A safety interlock system and method for a steel rolling operation

Abstract

The application discloses a kind of steel rolling operation safety linkage system and method, including UWB personnel positioning function module A, image acquisition module B, video identification module C, rolling mill equipment operation state tracking control module D, comprehensive analysis and judgment module E, cloud management and control platform F, three-dimensional visualization platform G, safety helmet device H and safety protection net device I;The safety behavior of its operating personnel is tracked by UWB position positioning, simultaneously by image acquisition, machine deep learning and video identification technology to face, billet derailment situation are matched, the safety state of equipment is intelligently identified and tracked in real time, when any security risk occurs, according to the security risk identification of different sources, the security event of different security levels, make different security linkage measures, to form a complete rolling line operation safety linkage mechanism, can effectively achieve all-round safety monitoring, eliminate security risk in time, and start safety measures in time Effect.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of rolling steel safety production technology, in particular to a rolling steel operation safety linkage system and method. BACKGROUND

[0002] After analyzing the types of accidents occurring in rolling steel production, the main types are mechanical injury, gas poisoning, fire, explosion, object impact, scalding, hoisting injury, falling from a height, electric shock injury, etc. The main reasons are concentrated in the following aspects: workers' violation of operation, defects in operation environment conditions, unfamiliarity with safety regulations and operation technology, defects in equipment design and protection devices, and imperfection of safety management rules and regulations and failure to implement them. Therefore, how to eliminate safety hazards in rolling steel production, control rolling steel production accidents, and reduce personnel casualties and property losses has become an important issue in rolling steel safety.

[0003] Rolling steel site operation has problems such as large space, complex and uncontrollable environment, high risk level, and great difficulty in supervision. How to prevent in advance and eliminate safety hazards in the embryonic state will greatly reduce the probability of accidents in production safety of steel enterprises and truly realize intrinsic safety. SUMMARY

[0004] The purpose of the present application is to provide a rolling steel operation safety linkage system and method that can effectively achieve all-round safety monitoring, timely eliminate safety hazards, and timely start safety measures to solve the problems raised in the background technology.

[0005] To achieve the above purpose, the present application provides the following technical solutions:

[0006] A rolling steel operation safety linkage system, comprising a UWB personnel positioning function module A, an image acquisition module B, a video recognition module C, a rolling mill equipment operation state tracking control module D, a comprehensive analysis and judgment module E, a cloud management and control platform F, a three-dimensional visualization platform G, a safety helmet device H, and a safety protection net device I;

[0007] The UWB personnel positioning function module A is used to real-time locate the workers in the rolling steel operation site operation area, obtain the position information of each operator, and transmit it to the comprehensive analysis and judgment module E;

[0008] The image acquisition module B is used to real-time collect image data of the operation area and send it to the video recognition module C;

[0009] The video recognition module C has two model libraries built-in after deep learning of all worker face images and rolling mill steel billet flying steel through a neural network. The two model libraries are identified and distinguished, and the results are encapsulated and sent to the comprehensive analysis and judgment module E;

[0010] The rolling mill equipment operation state tracking management module D is used for receiving the rolling mill operation data returned by the various PLC devices connected with the L1 system, tracking the operation state of all devices, and transmitting the operation state results of each device to the comprehensive analysis and judgment module E in real time; the rolling mill equipment operation state tracking management module D is also used for connecting the safety protection net of each rolling mill device, sending the start and stop instructions of the safety protection net, and receiving the braking signal sent by the comprehensive analysis and judgment module E;

[0011] The comprehensive analysis and judgment module E is used for receiving the messages from the UWB personnel positioning function module A, the video recognition module C, the rolling mill equipment operation state tracking management module D, and the cloud management and control platform F, and giving the corresponding safety linkage measures after analyzing and judging according to a set of local rules and combining the safety linkage strategies of each rule;

[0012] The cloud management and control platform F is used for storing the basic information related to the rolling mill operation of the plant personnel information, operation equipment information, operation area information, camera and the information of the operation area, personnel helmet information, alarm code, alarm type, alarm content, and safety protection net, and sending the basic information to the comprehensive analysis and judgment module E, the image acquisition module B, the video recognition module C, and the UWB personnel positioning function module A when online; the cloud management and control platform F is also used for receiving the alarm messages sent by the comprehensive analysis and judgment module E and performing analysis and processing;

[0013] The three-dimensional visualization platform G is used for receiving the real-time alarm messages sent by the comprehensive analysis and judgment module E, reminding the personnel in the central control room through a sound alarm when receiving the alarm, analyzing the position information, and synchronously displaying on the three-dimensional model for the personnel in the central control room to check the detailed dynamic position of the personnel at the alarm position at any time;

[0014] The safety hat device H is the safety hat worn by each personnel, and a small signal receiver and a sound player are built in each safety hat for receiving the alarm signals of the cloud management and control platform F and the three-dimensional visualization platform G;

[0015] The safety protection net device I is installed around the rolling mill equipment for receiving the braking signals of the rolling mill equipment operation state tracking management module D and the three-dimensional visualization platform G.

