Converter smelting stokehole safety risk monitoring and early warning system based on AI identification

Abstract

The invention relates to the field of risk monitoring, in particular to a converter smelting stokehole safety risk monitoring and early warning system based on AI identification. Comprising an image acquisition module, a furnace front area detection module, a furnace front area analysis module, an action execution detection module, an action execution analysis module, a personnel protection safety analysis module, a safety risk comprehensive evaluation module and a management database. Meanwhile, according to the furnace front area position parameters, an area position risk evaluation coefficient is obtained, according to the action execution parameters of the operating personnel, an action execution risk evaluation coefficient is obtained, according to the protection parameters of the operating personnel, a personnel protection safety coefficient is obtained, and then a furnace front safety risk comprehensive evaluation index is comprehensively analyzed; according to the invention, intelligent safety management can be realized, the pressure of manual monitoring is reduced, and the management efficiency and quality are improved.

Description

Technical Field

[0001] The present invention relates to the field of risk monitoring, and specifically, to a converter smelting furnace front safety risk monitoring and early warning system based on AI recognition. Background Art

[0002] With the continuous development of the steel industry, converter smelting has been widely used as an important steelmaking method. However, the converter smelting process is a complex and dangerous industrial production process. There are many safety risk factors in the furnace operation, such as high-temperature melt splashing, toxic and harmful gas release, mechanical failure, etc. These risks may not only pose a serious threat to the life safety of operators, but also have a huge impact on production equipment and the environment. Therefore, it is very important to effectively monitor and warn of safety risks in the converter smelting furnace, which not only helps to ensure the safety and health of workers, but also helps to improve the stability and efficiency of the entire production process.

[0003] For example, the existing Chinese patent with the announcement number CN110045594B discloses an intelligent management and control system and method for predicting the risk of boiler four-tube status. The solution uses a variety of sensors and acquisition modules to evaluate the risk of boiler heating surface status through three-dimensional monitoring, boiler heating surface tube leakage monitoring, heating surface wall temperature monitoring, three-dimensional temperature field simulation, equivalent temperature calculation and other methods.

[0004] However, the above patent has the following problems: First, the solution mainly targets the status risk of the boiler itself, and does not involve analyzing the operational risk of the operator. This may lead to risks in the actual operation process due to the operator's non-standard operation or improper handling of specific situations, and cannot fully guarantee the safety of boiler operation.

[0005] Second, the plan does not take into account the risks of the operators themselves. Operators may make mistakes during the operation due to negligence, lack of skills, inadequate protection, etc., which may cause a series of safety problems. Summary of the invention

[0006] In order to overcome the shortcomings of the background technology, the embodiment of the present invention provides a converter smelting furnace front safety risk monitoring and early warning system based on AI recognition, which can effectively solve the problems involved in the above-mentioned background technology.

[0007] The purpose of the present invention can be achieved through the following technical solutions: The present invention provides a converter smelting furnace front safety risk monitoring and early warning system based on AI recognition, including: an image acquisition module, which is used to perform video monitoring of the furnace front area through a camera to obtain videos of the furnace front area at different time periods.

[0008] The furnace front area detection module is used to determine whether the operators have entered the dangerous area based on the videos of the furnace front area at different time periods, and provide feedback to the system. At the same time, the furnace front area position parameters are obtained. The furnace front area position parameters include the number of times each operator has entered the dangerous area and the safety level of each operator's position.

[0009] The furnace front area analysis module is used to obtain the regional location risk assessment coefficient based on the furnace front area location parameters.

[0010] The action execution detection module is used to detect the action execution parameters of the operator. The action execution parameters of the operator include the operator's feeding compliance, the abnormality of the furnace pulling speed, and the angle compliance of the oxygen gun.

[0011] The action execution analysis module is used to obtain the action execution risk evaluation coefficient based on the action execution parameter analysis of the operator.

[0012] The personnel protection safety analysis module is used to detect the protection parameters of each operator and then analyze the personnel protection safety factor. The protection parameters include the position deviation of the protection equipment and the coverage area qualification of the protection equipment.

[0013] The safety risk comprehensive evaluation module is used to obtain the comprehensive evaluation index of furnace safety risk based on the regional location risk evaluation coefficient λ, the action execution risk evaluation coefficient ξ, and the personnel protection safety factor χ, and provide feedback to the system.

[0014] Management database for storing preset safety distances, reasonable range of pulling furnace speeds, and standard angles of oxygen lances.

[0015] Preferably, the specific analysis method of the image acquisition module is as follows: in the first step, the working time of the converter smelting furnace is divided into a number of equal-length time periods according to the set time period, which are recorded as each time period.

[0016] The second step is to obtain real-time videos of the key area in front of the furnace at each time period through a high-definition camera, which are recorded as the videos of the key area in front of the furnace at each time period.

