[0051] B: Select images containing animals, and use image annotation tools, such as LabelImg, to annotate the animals in the images. The annotation method is to indicate the position of the animals in the image by using bounding boxes, and divide the images into training image set and test image set;

[0052] C: The convolutional neural network (CNN) is trained using a combination of cross-entropy loss and localization loss, along with a training image set, and the trained CNN is validated using a test image set.

[0053] D: Use a trained convolutional neural network (CNN) to detect animal images of electrical equipment and power transmission lines in the acquired images.

[0054] Step 4: Set up a feedback module. The feedback module interfaces with the detection module. When an abnormal value occurs, if the detection module does not find any abnormal external environmental factors, and the equipment and data acquisition are not faulty, it means that the value is normal and not an abnormal value. The feedback module will then generate feedback, mark the abnormal data as a normal value, and optimize the standard fluctuation range of the equipment.

[0055] The specific implementation process of this invention is as follows: First, through the data collection module, based on the normal power supply and distribution data of the metallurgical plant over the previous year, a standard fluctuation range is set for each electrical device and power supply line. In the subsequent actual use stage, the data collection module continuously captures various operating data of the electrical devices and power supply lines in the power supply and distribution system of the metallurgical steel plant in real time. Then, these data are compared with the standard fluctuation range set for each device in advance, and values ​​exceeding the standard fluctuation range are marked as abnormal values. Once an abnormal value occurs, the system starts the data processing module. This module analyzes external factors, such as external environmental factors like sudden changes in temperature, abnormal increases in humidity, human error, or animal factors such as interference from birds, rats, etc., to power equipment and transmission lines. If it is caused by external factors, the abnormal value is processed according to the corresponding external factors. After detection by the weather factor detection unit, human factor detection unit, and animal factor detection unit, if the abnormal value is found to be caused by external factors, it is divided into two situations.

[0056] Firstly, if after the anomaly occurs, the collected values ​​return to normal as external factors, such as the weather returning to normal, correction of human error, or removal of animals approaching the equipment and transmission lines, indicate that the factor has not had a long-term impact on the equipment. In this case, the processing unit retains the value, since the anomaly was caused by a genuine external factor, and retaining the anomaly facilitates the investigation of the impact of changes in external factors on the equipment and transmission lines. Secondly, if after the anomaly occurs, the collected values ​​still do not return to normal as external factors, such as the weather returning to normal, correction of human error, or removal of animals approaching the equipment and transmission lines, indicate that the factor has had a long-term impact on the equipment or transmission lines. In this case, the equipment needs to be inspected and the anomaly needs to be deleted by the processing unit.

[0057] If the issue is not due to external factors, it indicates a possible equipment malfunction or data acquisition error. In this case, the processing unit generates a maintenance notification to instruct staff to inspect the equipment. If the equipment is functioning correctly, the abnormal data indicates a data acquisition error. In this case, the data acquisition process needs to be optimized, including but not limited to replacing the acquisition equipment. If equipment malfunction or problems in the data acquisition stage are also ruled out, it means the data is not abnormal and is caused by a low-probability event, which is normal. However, this data did not appear when the standard fluctuation range for power supply and distribution was set for each device. In this case, the feedback module generates feedback, marks the abnormal data as a normal value, and optimizes the standard fluctuation range for that device.

[0058] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Those skilled in the art can readily implement the present invention based on the accompanying drawings and the above description. However, any modifications, alterations, or variations made by those skilled in the art without departing from the scope of the present invention, utilizing the disclosed technical content, are equivalent embodiments of the present invention. Furthermore, any modifications, alterations, or variations made to the above embodiments based on the essential technology of the present invention are still within the protection scope of the present invention.