[0016] Further, the UWB personnel positioning function module A records the position information of each operator as P1, P2, P3…Pn, Pn={Xi, Yi}, wherein the values of X and Y are the position coordinates of the target detection object at each moment, and Pn represents the UWB base station number.

[0017] Further, the image acquisition module B realizes data acquisition of images by installing corresponding multiple high-definition cameras in the work area. The images acquired by the image acquisition module B are divided into two parts. One part is the images of the on-site workers, and the images of these cameras are marked as P1, P2, …, and Pn. The other part is the images of the theoretical positions of the flying slabs, and the images captured by these cameras are marked as B1, B2, …, and Bn. The theoretical position images are mainly collected from the images near the rolling mill port according to the trajectories of the flying slabs in the past.

[0018] Further, the two model libraries in the video recognition module C are as follows. Each operator has his own sub-model library Person_Name. Each work area has its own sub-model library Mill_AreaCode, and the AreaCode is composed of the production line and the specific work port code. The coding rules of the production line and the work port can be uniformly formulated on the cloud management and control platform F.

[0019] Further, the local analysis and judgment rules in the comprehensive analysis and judgment module E are as follows.

[0020] R1: According to the UWB positioning function, it is judged whether the personnel are in the high-risk work area. The logic is as follows. When receiving the position information sent by the UWB personnel positioning function module A, if any position information Pn(Xi, Yi) coordinate information is in the work area with a danger level of 1, it is judged that the personnel are in the high-risk work area.

[0021] R2: According to the image acquisition face and the video recognition technology, it is judged whether the personnel are in the high-risk work area. The logic is as follows.

[0022] When receiving the information with the Person prefix and the value of match_success sent by the video recognition module C, the Name, Area, and Helmet in the information are analyzed and matched with the Area in the real-time received device running state information. If the matching is successful and the work area device running state Status is 1, it is judged that the personnel are in the high-risk work area.

[0023] R3: According to the image acquisition slab derailment and flying slab video recognition technology, it is judged whether the slab has left the rolling track. The logic is as follows.

[0024] When receiving the information with the Mill prefix and the value of match_success sent by the video recognition module C, it is directly judged that the slab has derailed.

[0025] Further, the local safety policy rules in the comprehensive analysis and judgment module E are as follows.

[0026] S1: If the result of R1 is that the personnel is in a high-risk operation area, the message is encapsulated again, the alarm identifier AlertCode and the judgment result are added, and the location information Area is sent to the cloud management and control platform F, and simultaneously sent to the three-dimensional visualization platform G, wherein the alarm code and the judgment result are one-to-one corresponding;

[0027] S2: If the result of R2 is that the personnel is in a high-risk operation area, the message is encapsulated again, the alarm identifier AlertCode and the judgment result are added, and the location information Area is sent to the cloud management and control platform F, and simultaneously sent to the three-dimensional visualization platform G;

[0028] S3: If the result of R3 is that the billet is off the track, the message is added with the equipment braking instruction and the safety protection net braking instruction, and is sent to the rolling mill equipment operation state tracking management and control module D, and simultaneously sent to the cloud management and control platform F and the three-dimensional visualization platform G.

[0029] Further, the alarm processing measures of the cloud management and control platform F are as follows:

[0030] M1: If the alarm message contains the Person attribute, it means that the message comes from the alarm sent by the video recognition module C, and the measure taken is to send the alarm message to the safety helmet of the personnel recognized, and after the personnel receives the sound alarm prompt sent by the safety helmet, the personnel beside will remind the personnel to leave the dangerous area;

[0031] M2: If the alarm message contains the positon position attribute, a sound alarm is sent to the broadcasting equipment in the operation area, and the alarm content is that the personnel is in a high-risk operation area, please retreat outside the safety warning line;

[0032] M3: If the alarm message contains the Mill attribute, a sound alarm is sent to the broadcasting equipment in the operation area and the adjacent area of the operation area, and a red first-level sound alarm of the rolling line operation area is started, and the alarm content is that the billet is off the track, please evacuate quickly.

[0033] The application provides another technical scheme: a rolling steel operation safety linkage method, comprising the following steps:

[0034] S1: The rolling steel operation safety linkage system is started, and the cloud management and control platform F sends global configuration information to the comprehensive analysis and judgment module E, the UWB personnel positioning function module A, the image acquisition module B and the video recognition module C;

[0035] S2: The base station Pn for monitoring the finishing mill sends the monitored position information Pn={Xi, Yi} to the comprehensive analysis and judgment module E;

[0036] S3: The rolling mill equipment operation state tracking and control module D sends a message to the comprehensive analysis and judgment module E according to the real-time acquired data of the L1 interfaced PLC equipment, and the device at the finishing mill port is in a working state;

[0037] a. Upon receiving the braking signal sent by the comprehensive analysis and judgment module E, the device number and braking instruction are analyzed, and a shutdown signal is sent to the device and the upstream device of the production line where the device is located, and the safety protection net device I around the rolling mill equipment is triggered to rise;