[0017] Preferably, the specific analysis method of the furnace front area detection module is as follows: the first step is to read the videos of the furnace front area at each time period, use the target detection technology to identify each operator from the videos of the furnace front area at each time period, and mark them, record them as each operator in the videos of the furnace front area at each time period, track the movement trajectory of each operator in the videos of the furnace front area at each time period, and take a number of time points with equal time intervals in each time period, record them as each time point, obtain the position of each operator at each time point in the videos of the furnace front area at each time period, and compare the position of each operator at each time point in the videos of the furnace front area at each time period with the pre-set dangerous area boundary in real time. If the position of a certain operator at a certain time point in the videos of the furnace front area at a certain time period is completely or partially within the dangerous area, it is determined that the operator has entered the dangerous area at this time point in the period, and feedback is given to the system. If the position of a certain operator at a certain time point in the videos of the furnace front area at a certain time period is outside the dangerous area, it is determined that the operator has not entered the dangerous area, and the number of times each operator enters the dangerous area is counted, which is recorded as N m , m represents the number of the mth operator, m=1,2,...,q.

[0018] The second step is to read the position of each operator at each time point in the video of each period in the furnace area, recorded as (x imf ,y imf ), i represents the number of the i-th time period, i=1,2,...,n, f represents the number of the f-th time point, f=1,2,...,k, and at the same time, the position of the converter smelting furnace body is obtained, recorded as (x',y'), and substituted into the formula Get the distance d of each operator at each time point in the video of each period imf At the same time, the preset safety distance is read from the management database, recorded as d', and substituted into the formula Get the position safety level α of each operator m , k represents the number of time points, and n represents the number of time periods.

[0019] Preferably, the specific analysis method of the regional location risk assessment coefficient is: reading the number of times each operator enters the dangerous area N m 、Location safety level α m , the average number of times each operator enters the dangerous area is obtained by taking the average number of times the operator enters the dangerous area, recorded as Substituting this into the formula The regional location risk assessment coefficient λ is obtained, q represents the number of operators, φ1 and φ2 represent the weight factors of the number of times of entering the dangerous area and the safety degree of the location, respectively.

[0020] Preferably, the specific analysis method of the operator's feeding compliance degree is as follows: the first step is to use a high-definition camera to capture the video of each operator's feeding process, which is recorded as the feeding video of each operator.

[0021] The second step is to mark the position points in the feeding video of each operator according to the set feeding trajectory, and record them as marked position points.

[0022] The third step is to track the feeding action trajectory of each operator in the feeding video according to each marked position point to generate the feeding movement trajectory of each operator.

[0023] The fourth step is to compare the feeding motion trajectory of each operator with the preset feeding standard trajectory to analyze the feeding compliance degree of the operator.

[0024] Preferably, the specific analysis method of the abnormal degree of the pulling furnace speed is: by installing a speed sensor on the pulling furnace equipment to detect the pulling furnace speed at each time point in each period, recorded as the pulling furnace speed at each time point in each period, and by taking the average value of the pulling furnace speed at each time point in each period to obtain the pulling furnace speed in each period, recorded as v i At the same time, the preset reasonable range of pulling furnace speed is read from the management database, and the maximum and minimum values ​​corresponding to the preset reasonable range of pulling furnace speed are recorded as the maximum reference value v of pulling furnace speed. max , minimum reference value of pulling furnace speed v min , substituting it into the formula Obtain the abnormal degree of pulling furnace speed in each period δ i , and take the average value to get the abnormal degree of pulling furnace speed, which is recorded as δ.

[0025] Preferably, the specific analysis method of the angle compliance of the oxygen lance is: the angular velocity of the gyroscope at each time point in each period is detected by the gyroscope installed on the oxygen lance, which is recorded as ω if , through the formula Get the angle θ of the oxygen gun in each time period i , Δt represents the duration of the time interval, k represents the number of time points, and the preset oxygen gun standard angle is read from the management database, recorded as θ', and the formula is used to obtain Obtain the angle compliance of the oxygen gun n represents the number of time periods.

[0026] Preferably, the specific analysis method of the action execution risk assessment coefficient is: respectively reading the operator's feeding compliance, the abnormality degree of the pulling furnace speed δ, the angle compliance degree of the oxygen gun Substituting this into the formula The action execution risk evaluation coefficient ξ is obtained, and γ represents the feeding compliance degree of the operator. The weighting factors represent the preset charging compliance, furnace pulling speed abnormality, and oxygen lance angle compliance.

[0027] Preferably, the specific analysis method of the personnel protection safety analysis module is as follows: in the first step, images of each operator wearing protective equipment are acquired through a camera, recorded as images of each operator, a number of key feature points are selected on the protective equipment according to the set position, and the protective equipment is worn on a human body model to acquire the position coordinates of each key feature point, recorded as the coordinates of each key feature point, and at the same time, the coordinates of each key feature point of the protective equipment worn by each operator are extracted from the images of each operator through image recognition technology, recorded as the coordinates of each key feature point of the protective equipment of each operator, and the position deviation of the protective equipment of each operator is obtained by calculation, recorded as d' m .