Claims

1. A method for handling abnormal values ​​in power supply and distribution in a metallurgical steel plant, characterized in that: Includes the following steps: Step 1: Data collection phase. Set up a data collection module to collect normal power supply and distribution data of the metallurgical steel plant over a period of time. Based on the power supply and distribution data for this period of time, set a standard fluctuation range for power supply and distribution for each device, and then store the standard fluctuation range for each device. Step 2: When collecting power supply and distribution data from the metallurgical steel plant, compare the collected data with the standard fluctuation range of the corresponding equipment to check whether the newly collected power supply and distribution data belongs to the standard fluctuation range. If so, it means that the data is normal and the normal value is stored in the data collection module. If not, mark the data as an abnormal value and mark the timestamp, equipment and related parameters of the abnormal value. Step 3: Configure a data processing module. This module detects external factors when anomalies occur. These external environmental factors include temperature, humidity, human error, and animal factors. The detected data is stored in the data collection module. The data processing module includes weather factor detection, human factor detection, animal factor detection, and a response unit. When anomalies occur, the weather factor detection, human factor detection, and animal factor detection units detect whether extreme weather conditions have occurred; whether human error has occurred; and whether animals are present. If so, the anomaly is caused by an external factor, and subsequent data recordings without anomalies indicate that this factor will not have a long-term impact on the equipment. If not, it means that the abnormal value is caused by equipment failure or data acquisition error. In this case, the handling unit will generate a corresponding maintenance notice, perform maintenance on the equipment with the abnormal value, and retain the abnormal value for subsequent analysis. Step 4: Set up a feedback module. The feedback module interfaces with the detection module. When an abnormal value occurs, if the detection module does not find any abnormal external environmental factors, and the equipment and data acquisition are not faulty, it means that the value is normal and not an abnormal value. The feedback module will then generate feedback, mark the abnormal data as a normal value, and optimize the standard fluctuation range of the equipment.

2. The method for handling abnormal power supply and distribution values ​​in a metallurgical steel plant according to claim 1, characterized in that: The specific steps for detecting external environmental factors in step three, using the weather factor detection unit, human factor detection unit, and animal factor detection unit, are as follows: Climate factor detection: The weather factor detection unit includes multiple sensors, including temperature sensors and humidity sensors. The temperature and humidity sensors intermittently detect the external temperature and humidity and set detection intervals. When an abnormal value is found, the data collection module searches for the timestamp of the abnormal value and checks whether the difference between the temperature / humidity of the detection interval under that timestamp and the temperature / humidity of the previous detection interval is greater than the set threshold. If it is, it means that the temperature / humidity is abnormal; if not, it means that the temperature / humidity is normal. Human factor detection involves recording staff operation logs through a human factor detection unit. When an anomaly occurs, the human factor detection unit searches the operation logs under the timestamp of the anomaly and confirms whether the staff operation was abnormal. Animal factor detection: The animal factor detection unit consists of multiple monitoring cameras. These cameras are used to collect images of electrical equipment and power transmission lines in the metallurgical and steel plant in real time. When an anomaly occurs, the animal factor detection unit searches for whether any animals have approached the electrical equipment and power transmission lines at that time, based on the timestamp of the anomaly.

3. The method for handling abnormal power supply and distribution values ​​in a metallurgical steel plant according to claim 2, characterized in that: The specific method for determining whether temperature / humidity is abnormal is as follows: First, label each detection interval sequentially as 1, 2, ..., N, where N is a positive integer ≥ 2. Label the temperature data in the environmental factor data of the Mth time interval as X, and the temperature data in the (M+1)th time interval as Y, where 1 ≤ M ≤ N, and M is a positive integer. Then, calculate the temperature difference value using the following formula: Temperature difference value = |YX|; Among them, the threshold for temperature difference is set to 15. When the temperature difference is greater than 15, the temperature data recorded by the environmental factor collection unit in that time interval is marked as abnormal. Label the air humidity in the Mth time interval as A, and the air humidity in the (M+1)th time interval as B, where 1 ≤ M ≤ N, and M is a positive integer. Then, calculate the percentage difference in air humidity using the following formula: Humidity difference percentage = |BA| / A × 100%; The threshold for the percentage difference in air humidity is set at 20%. When the percentage difference in temperature is greater than 20%, the air humidity data recorded by the environmental factor collection unit in that time interval is marked as abnormal.

4. The method for handling abnormal power supply and distribution values ​​in a metallurgical steel plant according to claim 3, characterized in that: The specific method for detecting animals at the locations of electrical equipment and transmission lines using an animal factor detection unit is as follows: A: Use multiple surveillance cameras to conduct all-weather, all-round image monitoring of the electrical equipment and transmission lines of the metallurgical and steel plant, and obtain clear images of the electrical equipment and transmission lines of the metallurgical and steel plant through image denoising, enhancement, and binarization. B: Select images containing animals, use the image annotation tool LabelImg to annotate the animals in the images, and use bounding boxes to indicate the positions of the animals in the images. Then, divide the images into training image sets and test image sets. C: The convolutional neural network (CNN) is trained using a combination of cross-entropy loss and localization loss, along with a training image set, and the trained CNN is validated using a test image set. D: Use a trained convolutional neural network (CNN) to detect animal images of electrical equipment and power transmission lines in the acquired images.

Images