[0038] S4: The video recognition module C outputs different results according to different image acquisition results:

[0039] a. If the image acquisition module B collects images that are not successfully matched by the video recognition model library due to personnel positioning, overlapping, or night factors, jump to S5a for execution;

[0040] b. If the image acquisition module B collects the front faces of any number of operators, which are successfully matched by the video recognition model library, send a message to the comprehensive analysis and judgment module E, and jump to S5b for execution;

[0041] c. Based on S3, if there are signs of billet disengaging from the track, the image acquisition module B matches the collected images with the model library of the video recognition module C, and sends a message to the comprehensive analysis and judgment module E if the matching is successful, and jumps to S5c for execution;

[0042] S5: The operation steps of the comprehensive analysis and judgment module E are as follows:

[0043] a. According to the local R1 judgment rule, the secondary packaged alarm message is sent to the cloud management and control platform F and the three-dimensional visualization platform G, and jumps to S6a for execution;

[0044] b. According to the local R2 judgment rule, the secondary packaged alarm message is sent to the cloud management and control platform F and the three-dimensional visualization platform G, and jumps to S6b for execution;

[0045] c. According to the local R3 judgment rule, the secondary packaged alarm message is sent to the rolling mill equipment operation state tracking and control module D, the cloud management and control platform F, and the three-dimensional visualization platform G, and jumps to S6c and S3a for execution;

[0046] S6: After receiving the alarm, the cloud management and control platform F has the following linkage measures according to the different identifiers of the message:

[0047] a. The alarm contains the positon position attribute, and the regional broadcasting device of the operation area is started, with the sound content being: "Personnel is in a high-risk operation area, please retreat outside the safety warning line";

[0048] b, the alarm message contains a Person attribute, a dangerous alarm message is sent to the safety helmet device H of the personnel, the alarm sound content is: "you are in a high-risk operation area, please retreat outside the safety warning line", after the personnel receive the sound alarm prompt sent by the safety helmet device H, retreat outside the safety warning line;

[0049] c, if the alarm message contains a Mill attribute, a dangerous alarm is sent to the broadcast equipment in the operation area and the adjacent area of the operation area, and a red first-level sound alarm in the rolling line operation area is started, the alarm content is: "billet derailment, please evacuate quickly";

[0050] S7: the three-dimensional visualization platform G receives the real-time alarm message sent by the comprehensive analysis and judgment module E, reminds the personnel in the central control room through a sound alarm at the same time, analyzes the position information in it, and synchronously displays it on the three-dimensional visualization platform G, the personnel in the central control room notifies the personnel in the area through a interphone according to the different types of alarm events, retreats outside the safety line, or starts the area sound broadcast device manually, reminds the personnel to retreat outside the safety line, at the same time, notifies the relevant accident handling personnel, until the central control room confirms that all risks have disappeared, and the safety alarm is released.

[0051] Compared with the prior art, the beneficial effects of the present application are:

[0052] The rolling steel operation safety linkage system and method of the present application not only track the safety behavior of the operation personnel through UWB position positioning, but also match the face and billet derailment condition through image acquisition, machine deep learning and video recognition technology, and realize omnidirectional intelligent identification tracking of the safety state of the equipment in real time, when any safety hazard occurs, different safety linkage measures are made according to different sources of safety hazard identification and different safety levels of safety events, so that a complete rolling line operation safety linkage mechanism is formed, which can effectively achieve the effect of omnidirectional safety monitoring, timely elimination of safety hazards and timely start of safety measures. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 is the system module connection diagram of the present application;

[0054] Figure 2 is the method logic flowchart of the present application. DETAILED DESCRIPTION

[0055] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0056] Please refer to Figure 1 The present application provides a rolling operation safety linkage system, which comprises a UWB personnel positioning function module A, an image acquisition module B, a video recognition module C, a rolling mill equipment operation state tracking management module D, a comprehensive analysis and judgment module E, a cloud management and control platform F, a three-dimensional visualization platform G, a safety helmet device H and a safety protection net device I.

[0057] The UWB personnel positioning function module A is used for real-time positioning of workers in the rolling operation site operation area, obtaining the position information of each operator and transmitting it to the comprehensive analysis and judgment module E. UWB is a relatively new wireless communication technology, mainly suitable for indoor use, which can achieve centimeter-level positioning accuracy, has strong anti-interference ability and penetration, and is an ideal industrial-level positioning technology. In the present application, this technology is applied to real-time positioning of workers in the rolling operation site operation area, obtaining the position information P1, P2, P3…Pn of each operator, Pn={Xi, Yi}, wherein the values of X and Y are the position coordinates of the target detection object at each time, Pn represents the UWB base station number, and is transmitted to the comprehensive analysis and judgment module E.