[0028] In the second step, each worker image is divided into several grids of equal area, and the number of grids covered by protective equipment in each worker image is counted. At the same time, the total number of grids in each worker image is extracted. The number of grids covered by protective equipment in each worker image is divided by the total number of grids in each worker image to obtain the protective equipment coverage area of ​​each worker, which is recorded as M. m At the same time, the image of the human body model wearing protective equipment is acquired, recorded as the standard wearing image, and the standard wearing image is analyzed according to the method of analyzing the coverage area of ​​the protective equipment of each operator to obtain the standard coverage area of ​​the protective equipment, recorded as M', and the formula Obtain the qualified coverage of protective equipment for each operator

[0029] The third step is to calculate the position deviation d' of each operator's protective equipment. m , Qualification of protective equipment coverage area Substitute into the formula The personnel protection safety factor χ is obtained, κ1 and κ2 represent the weight factors of the set protective equipment position deviation and coverage area qualification, and q represents the number of operators.

[0030] Preferably, the specific analysis method of the safety risk comprehensive evaluation module is: read the regional location risk evaluation coefficient λ, the action execution risk evaluation coefficient ξ, and the personnel protection safety factor χ respectively, and substitute them into the formula Obtain the comprehensive evaluation index of furnace safety risk w1, w2, and w3 respectively represent the weight factors of the set regional location risk assessment coefficient, action execution risk assessment coefficient, and personnel protection safety factor. The comprehensive evaluation index of the furnace-front safety risk is compared with the preset threshold of the comprehensive evaluation index of the furnace-front safety risk. If the comprehensive evaluation index of the furnace-front safety risk is greater than or equal to the preset threshold of the comprehensive evaluation index of the furnace-front safety risk, it means that the comprehensive evaluation index of the furnace-front safety risk is qualified, otherwise it means that it is unqualified, and feedback is given to the system.

[0031] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention uses a camera to perform video monitoring of the area in front of the furnace, obtains videos of the area in front of the furnace at different time periods, thereby determining whether the operating personnel have entered the dangerous area, and provides feedback to the system, effectively reducing safety accidents caused by operating personnel mistakenly entering the dangerous area, and ensuring the safety of personnel lives.

[0032] 2. The present invention obtains a comprehensive evaluation index of furnace safety risk based on a comprehensive analysis of the regional location risk evaluation coefficient, the action execution risk evaluation coefficient, and the personnel protection safety factor, and provides feedback to the system, thereby realizing intelligent safety management, reducing the pressure of manual monitoring, and improving management efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0034] Figure 1 This is a system module connection diagram of the present invention.

[0035] Figure 2 for Figure 1 Schematic diagram of the process of image acquisition module.

[0036] Figure 3 It is a flow chart of the material feeding compliance of the operator in the action execution detection module. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] See also Figure 1As shown, the converter smelting furnace front safety risk monitoring and early warning system based on AI recognition includes an image acquisition module, a furnace front area detection module, a furnace front area analysis module, an action execution detection module, an action execution analysis module, a personnel protection safety analysis module, a safety risk comprehensive evaluation module, and a management database.

[0039] The management database is connected to the furnace front area analysis module, the action execution analysis module, the personnel protection safety analysis module, and the comprehensive safety risk evaluation module; the comprehensive safety risk evaluation module is connected to the furnace front area analysis module, the action execution analysis module, and the personnel protection safety analysis module; the furnace front area detection module is connected to the furnace front area analysis module; and the furnace front area detection module is connected to the furnace front area analysis module.

[0040] The image acquisition module is used to monitor the furnace front area through a camera and obtain videos of the furnace front area at different time periods.

[0041] See also Figure 2 As shown, the specific analysis method of the image acquisition module is: the first step is to divide the working time of the converter smelting furnace into several equal-length time periods according to the set time, which are recorded as each time period; to improve production efficiency and quality control, through targeted management of each time period, problems can be discovered and adjusted in time.

[0042] The second step is to obtain real-time video of the key area in front of the furnace at each time period through a high-definition camera, which is recorded as the video of the area in front of the furnace at each time period; this improves safety and enables timely detection of potential dangerous situations, such as abnormal flames, equipment failures, etc., so that countermeasures can be taken quickly.

[0043] The furnace front area detection module is used to determine whether the operators have entered the dangerous area based on the videos of the furnace front area at different time periods, and provide feedback to the system. At the same time, the furnace front area position parameters are obtained. The furnace front area position parameters include the number of times each operator has entered the dangerous area and the safety level of each operator's position.

