Discrimination method and discrimination system for illegal work

Abstract

The invention relates to a method and a system for discriminating illegal work, and belongs to the technical field of work management and control. Acquiring information of the current operation by extracting characters in the current operation ticket, if the current operation condition does not have the information of the current operation, judging that the current operation is a ticket-free operation in illegal operations, ending judgment and carrying out illegal operation early warning, otherwise, judging that the current operation is a ticket operation, and carrying out next judgment; acquiring the weather information of the current operation area, studying and judging whether the current weather meets the requirements of the current field operation or not through a weather and operation information association model, if not, judging that the operation is illegal operation, ending judgment and carrying out illegal operation early warning, and if yes, carrying out next judgment; by monitoring the field operation condition in real time, the problem behavior of the field operation is judged in real time, judgment is finished, and illegal operation early warning is carried out. According to the invention, the violation work is discriminated through the violation work discrimination method and the corresponding violation work early warning is carried out, so that the safety risk increase caused by the violation work is avoided.

Description

Technical Field

[0001] The invention relates to a method and a system for distinguishing illegal operations, belonging to the technical field of operation control. Background Art

[0002] Power grid infrastructure projects are characterized by multiple sites, wide areas, complex and changeable environments, many constraints on site, and great management difficulties. With the continuous promotion and application of new equipment, new technologies, and new processes at infrastructure construction sites, not only higher requirements are placed on the operating capabilities and behavioral standardization of infrastructure construction personnel, but also new challenges are posed to construction site safety and quality control.

[0003] At present, power grid infrastructure construction work still mainly relies on manual operation of construction site workers, and there is still a lack of effective means to control construction site work behavior, making it difficult to avoid safety accidents such as personal injury, equipment damage, etc. caused by factors such as poor physical condition, insufficient knowledge and skills, and illegal work behavior. The existing ways to identify and control safety risks in infrastructure construction work are mainly through on-site surveys and manual supervision of personnel work behavior. It has not yet achieved automatic and comprehensive evaluation of the safety work ability and physical health status of infrastructure construction workers before work and organized targeted pre-work drills; it has not yet achieved real-time automatic identification of personnel's physical health status, unsafe behaviors, unsafe conditions of objects, and unsafe factors in the environment during work, and provided active warnings, automatic intervention, and intelligent guidance; it has not yet achieved the full process of the work, multi-perspective inversion and post-evaluation, and it is impossible to intelligently perceive, predict, intervene, and supervise the potential risks and unsafe behaviors of workers. At the same time, in terms of infrastructure construction quality control, it is currently mostly achieved through manual supervision. Since construction quality control based on manual supervision is usually highly subjective, has poor real-time performance, and has high work intensity, such control measures are gradually unable to meet the increasingly sophisticated construction quality control requirements. On the other hand, the growing scale of infrastructure construction, the continuously improving management requirements, and the substandard quality of operators, unauthorized operations, and operations in extreme weather conditions have increased safety risks and constituted the main difficulty in infrastructure safety management.

[0004] The patent document with application publication number CN115496398A discloses a method and system for the safety management of power operations, including establishing a static three-dimensional model of the power operation site; collecting electrical information of each device at the power operation site; collecting video surveillance information, operation information and Beidou positioning information of mobile targets entering the power operation site; performing dangerous situation perception and distance detection based on the collision principle according to the electrical information of each device, the video surveillance information of the mobile target, the operation information of the mobile target and the Beidou positioning information of the mobile target; if the result of the dangerous situation perception is that the mobile target mistakenly enters the live area or the mobile target accidentally touches the live equipment, or the result of the distance detection is that the distance between the mobile target and the live equipment is less than the safe distance, a real-time warning is issued. This invention does not accurately control the safety risks caused by violations such as the substandard ability and quality of the operators, working without tickets, and working in extreme weather. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for distinguishing illegal operations, so as to solve the existing problem of increased safety risks caused by illegal phenomena such as operators' substandard capabilities and qualities, working without a ticket, and working in extreme weather.

[0006] To achieve the above object, the solution of the present invention includes:

[0007] A method for distinguishing illegal operations of the present invention comprises the following steps:

[0008] The operation information is obtained by extracting the text in the operation ticket. If there is no operation information in the current operation, the operation is judged as an unticketed operation in the illegal operation. The judgment is completed and an illegal operation warning is issued. Otherwise, it is a ticketed operation and the next step of judgment is carried out;

[0009] Obtain the weather information of the current operation area, and determine whether the current weather meets the requirements of the current on-site operation through the correlation model of weather and operation information. If not, the operation is judged as illegal operation, the judgment is completed and an illegal operation warning is issued. If it meets the requirements, the next step of judgment is carried out;

[0010] By real-time monitoring of on-site operation conditions, problematic behaviors in on-site operations can be identified in real time, and the judgment is completed and an early warning of illegal operations is issued.