[0058] The image acquisition module B is used for real-time acquisition of image data of the operation area and sending it to the video recognition module C. Specifically, by installing a high-definition camera in the operation area, the image of the operation area is acquired in real time and returned to the video recognition module C. One operation area may correspond to multiple cameras. The acquired image is divided into two parts. One part is the image of the on-site operator, and the image of these cameras is identified as P1, P2…Pn. The other part is the theoretical position image of the flying billet, and the image captured by these cameras is identified as B1, B2…Bn. The theoretical position here is mainly based on the trajectory of the flying billet in the past, especially focusing on collecting images near the rolling mill port.

[0059] The video recognition module C is built-in two model libraries of all staff face images and billet flying steel depth learning through neural network, and respectively identifies and distinguishes the two model libraries; the identification result of each sub-model library is sent to the comprehensive analysis and judgment module E after encapsulation. Each operator has his own set of sub-model library Person_Name, for example, the model library of Name "Zhang San" is Person_Zhang San; the model library is bound with the operator's name Name, safety helmet number Helmet00000n (n is 0, 1, 2…n), and work area identification, and each operator is bound with his own helmet. If the image recognition module collects the image and the module C performs deep learning identification, it is found that the result of Name "Zhang San" is match_success, otherwise it is match_failed; the information of match_success identification result, personnel Name and work area identification, and the helmet number worn by the personnel are encapsulated into a similar jason format:

[0060] Person: {Name: Zhang San, Area: rolling06#_filigree01, Helmet: Helmet00000n, value: match_success}

[0061] The comprehensive analysis and judgment module E is sent. Each work area has its own set of sub-model library Mill_AreaCode, and AreaCode is composed of production line and specific work port code. The coding rules of production line and work port can be uniformly formulated in cloud platform, for example, the work area of rough rolling port 2# of production line 5# is 5# production line, and the coding code of rough rolling 2# is rough02, so the model library is Mill_rolling05_rough02; if the image collected through identification is found to be Mill_rolling05_rough02, the identification result is match_success, which means that the billet has deviated from the track and the flying steel situation has occurred. Otherwise, it is match_failed; the identification result and work area identification are encapsulated into a similar format as follows:

[0062] Mill: {Area: rolling06#_filigree01, value: match_success}

[0063] The comprehensive analysis and judgment module E is sent.

[0064] The rolling mill equipment operation state tracking and control module D is used to interface two parts, one is the L1 system, used to receive the rolling mill operation data returned by the various PLC devices interfaced by the L1 system, the module divides the production line as a large item, and tracks the running state of all devices, the device is running, the risk level is 1, otherwise it is 0; for example, the finishing 1# port of the 6# production line is running, so the output information is {Area: rolling06#_filigree01, Status: 1}, the running state of each device is transmitted to the comprehensive analysis and judgment module E in real time; the second is to interface the safety protection network of each rolling mill equipment, which can send start and stop instructions to the safety protection network, and receive the brake signal sent by the comprehensive analysis and judgment module E. When receiving the brake signal sent by the comprehensive analysis and judgment module E, analyze the device number and brake instruction, if the stop signal isShutDown is 1, isProcValid: directly trigger the related equipment to stop and all upstream equipment of the production line to stop, if the safety protection network effective signal is 1, trigger the safety protection network device I around the rolling mill equipment to rise.

[0065] The comprehensive analysis and judgment module E is used to receive messages from the UWB personnel positioning function module A, the video recognition module C, the rolling mill equipment operation state tracking and control module D and the cloud management and control platform F, and after a set of local analysis and judgment rules, combined with the safety linkage strategy of each rule, the corresponding safety linkage measures are given; wherein the local analysis and judgment rules are as follows:

[0066] R1: According to the UWB positioning function to judge whether the personnel is in the high-risk operation area, the logic is as follows: when receiving the position information sent by the UWB personnel positioning function module A, if any position information Pn(Xi, Yi) coordinate information is in the operation area with a risk level of 1, it is judged that the personnel is in the high-risk operation area.

[0067] R2: According to the image acquisition face, video recognition technology to judge whether the personnel is in the high-risk operation area, the logic is as follows: when receiving the information with Person at the beginning and value as match_success sent by the video recognition module C, analyze the Name, Area, Helmet in it, and match with the Area in the device running state information received in real time, if the matching is successful, and the operation area device running state Status is 1, the judgment result is: the personnel is in the high-risk operation area.

[0068] R3: According to the image acquisition billet derailment, flying steel and other video recognition technologies to judge whether the billet is off the rolling track, the logic is as follows: when receiving the information with Mill at the beginning and value as match_success sent by the video recognition module C, it is directly judged that the billet is off the track.

[0069] The local security policy rule is as follows:

[0070] S1: If the result of R1 is that the personnel is in a high-risk operation area, the message is encapsulated again, and the alarm identifier AlertCode, the location area information Area, and the judgment result are sent to the cloud management and control platform F, and the three-dimensional visualization platform G. The alarm code and the judgment result (alarm content) correspond one by one. For example, AlertCode:001 represents the judgment result: "personnel is in a high-risk operation risk area", and AlertCode:002 represents the judgment result: "billet derailment".