[0044] The specific analysis method of the furnace front area detection module is as follows: the first step is to read the videos of the furnace front area at each time period, use the target detection technology to identify each operator from the videos of the furnace front area at each time period, and mark them, record them as each operator in the videos of the furnace front area at each time period, track the movement trajectory of each operator in the videos of the furnace front area at each time period, and take a number of time points with equal time intervals in each time period, record them as each time point, obtain the position of each operator at each time point in the videos of the furnace front area at each time period, and compare the position of each operator at each time point in the videos of the furnace front area at each time period with the pre-set dangerous area boundary in real time. If the position of a certain operator at a certain time point in the video of a certain time period in the furnace front area is completely or partially within the dangerous area, it is determined that the operator has entered the dangerous area at this time point in this period, and feedback is given to the system. If the position of a certain operator at a certain time point in the video of a certain time period in the furnace front area is outside the dangerous area, it is determined that the operator has not entered the dangerous area, and the number of times each operator enters the dangerous area is counted, which is recorded as N m , m represents the number of the mth operator, m=1,2,...,q; it can greatly improve the safety of the work site, timely warn operators entering dangerous areas, and reduce the risk of accidents.

[0045] It should be noted that the specific analysis method for judging whether the operator has entered the danger zone based on the video of the furnace front area in each time period is as follows: obtaining the position of each operator at each time point in the video of the furnace front area in each time period, and reading the preset boundary of the danger zone at the same time, making a ray from the position of each operator at each time point in the video of the furnace front area in each time period in any direction, and calculating the number of intersections of the ray with the boundary line segment of the danger zone; if the number of intersections of the ray of a certain operator at a certain time point in the video of the furnace front area in a certain time period and the boundary line segment of the danger zone is an odd number, then it is judged that the operator has entered the danger zone at that time point in that time period, and feedback is given to the system; otherwise, it is judged that the operator has not entered the danger zone at that time point in that time period; the accuracy and reliability of the judgment of the danger zone are improved, the possibility of misjudgment is reduced, the safety of the operators can be better protected, and potential risks caused by inaccurate judgment can be avoided.

[0046] The second step is to read the position of each operator at each time point in the video of each period in the furnace area, recorded as (x imf ,y imf ), i represents the number of the i-th time period, i=1,2,...,n, f represents the number of the f-th time point, f=1,2,...,k, and at the same time, the position of the converter smelting furnace body is obtained, recorded as (x',y'), and substituted into the formula Get the distance d of each operator at each time point in the video of each period imf At the same time, the preset safety distance is read from the management database, recorded as d', and substituted into the formula Get the position safety level α of each operator m , k represents the number of time points, and n represents the number of time periods; it can grasp the position of workers in the furnace front area and their distance from the furnace body in real time, discover possible safety hazards in time and issue early warnings, and improve the safety of the work site.

[0047] The furnace front area analysis module is used to obtain the regional location risk assessment coefficient based on the furnace front area location parameters.

[0048] The specific analysis method of the regional location risk assessment coefficient is: read the number of times each operator enters the dangerous area N m 、Location safety level α m , the average number of times each operator enters the dangerous area is obtained by taking the average number of times the operator enters the dangerous area, recorded as Substituting this into the formula The regional location risk assessment coefficient λ is obtained, q represents the number of operators, φ1 and φ2 represent the set number of times of entering the dangerous area and the weight factors of the location safety level respectively; the situation of each operator entering the dangerous area is accurately understood so as to carry out more targeted management and intervention.

[0049] It should be noted that, in a specific embodiment, φ1 can be set to 0.6 and φ2 can be set to 0.4. The number of times entering the dangerous area directly reflects the frequency of exposure to risks. Frequent entry into the dangerous area means that the possibility of facing danger and the probability of potential consequences are higher. In comparison, although the degree of location safety is important, it is more of a static assessment, and the actual risk is often closely related to the actual entry and interaction of personnel. Therefore, the number of times entering the dangerous area has a higher weight.

[0050] The action execution detection module is used to detect the action execution parameters of the operator. The action execution parameters of the operator include the operator's feeding compliance, the abnormality of the furnace pulling speed, and the angle compliance of the oxygen gun.

[0051] See also Figure 3 As shown, the specific analysis method of the operator's feeding compliance degree is: the first step is to use a high-definition camera to capture the video of each operator's feeding process, which is recorded as the feeding video of each operator; it is helpful to supervise whether the operators feed in accordance with the specifications and ensure production quality and safety.

[0052] The second step is to mark the position points in the feeding video of each operator according to the set feeding trajectory, and record them as marked position points; the deviation from the standard trajectory during the feeding process can be discovered in time for timely correction and adjustment.

[0053] The third step is to track the trajectory of each operator's feeding action in the feeding video according to the marked position points to generate the feeding movement trajectory of each operator; this is conducive to ensuring the accuracy and consistency of feeding, improving the stability of product quality, and timely discovering irregular or unreasonable feeding action trajectories, thereby optimizing the operation process.

[0054] The fourth step is to compare the feeding movement trajectory of each operator with the preset feeding standard trajectory to analyze the feeding compliance of the operators; this is to facilitate targeted training and guidance for operators who do not meet the requirements and improve the overall work level.

[0055] It should be noted that the specific analysis method of the feeding compliance degree of the operators is: obtaining the feeding motion trajectory of each operator, and extracting the preset feeding standard trajectory from the management database, respectively overlapping the feeding motion trajectory of each operator with the preset feeding standard trajectory, and obtaining the overlapping length of the feeding motion trajectory of each operator with the preset feeding standard trajectory, which is recorded as L m At the same time, the preset standard feeding trajectory length is recorded as L', and the formula The degree of compliance of the operators in feeding is obtained, and q represents the number of operators; it can better grasp the actual situation in the production process, optimize the production process and standards in time, ensure the standardization and consistency of the entire feeding operation, and ensure the stability of product quality.

[0056] The specific analysis method of the abnormal degree of the pulling furnace speed is as follows: by installing a speed sensor on the pulling furnace equipment to detect the pulling furnace speed at each time point in each period, recorded as the pulling furnace speed at each time point in each period, and by taking the average value of the pulling furnace speed at each time point in each period to obtain the pulling furnace speed in each period, recorded as v i At the same time, the preset reasonable range of pulling furnace speed is read from the management database, and the maximum and minimum values ​​corresponding to the preset reasonable range of pulling furnace speed are recorded as the maximum reference value v of pulling furnace speed. max , minimum reference value of pulling furnace speed v min , substituting it into the formula Obtain the abnormal degree of pulling furnace speed in each period δ i , and take the average value to get the abnormal degree of pulling furnace speed, recorded as δ; accurately grasp the actual speed of the pulling furnace in each period, discover abnormal fluctuations in time, and be able to discover potential problems in time to prevent production failures or quality problems caused by unreasonable pulling furnace speed.

[0057] The specific analysis method of the angle compliance of the oxygen lance is as follows: the angular velocity of the gyroscope at each time point in each period is detected by the gyroscope installed on the oxygen lance, which is recorded as ω if , through the formula Get the angle θ of the oxygen gun in each time period i, Δt represents the duration of the time interval, k represents the number of time points, and the preset oxygen gun standard angle is read from the management database, recorded as θ', and the formula is used to obtain Obtain the angle compliance of the oxygen gun n represents the number of time periods; being able to accurately grasp the angle status of the oxygen lance at different time periods to ensure that it is within a reasonable range helps to ensure the smooth progress of steelmaking and other process and improve production quality and efficiency.

[0058] The action execution analysis module is used to obtain the action execution risk evaluation coefficient based on the action execution parameter analysis of the operator.

[0059] The specific analysis method of the action execution risk assessment coefficient is: respectively reading the operator's feeding compliance, the abnormal degree of furnace pulling speed δ, the angle compliance of the oxygen gun Substituting this into the formula The action execution risk evaluation coefficient ξ is obtained, and γ represents the feeding compliance degree of the operator. The weight factors represent the preset charging compliance, furnace pulling speed abnormality, and oxygen lance angle compliance. They help timely identify high-risk points that may exist in the operator's operation and improve the safety and standardization of the overall operation.

[0060] It should be noted that, in a specific embodiment, It can be set to 0.4. It can be set to 0.4. It can be set to 0.2. The degree of compliance with feeding is more important because the accuracy of feeding has a far-reaching impact on the entire production process. If the feeding does not meet the requirements, it may directly lead to product quality problems, reduced production efficiency, and even cause a series of serious consequences such as safety accidents. It is the key starting link of the entire production process. The degree of abnormality in furnace pulling speed is also very important because inappropriate furnace pulling speed may affect key factors such as the reaction process and temperature distribution in the furnace, which in turn has an important impact on product quality and equipment safety. The degree of compliance with the angle of the oxygen gun may be relatively less important. Compared with feeding and furnace pulling speed, its impact on the overall risk may not be so direct and critical. Therefore, the weights corresponding to the degree of compliance with feeding and the degree of abnormality in furnace pulling speed are higher.

[0061] The personnel protection safety analysis module is used to detect the protection parameters of each operator and then analyze the personnel protection safety factor. The protection parameters include the position deviation of the protection equipment and the coverage area qualification of the protection equipment.

[0062] The specific analysis method of the personnel protection safety analysis module is as follows: the first step is to obtain images of each operator wearing protective equipment through a camera, recorded as each operator image, select a number of key feature points on the protective equipment according to the set position, and wear the protective equipment on a human body model to obtain the position coordinates of each key feature point, recorded as each key feature point coordinates, and at the same time, extract the coordinates of each key feature point of the protective equipment worn by each operator from each operator image through image recognition technology, recorded as each key feature point coordinates of each operator protective equipment, and calculate the position deviation of each operator's protective equipment, recorded as d' m ; It can ensure that workers wear protective equipment correctly, improve work safety, and promptly detect situations where protective equipment is not worn properly so that they can correct them in time.