[0071] The structure of the encapsulated alarm message body is: Alert: {AlertCode:001, AlertContent: "personnel is in a high-risk operation risk area", Area: rolling06#_filigree01, positon: Pn(Xi, Yi)}.

[0072] S2: If the result of R2 is that the personnel is in a high-risk operation area, the message is encapsulated again, the alarm identifier AlertCode and the judgment result are added, and are sent to the cloud management and control platform F, and the three-dimensional visualization platform G.

[0073] The structure of the encapsulated alarm message body is:

[0074] Alert: {AlertCode:001, AlertContent: "personnel is in a high-risk operation risk area", Person: {Name: Zhang San, Area: rolling06#_filigree01, Helmet: Helmet00000n, value: match_success}}.

[0075] S3: If the result of R3 is that the billet is derailed, the message is added with the device braking instruction and the safety protection net braking instruction, and is sent to the rolling mill device operation state tracking and control module D; at the same time, it is sent to the cloud management and control platform F and the three-dimensional visualization platform G.

[0076] The structure of the encapsulated alarm message body is:

[0077] Alert: {AlertCode:002, AlertContent: "billet derailment", isShutDown:1, isProcValid:1, Mill: {Area: rolling06#_filigree01, value: match_success}};

[0078] Wherein, isShutDown represents the device brake instruction, 1 represents the device shutdown, isProcValid represents the safety fence instruction, 1 represents the safety fence rising, and the safety fence takes effect.

[0079] The cloud management and control platform F is used to store basic information related to the rolling mill operation, such as personnel information, operation equipment information, operation area information, camera and belonging operation area information, personnel helmet information, alarm code, alarm type, alarm content, safety fence, etc., and sends these global configuration information to the comprehensive analysis and judgment module E, the image acquisition module B, the video recognition module C, and the UWB personnel positioning function module A when online; if any of the above information changes, the updated message is sent;

[0080] The cloud management and control platform F is also used to receive and analyze the alarm message sent by the comprehensive analysis and judgment module E; the specific alarm processing measures are as follows:

[0081] M1: If the alarm message contains the Person attribute, it means that the message comes from the alarm sent by the video recognition module C. Since the video recognition cannot accurately recognize the overlapping personnel and the face being blocked, etc., the alarm message obtained through the video recognition may only match part of the people. At this time, the measure taken should be to send the alarm message to the helmet of the recognized personnel. After receiving the sound alarm prompt from the helmet, the personnel will remind the person next to him to leave the dangerous area.

[0082] M2: If the alarm message contains the positon position attribute, send a sound alarm to the operation area broadcast device, and the alarm content is: personnel is in a high-risk operation area, please retreat outside the safety warning line.

[0083] M3: If the alarm message contains the Mill attribute, send a sound alarm to the operation area broadcast device and the broadcast device of the adjacent area of the operation area, and turn on the red first-level sound alarm of the rolling line operation area, and the alarm content is: the billet is derailed, please evacuate quickly.

[0084] The three-dimensional visualization platform G is used to receive real-time alarm messages sent by the comprehensive analysis and judgment module E, and at the same time of receiving the alarm, remind the personnel in the control room through the sound alarm, and analyze the position information therein to display on the three-dimensional model synchronously, so that the personnel in the control room can check the detailed dynamic position of the personnel at the alarm at any time.

[0085] The safety helmet device H is a safety helmet worn by each personnel, and each safety helmet is built-in with a small signal receiver and a sound player. When receiving the alarm signal, the sound player plays: you are in a high-risk operation area, please retreat outside the safety line.

[0086] The safety fence device I is installed around the rolling mill equipment, and is used for receiving a brake signal of the rolling mill equipment operation state tracking control module D and the three-dimensional visualization platform G. When receiving the brake signal as 1, the safety fence is raised to prevent the billet from flying out; otherwise, the safety fence is retracted.

[0087] Please refer to Figure 2 In order to further better explain the embodiment of the present application, a rolling operation safety linkage method is also provided, and the application scene is as follows: assuming that the 6# rolling line (coded as rolling06#) is in a normal rolling state, and three persons enter the defined operation area in the operation area of the finishing port 2# (coded as filigree02); the operation steps are as follows:

[0088] Step 1: The rolling operation safety linkage system is started, and the cloud control platform F sends global configuration information to the comprehensive analysis and judgment module E, the UWB personnel positioning function module A, the image acquisition module B and the video recognition module C;

[0089] Step 2: The base station Pn for monitoring the finishing port 2# sends the monitored position information Pn={Xi,Yi} to the comprehensive analysis and judgment module E;

[0090] Step 3: The rolling mill equipment operation state tracking control module D sends a message {Area:rolling06#_filigree02,Status:1} to the comprehensive analysis and judgment module E according to the real-time acquired data of the L1 docking PLC device, and the device of the finishing port 2# is in a working state;

[0091] a, the brake signal sent by the comprehensive analysis and judgment module E is as follows: the device number and the brake instruction are analyzed, Alert: {AlertCode:002,AlertContent:“steel rail derailment”, isShutDown:1,isProcValid:1,Mill: {Area:rolling06#_filigree02,value:match_success}}