[0063] It should be noted that the specific analysis method of the position deviation of the protective equipment of each operator is: read the coordinates of each key feature point and the coordinates of each key feature point of each operator's protective equipment respectively, and record them as (x' s ,y' s ,z' s )、(x ms ,y ms ,z ms ), s represents the number of the sth key feature point, s = 1, 2, ..., r, and substitute it into the formula Get the position deviation d' of each operator's protective equipment m , r represents the number of key feature points; clarifying the specific deviation degree of each key feature point will help to make targeted improvements and adjustments, prompting operators to pay more attention to the correct wearing of protective equipment and enhance safety awareness.

[0064] In the second step, each worker image is divided into several grids of equal area, and the number of grids covered by protective equipment in each worker image is counted. At the same time, the total number of grids in each worker image is extracted. The number of grids covered by protective equipment in each worker image is divided by the total number of grids in each worker image to obtain the protective equipment coverage area of ​​each worker, which is recorded as M. m At the same time, the image of the human body model wearing protective equipment is acquired, recorded as the standard wearing image, and the standard wearing image is analyzed according to the method of analyzing the coverage area of ​​the protective equipment of each operator to obtain the standard coverage area of ​​the protective equipment, recorded as M', and the formula Obtain the qualified coverage of protective equipment for each operator It can accurately measure the area of ​​the body actually covered by protective equipment, intuitively reflect the compliance level of operators, and ensure effective protection of various parts of the body during the operation.

[0065] The third step is to calculate the position deviation d' of each operator's protective equipment. m , Qualification of protective equipment coverage area Substitute into the formula The personnel protection safety coefficient χ is obtained, κ1 and κ2 represent the weight factors of the set protective equipment position deviation and coverage area qualification, respectively, and q represents the number of operators. It can comprehensively evaluate the overall protection safety status of the operators and promptly discover the weak links in protection safety so as to take effective measures to improve them.

[0066] It should be noted that, in a specific embodiment, κ1 can be set to 0.6, κ2 can be set to 0.4. In the personnel protection safety factor, the influence of the position deviation of the protective equipment may be relatively large. If the position deviation of the protective equipment is large, it may cause it to fail to accurately play a protective role at a critical moment and fail to effectively protect personnel. The accuracy of the position is directly related to the actual effect of the protection, and the qualified coverage area is also very important. However, relatively speaking, as long as there is a certain coverage area, it can usually provide certain protection, unlike position deviation, which may completely make the protection ineffective. Therefore, the weight corresponding to the position deviation of the protective equipment is higher.

[0067] The safety risk comprehensive evaluation module is used to obtain the comprehensive evaluation index of furnace safety risk based on the regional location risk evaluation coefficient λ, the action execution risk evaluation coefficient ξ, and the personnel protection safety factor χ, and provide feedback to the system.

[0068] The specific analysis method of the safety risk comprehensive evaluation module is: read the regional location risk evaluation coefficient λ, the action execution risk evaluation coefficient ξ, and the personnel protection safety factor χ respectively, and substitute them into the formula Obtain the comprehensive evaluation index of furnace safety risk w1, w2, and w3 represent the weight factors of the set regional location risk assessment coefficient, action execution risk assessment coefficient, and personnel protection safety factor respectively. The comprehensive evaluation index of the furnace-front safety risk is compared with the preset threshold of the comprehensive evaluation index of the furnace-front safety risk. If the comprehensive evaluation index of the furnace-front safety risk is greater than or equal to the preset threshold of the comprehensive evaluation index of the furnace-front safety risk, it means that the comprehensive evaluation index of the furnace-front safety risk is qualified, otherwise it means that it is unqualified, and feedback is given to the system; timely feedback helps relevant personnel to quickly understand the safety situation so that they can take targeted measures to promote the refined management and continuous improvement of risks in all aspects of the furnace.

[0069] It should be noted that, in a specific embodiment, w1 can be set to 0.4, w2 can be set to 0.4, and w3 can be set to 0.2. The regional location involves the safety of the overall environment, such as whether it is in an area prone to danger, the distance from other potential risk sources, etc. The safety of the regional location has a fundamental and global impact on the entire furnace operation. The impact of the action execution risk assessment coefficient is also relatively high, because whether the operator's action execution is correct and standardized is directly related to whether the operation process will cause accidents. Incorrect action execution may quickly lead to serious consequences. The impact of the personnel protection safety factor may be relatively lower. Although personnel protection is very important, it plays a more important role in reducing injuries when accidents occur. In comparison, the former two are more critical to preventing the occurrence of risks. Therefore, the weights corresponding to the regional location risk assessment coefficient and the action execution risk assessment coefficient are higher.

[0070] Management database for storing preset safety distances, reasonable range of pulling furnace speeds, and standard angles of oxygen lances.