[0092] Then, a shutdown signal is sent to the device filigree02 and the upstream device of the production line rolling06, and the safety fence around the rolling mill equipment is raised;

[0093] Step 4: The video recognition module C outputs different results according to different image acquisition results:

[0094] a, if the image acquisition module B collects images due to factors such as personnel station, overlap, or night, and the images are matched by the video recognition model library, and all the matching is unsuccessful, then jump to step 5a for execution;

[0095] b、If image acquisition module B collects the front faces of any number of (assuming three) operators, and after matching with the video recognition model library, all are matched successfully. Send the following message to the comprehensive analysis and judgment module E respectively, and jump to step 5b for execution:

[0096] Person: {Name: Zhang San, Area: rolling06#_filigree01, Helmet: Helmet000001, value: match_success}

[0097] Person: {Name: Li Si, Area: rolling06#_filigree01, Helmet: Helmet000002, value: match_success}

[0098] Person: {Name: Zhao Wu, Area: rolling06#_filigree01, Helmet: Helmet000003, value: match_success}

[0099] c、On the basis of step 3, if at this time there is a sign that the billet has deviated from the track, image acquisition module B matches the collected image through the video recognition model library, and the matching is successful. Send the following message to the comprehensive analysis and judgment module E respectively, and jump to step 5c for execution;

[0100] Mill: {Area: rolling06#_filigree02, value: match_success};

[0101] S5: The operation steps of the comprehensive analysis and judgment module E are as follows:

[0102] a、According to the local R1 judgment rule, the alarm message after secondary packaging is sent to the cloud management and control platform F and the three-dimensional visualization platform G as follows, and jumps to step 6a for execution:

[0103] Alert: {AlertCode: 001, AlertContent: "Personnel is in a high-risk operation area", Area: rolling06#_filigree02, positon: Pn(Xi, Yi)};

[0104] b、According to the local R2 judgment rule, the alarm message after secondary packaging is sent to the cloud management and control platform F and the three-dimensional visualization platform G, and jumps to step 6b for execution:

[0105] Alert: {AlertCode: 001, AlertContent: "Personnel is in a high-risk operation risk area", Person: {Name: Zhang San, Area: rolling06#_filigree02, Helmet: Helmet000001, value: match_success}}

[0106] Alert: {AlertCode: 001, AlertContent: "Personnel is in a high-risk operation risk area", Person: {Name: Li Si, Area: rolling06#_filigree02, Helmet: Helmet000002, value: match_success}}

[0107] Alert: {AlertCode: 001, AlertContent: "Personnel is in a high-risk operation risk area", Person: {Name: Zhao Wu, Area: rolling06#_filigree02, Helmet: Helmet000003, value: match_success}};

[0108] c. According to the local R3 judgment rule, the secondary encapsulated alarm message is sent to the rolling mill equipment operation state tracking management module D, the cloud management and control platform F and the three-dimensional visualization platform G, and steps 6c and 3a are executed respectively;

[0109] Step 6: After the cloud management and control platform F receives the alarm, according to the different identification of the message, there are the following linkage measures:

[0110] a. If the alarm message contains the positon position attribute, then the area broadcast device of the operation area is started, and the sound content is: "Personnel is in a high-risk operation risk area, please retreat outside the safety warning line."

[0111] b. If the alarm message contains the Person attribute, send a dangerous alarm message to the helmets of "Zhang San", "Li Si" and "Zhao Wu", and the alarm sound content is: "You are in a high-risk operation risk area, please retreat outside the safety warning line." After the personnel receive the sound alarm prompt from the helmet, retreat outside the safety warning line.

[0112] c. If the alarm message contains the Mill attribute, send a dangerous alarm to the broadcast devices in the operation area and the adjacent area of the operation area, and start the red level one sound alarm of the rolling line operation area. The alarm content is: "Steel billet derailment, please evacuate quickly."

[0113] Step 7: The three-dimensional visualization platform G receives the real-time alarm message sent by the comprehensive analysis and judgment module E, and at the same time reminds the personnel in the central control room through a sound alarm, analyzes the position information therein, and synchronously displays on the three-dimensional visualization platform G; the personnel in the central control room retreats outside the safety line according to the different types of alarm events in real time through a intercom, or retreats outside the safety line through a manual area sound broadcast device, and notifies the relevant accident handling personnel until the central control room confirms that all risks have disappeared and the safety alarm is released.

[0114] In summary: the rolling operation safety linkage system and method provided by the application makes full use of the advantages of UWB positioning technology in indoor positioning, combines the very mature video recognition technology at present, more comprehensively, timely and accurately locates the level of safety events, and then combines the safety linkage method deduced according to different safety positioning sources and different safety alarm types, which is proposed by the application, to well avoid the occurrence of rolling operation safety accidents.

[0115] The above is only the preferred specific embodiment of the application, but the protection scope of the application is not limited to this, any person skilled in the art in the technical range disclosed by the application, according to the technical scheme and the inventive concept of the application, makes equivalent replacement or change, which should be covered in the protection scope of the application.