[0071] The present invention performs video monitoring of the furnace front area through a camera, obtains videos of the furnace front area in each time period, thereby judging whether the operating personnel have entered the dangerous area, and feeds back to the system. At the same time, the regional position risk assessment coefficient is obtained according to the position parameter analysis of the furnace front area, the action execution risk assessment coefficient is obtained according to the action execution parameter analysis of the operating personnel, and the personnel protection safety factor is obtained according to the protection parameter analysis of each operating personnel. Then, a comprehensive evaluation index of furnace front safety risk is obtained through comprehensive analysis, and the index is fed back to the system. Intelligent safety management can be realized, the pressure of manual monitoring can be reduced, and management efficiency and quality can be improved.

[0072] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention and they are still covered by the protection scope of the present invention.

Claims

1. The AI-based converter smelting furnace front safety risk monitoring and early warning system is characterized by: The system specifically includes the following modules: An image acquisition module is used to monitor the furnace front area through a camera and obtain videos of the furnace front area at different time periods; The furnace front area detection module is used to determine whether the operator has entered the dangerous area based on the video of the furnace front area at each time period, and provide feedback to the system. At the same time, the furnace front area position parameters are obtained. The furnace front area position parameters include the number of times each operator has entered the dangerous area and the safety level of each operator's position; The furnace front area analysis module is used to obtain the regional location risk assessment coefficient based on the furnace front area location parameters; The action execution detection module is used to detect the action execution parameters of the operator, including the operator's feeding compliance, the abnormality of the furnace pulling speed, and the angle compliance of the oxygen gun; An action execution analysis module is used to obtain an action execution risk evaluation coefficient based on the action execution parameters of the operator; Personnel protection safety analysis module is used to detect the protection parameters of each operator and then analyze the personnel protection safety factor. The protection parameters include the position deviation of the protective equipment and the coverage area qualification of the protective equipment. The safety risk comprehensive evaluation module is used to analyze and obtain the comprehensive evaluation index of the furnace safety risk based on the regional location risk evaluation coefficient λ, the action execution risk evaluation coefficient ξ, and the personnel protection safety factor χ, and provide feedback to the system; Management database for storing preset safety distances, reasonable range of pulling furnace speeds, and standard angles of oxygen lances.

2. The AI-based converter smelting furnace safety risk monitoring and early warning system according to claim 1 is characterized in that: The specific analysis method of the image acquisition module is: The first step is to divide the working time of the converter smelting furnace into a number of equal-length time periods according to the set time period, which are recorded as each time period; The second step is to obtain real-time videos of the key area in front of the furnace at each time period through a high-definition camera, which are recorded as the videos of the key area in front of the furnace at each time period.

3. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 2 is characterized in that: The specific analysis method of the furnace front area detection module is: The first step is to read the videos of the furnace front area at each time period, use the target detection technology to identify each operator from the videos of the furnace front area at each time period, and mark them, recorded as each operator in the videos of the furnace front area at each time period, track the movement trajectory of each operator in the videos of the furnace front area at each time period, and take several time points with equal time intervals in each time period, recorded as each time point, obtain the position of each operator at each time point in the videos of the furnace front area at each time period, and compare the position of each operator at each time point in the videos of the furnace front area at each time period with the pre-set dangerous area boundary in real time. If the position of a certain operator at a certain time point in the video of a certain time period in the furnace front area is completely or partially within the dangerous area, it is determined that the operator has entered the dangerous area at this time point in the period, and feedback is given to the system. If the position of a certain operator at a certain time point in the video of a certain time period in the furnace front area is outside the dangerous area, it is determined that the operator has not entered the dangerous area, and the number of times each operator enters the dangerous area is counted, recorded as N. m , m represents the number of the mth operator, m = 1, 2, ..., q; The second step is to read the position of each operator at each time point in the video of each period in the furnace area, recorded as (x imf ,y imf ), i represents the number of the i-th time period, i=1,2,...,n, f represents the number of the f-th time point, f=1,2,...,k, and at the same time, the position of the converter smelting furnace body is obtained, recorded as (x',y'), and substituted into the formula Get the distance d of each operator at each time point in the video of each period imf At the same time, the preset safety distance is read from the management database, recorded as d', and substituted into the formula Get the position safety level α of each operator m , k represents the number of time points, and n represents the number of time periods.

4. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 3 is characterized in that: The specific analysis method of the regional location risk assessment coefficient is: Read the number of times each operator enters the dangerous area N m 、Location safety level α m , the average number of times each operator enters the dangerous area is obtained by taking the average number of times the operator enters the dangerous area, recorded as Substituting this into the formula The regional location risk assessment coefficient λ is obtained, q represents the number of operators, φ1 and φ2 represent the weight factors of the number of times entering the dangerous area and the safety degree of the location, respectively.

5. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 1 is characterized in that: The specific analysis method of the operator's feeding compliance degree is as follows: The first step is to use a high-definition camera to capture the video of each operator's feeding process, which is recorded as the feeding video of each operator; The second step is to mark the position points in the feeding video of each operator according to the set feeding trajectory, and record them as marked position points; The third step is to track the feeding action trajectory of each operator in the feeding video of each operator according to each marked position point, and generate the feeding motion trajectory of each operator; The fourth step is to compare the feeding motion trajectory of each operator with the preset feeding standard trajectory to analyze the feeding compliance degree of the operator.

6. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 5 is characterized in that: The specific analysis method of the abnormal degree of the pulling furnace speed is as follows: The pulling furnace speed at each time point in each period is detected by installing a speed sensor on the pulling furnace equipment, which is recorded as the pulling furnace speed at each time point in each period. The pulling furnace speed at each time point in each period is obtained by taking the average value of the pulling furnace speed at each time point in each period, which is recorded as v i At the same time, the preset reasonable range of pulling furnace speed is read from the management database, and the maximum and minimum values ​​corresponding to the preset reasonable range of pulling furnace speed are recorded as the maximum reference value v of pulling furnace speed. max , minimum reference value of pulling furnace speed v min , substituting it into the formula Obtain the abnormal degree of pulling furnace speed in each period δ i , and take the average value to get the abnormal degree of pulling furnace speed, which is recorded as δ.

7. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 6 is characterized in that: The specific analysis method of the angle compliance of the oxygen lance is: The angular velocity of the gyroscope at each time point in each period is detected by the gyroscope installed on the oxygen gun, which is recorded as ω if , through the formula Get the angle θ of the oxygen gun in each time period i , Δt represents the duration of the time interval, k represents the number of time points, and the preset oxygen gun standard angle is read from the management database, recorded as θ', and the formula is used to obtain Obtain the angle compliance of the oxygen gun n represents the number of time periods.

8. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 7 is characterized in that: The specific analysis method of the action execution risk assessment coefficient is: Read the operator's feeding compliance, furnace pulling speed abnormality δ, oxygen gun angle compliance Substituting this into the formula The action execution risk evaluation coefficient ξ is obtained, and γ represents the feeding compliance degree of the operator. The weighting factors represent the preset charging compliance, furnace pulling speed abnormality, and oxygen lance angle compliance.

9. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 1 is characterized in that: The specific analysis method of the personnel protection safety analysis module is: In the first step, the image of each worker wearing protective equipment is acquired through the camera, recorded as the image of each worker, and several key feature points are selected on the protective equipment according to the set position. The protective equipment is worn on the human body model to obtain the position coordinates of each key feature point, recorded as the coordinates of each key feature point. At the same time, the coordinates of each key feature point of the protective equipment worn by each worker are extracted from the image of each worker through image recognition technology, recorded as the coordinates of each key feature point of the protective equipment of each worker, and the position deviation of the protective equipment of each worker is obtained by calculation, recorded as d' m ; In the second step, each worker image is divided into several grids of equal area, and the number of grids covered by protective equipment in each worker image is counted. At the same time, the total number of grids in each worker image is extracted. The number of grids covered by protective equipment in each worker image is divided by the total number of grids in each worker image to obtain the protective equipment coverage area of ​​each worker, which is recorded as M. m At the same time, the image of the human body model wearing protective equipment is acquired, recorded as the standard wearing image, and the standard wearing image is analyzed according to the method of analyzing the coverage area of ​​the protective equipment of each operator to obtain the standard coverage area of ​​the protective equipment, recorded as M', and the formula Get the qualified degree θ of the protective equipment coverage area of ​​each operator m ; The third step is to calculate the position deviation d' of each operator's protective equipment. m 、Quality of protective equipment coverage areaθ m Substitute into the formula The personnel protection safety factor χ is obtained, κ1 and κ2 represent the weight factors of the set protection equipment position deviation and coverage area qualification, and q represents the number of operators.

10. The AI-based converter smelting furnace front safety risk monitoring and early warning system according to claim 1 is characterized in that: The specific analysis method of the safety risk comprehensive evaluation module is: Read the regional location risk assessment coefficient λ, action execution risk assessment coefficient ξ, and personnel protection safety factor χ respectively, and substitute them into the formula Obtain the comprehensive evaluation index of furnace safety risk w1, w2, and w3 respectively represent the weight factors of the set regional location risk assessment coefficient, action execution risk assessment coefficient, and personnel protection safety factor. The comprehensive evaluation index of the furnace-front safety risk is compared with the preset threshold of the comprehensive evaluation index of the furnace-front safety risk. If the comprehensive evaluation index of the furnace-front safety risk is greater than or equal to the preset threshold of the comprehensive evaluation index of the furnace-front safety risk, it means that the comprehensive evaluation index of the furnace-front safety risk is qualified, otherwise it means that it is unqualified, and feedback is given to the system.

Images