Claims

1. A safety interlock system for a steel rolling operation, characterized by, The UWB personnel positioning function module A, the image acquisition module B, the video recognition module C, the rolling mill equipment operation state tracking control module D, the comprehensive analysis and judgment module E, the cloud management and control platform F, the three-dimensional visualization platform G, the safety helmet device H, and the safety protection net device I are included. The UWB personnel positioning function module A is used for real-time positioning of workers in the rolling mill operation site operation area, obtaining the position information of each operator and transmitting it to the comprehensive analysis and judgment module E. The image acquisition module B is used for real-time acquisition of image data of the operation area and sending it to the video recognition module C. The video recognition module C has two model libraries built-in after deep learning of all worker face images and rolling mill billet flying steel through neural networks, and distinguishes them by identifying the two model libraries, and sends the results to the comprehensive analysis and judgment module E after encapsulation. The rolling mill equipment operation state tracking control module D is used for receiving rolling mill operation data returned by various PLC devices connected by the L1 system, tracking the operation state of all devices, and transmitting the operation state results of each device to the comprehensive analysis and judgment module E in real time. The comprehensive analysis and judgment module E is used for receiving messages from the UWB personnel positioning function module A, the video recognition module C, the rolling mill equipment operation state tracking control module D, and the cloud management and control platform F, and after a set of analysis and judgment rules are applied, combined with the safety linkage strategy of each rule, the corresponding safety linkage measures are given. The cloud management and control platform F is used to store basic information related to rolling mill operation, such as personnel information, operation equipment information, operation area information, camera and related operation area information, personnel helmet information, alarm code, alarm type, alarm content, and safety protection net information, and send these basic information to the comprehensive analysis and judgment module E, the image acquisition module B, the video recognition module C, and the UWB personnel positioning function module A when online. The three-dimensional visualization platform G is used to receive real-time alarm messages sent by the comprehensive analysis and judgment module E, and to analyze the location information in the alarm messages and display them on the three-dimensional model in real time. The safety helmet device H is a safety helmet worn by each person, and each safety helmet has a small signal receiver and a sound player built-in, which is used to receive alarm signals from the cloud management and control platform F and the three-dimensional visualization platform G. The safety protection net device I is installed around the rolling mill equipment, and is used to receive braking signals from the rolling mill equipment operation state tracking control module D and the three-dimensional visualization platform G.

2. A safety interlock system for a steel rolling operation as defined in claim 1 wherein: The UWB personnel positioning function module A acquires position information records of each operator as P1, P2, P3...Pn, Pn={Xi, Yi}, wherein the values of X and Y are position coordinates of the target detection object at each time, and Pn represents the UWB base station number.

3. A safety interlock system for a steel rolling operation as defined in claim 1 wherein: The image acquisition module B realizes image data acquisition by installing corresponding multiple high-definition cameras in the working area. The acquired images are divided into two parts. One part is the image of the on-site operator, and the image identification of these cameras is P1, P2...Pn. The other part is the theoretical position image of the flying billet, and the image identification of these cameras is B1, B2...Bn. The theoretical position image is mainly collected from the image near the rolling mill port according to the track of the flying billet in the past.

4. A safety interlock system for a steel rolling operation as defined in claim 1 wherein: The two model libraries in the video recognition module C are as follows: each operator has his own sub-model library Person_Name; and each working area has its own sub-model library Mill_AreaCode. The AreaCode is composed of the production line and the specific working port code. The coding rules of the production line and the working port can be uniformly formulated in the cloud management and control platform F.

5. A safety interlock system for a steel rolling operation as defined in claim 4 wherein: The local analysis and judgment rules in the comprehensive analysis and judgment module E are as follows: R1: According to the UWB positioning function, it is judged whether the personnel are in the high-risk working area. The logic is as follows: after receiving the position information sent by the UWB personnel positioning function module A, if any position information Pn(Xi, Yi) coordinate information is in the working area with a danger level of 1, it is judged that the personnel are in the high-risk working area. R2: According to the image acquisition face and video recognition technology, it is judged whether the personnel are in the high-risk working area. The logic is as follows: After receiving the information with the Person prefix and the value of match_success sent by the video recognition module C, the Name, Area, and Helmet in the information are analyzed and matched with the Area in the real-time received device running state information. If the matching is successful and the working area device running state Status is 1, the judgment result is that the personnel are in the high-risk working area. R3: According to the image acquisition billet derailment and flying steel video recognition technology, it is judged whether the billet is off the rolling track. The logic is as follows: After receiving the information with the Mill prefix and the value of match_success sent by the video recognition module C, it is directly judged that the billet is off the track.

6. A safety interlock system for a steel rolling operation as defined in claim 5 wherein: The local safety policy rules in the comprehensive analysis and judgment module E are as follows: S1: If the judgment result of R1 is that the personnel are in the high-risk working area, the message is encapsulated twice, the alarm identification AlertCode is added, the position information Area and the judgment result are sent to the cloud management and control platform F, and the three-dimensional visualization platform G is also sent. S2: If the judgment result of R2 is that the personnel are in the high-risk working area, the message is encapsulated twice, the alarm identification AlertCode and the judgment result are added and sent to the cloud management and control platform F, and the three-dimensional visualization platform G is also sent. S3: If the result of the above R3 is that the billet is off the track, add the device braking instruction and the safety fence braking instruction to the message and send it to the rolling mill device operation state tracking and control module D, and send it to the cloud management and control platform F and the three-dimensional visualization platform G.

7. A safety interlock system for a steel rolling operation as defined in claim 1 wherein: The alarm handling measures of the cloud management and control platform F are as follows: M1: If the alarm message contains the Person attribute, it means that the message comes from the alarm sent by the video recognition module C. The measure taken is to send an alarm message to the safety helmet of the identified personnel. After receiving the sound alarm prompt from the safety helmet, the personnel will remind the people around them to leave the dangerous area together; M2: If the alarm message contains the positon attribute, send a sound alarm to the work area broadcast device, and the alarm content is: personnel are in a high-risk work area, please retreat outside the safety warning line; M3: If the alarm message contains the Mill attribute, send a sound alarm to the work area and the adjacent area of the broadcast device, and turn on the red level one sound alarm of the rolling line work area, and the alarm content is: the billet is off the track, please evacuate quickly.

8. A method of safety interlocking of a steel rolling operation, characterized in that, The steps include: S1: The rolling steel operation safety linkage system starts, and the cloud management and control platform F sends global configuration information to the comprehensive analysis and judgment module E, the UWB personnel positioning function module A, the image acquisition module B and the video recognition module C; S2: The base station Pn for monitoring the finishing mill sends the monitored position information Pn={Xi, Yi} to the comprehensive analysis and judgment module E; S3: The rolling mill device operation state tracking and control module D sends a message to the comprehensive analysis and judgment module E according to the real-time acquired data of the L1 interfaced PLC device, and the finishing mill device is in a working state; a. Receive the braking signal sent by the comprehensive analysis and judgment module E, analyze the device number and braking instruction in it, and send a shutdown signal to the device and the upstream device of the production line where the device is located, and trigger the safety fence device I around the rolling mill device to rise; S4: The video recognition module C outputs different results according to different image acquisition results: a. If the image acquisition module B collects images due to personnel positioning, overlapping, or night factors, and the images are not matched successfully after being matched by the video recognition model library, jump to S5a for execution; b. If the image acquisition module B collects the front faces of any number of operators, and the matching is successful after being matched by the video recognition model library, send a message to the comprehensive analysis and judgment module E, and jump to S5b for execution; c. On the basis of S3, if there are signs that the billet has deviated from the track, the image acquisition module B matches the collected images through the model library of the video recognition module C, and the matching is successful, sends a message to the comprehensive analysis and judgment module E, and jumps to S5c for execution; S5: The operation steps of the comprehensive analysis and judgment module E include the following: a. According to the local R1 judgment rule, the twice encapsulated alarm message is sent to the cloud management and control platform F and the three-dimensional visualization platform G, and jumps to S6a for execution; b、According to the local R2 judgment rule, the twice encapsulated alarm message is sent to the cloud management and control platform F and the three-dimensional visualization platform G, and jumps to the execution of S6b; c、According to the local R3 judgment rule, the twice encapsulated alarm message is sent to the rolling mill equipment operation state tracking management and control module D, the cloud management and control platform F and the three-dimensional visualization platform G, and jumps to the execution of S6c and S3a respectively; S6: After the cloud management and control platform F receives the alarm, according to the different identification of the message, there are the following several linkage measures: a、The alarm message contains the positon position attribute, the area broadcast device in the operation area is started, and the sound content is: "personnel is in the high-risk operation area, please retreat outside the safety warning line"; b、The alarm message contains the Person attribute, sends the dangerous alarm message to the safety helmet device H of personnel, and the alarm sound content is: "you are in the high-risk operation area, please retreat outside the safety warning line", personnel receives the sound alarm prompt sent by the safety helmet device H, and retreats outside the safety warning line; c、The alarm message contains the Mill attribute, then sends the dangerous alarm to the broadcast device in the operation area and the adjacent area of the operation area, and starts the red first-level sound alarm in the rolling line operation area, and the alarm content is: "billet derailment, please evacuate the operation personnel quickly"; S7: The three-dimensional visualization platform G receives the real-time alarm message sent by the comprehensive analysis and judgment module E, and at the same time reminds the control room personnel through the sound alarm, and analyzes the position information in it, and synchronously displays on the three-dimensional visualization platform G, the control room personnel according to the different types of alarm events, through the intercom, informs the personnel in the area to retreat outside the safety line, or through the manual area sound broadcast device, reminds the personnel to retreat outside the safety line, at the same time, informs the relevant accident handling personnel, until the control room confirms that all risks have disappeared, and the safety alarm is released.

Images