[0011] The present invention extracts the text in the current operation ticket and judges whether the current operation is a ticketless operation in the illegal operation in combination with the current operation situation; the present invention obtains the weather information of the current operation area and accurately judges whether the current weather conditions meet the needs of the on-site operation through the association model. If the weather forecast data obviously does not meet the requirements of the current on-site operation, the operation should be stopped immediately to avoid accidental injuries and losses; the present invention monitors the on-site operation situation in real time, and distinguishes the problem behavior of the on-site operation in real time. When the problem behavior of the on-site operation is distinguished, the distinction ends and an early warning of illegal operation is made; the weather information of the current operation area of ​​the present invention can be obtained using the weather website interface. The present invention distinguishes the ticketless operation, operation under extreme weather and problem behavior operation in the illegal operation through the above method, and makes corresponding warnings of illegal operation, accurately controls the ticketless operation, operation under extreme weather and problem behavior operation, and avoids the above illegal operation leading to increased safety risks; wherein the problem behavior operation is mainly related to the substandard ability and quality of the operator. The present invention distinguishes the illegal operation through the method of distinguishing illegal operation and makes corresponding warnings of illegal operation to avoid the increase of safety risks caused by illegal operation.

[0012] Furthermore, the job information is obtained in the following manner: after preprocessing the image of the job ticket, a region image containing the text position information is obtained through a pre-acquired text detection model, and finally the region image is recognized through a pre-acquired text recognition model.

[0013] The job information of the present invention is obtained by pre-processing the image of the job ticket by grayscale processing and image scaling, and then obtaining a regional image containing text position information through a pre-acquired text detection model, and finally identifying the regional image through a pre-acquired text recognition model to extract the text in the job ticket; wherein the pre-acquired text detection model of the present invention is obtained through data enhancement and data set training.

[0014] Furthermore, after the pre-acquired text recognition model recognizes the regional image, text fusion and residue removal processing are also performed.

[0015] After the pre-acquired text recognition model of the present invention recognizes the regional image, it also performs text fusion and residue removal processing (removing redundant text), providing a more accurate and reliable way to obtain the information of this operation.

[0016] Furthermore, the on-site operation conditions are monitored in real time through sensors and / or cameras.

[0017] The on-site operation conditions of the present invention are monitored in real time through sensors and / or cameras.

[0018] Furthermore, problematic behaviors in field operations include warning behaviors and / or illegal behaviors.

[0019] The problematic behaviors in the field operation of the present invention include warning behaviors and / or illegal behaviors.

[0020] A system for distinguishing illegal operations includes a processor, and the processor is used to execute instructions to implement the following method:

[0021] The operation information is obtained by extracting the text in the operation ticket. If there is no operation information in the current operation, the operation is judged as an unticketed operation in the illegal operation. The judgment is completed and an illegal operation warning is issued. Otherwise, it is a ticketed operation and the next step of judgment is carried out;

[0022] Obtain weather information of the current operation area, and determine whether the current weather meets the requirements of the current on-site operation through the correlation model of weather and operation information. If not, the operation is judged as illegal operation, the judgment is completed and an illegal operation warning is issued. If it meets the requirements, the next step of judgment is carried out;

[0023] By real-time monitoring of on-site operation conditions, problematic behaviors in on-site operations can be identified in real time, and the judgment is completed and an early warning of illegal operations is issued.

[0024] The present invention extracts the text in the current operation ticket and judges whether the current operation is a ticketless operation in the illegal operation in combination with the current operation situation; the present invention obtains the weather information of the current operation area and accurately judges whether the current weather conditions meet the needs of the on-site operation through the association model. If the weather forecast data obviously does not meet the requirements of the current on-site operation, the operation should be stopped immediately to avoid accidental injuries and losses; the present invention monitors the on-site operation situation in real time, and distinguishes the problem behavior of the on-site operation in real time. When the problem behavior of the on-site operation is distinguished, the judgment ends and the illegal operation warning is performed; the weather information of the current operation area of ​​the present invention can be obtained using the weather website interface. The present invention implements the above method by executing instructions through the processor, and then distinguishes the ticketless operation, operation under extreme weather and problem behavior operation in the illegal operation, and makes corresponding warnings for illegal operations, accurately controls the ticketless operation, operation under extreme weather and problem behavior operation, and avoids the above illegal operations causing increased safety risks; wherein the problem behavior operation is mainly related to the substandard ability and quality of the operator. The present invention distinguishes illegal operations through the illegal operation discrimination method and makes corresponding warnings for illegal operations to avoid increased safety risks caused by illegal operations.

[0025] Furthermore, the job information is obtained in the following manner: after preprocessing the image of the job ticket, a region image containing the text position information is obtained through a pre-acquired text detection model, and finally the region image is recognized through a pre-acquired text recognition model.

[0026] The job information of the present invention is obtained by pre-processing the image of the job ticket by grayscale processing and image scaling, and then obtaining a regional image containing text position information through a pre-acquired text detection model, and finally identifying the regional image through a pre-acquired text recognition model to extract the text in the job ticket; wherein the pre-acquired text detection model of the present invention is obtained through data enhancement and data set training.

[0027] Furthermore, after the pre-acquired text recognition model recognizes the regional image, text fusion and residue removal processing are also performed.

[0028] After the pre-acquired text recognition model of the present invention recognizes the regional image, it also performs text fusion and residue removal processing (removing redundant text), providing a more accurate and reliable way to obtain the information of this operation.

[0029] Furthermore, the on-site operation conditions are monitored in real time through sensors and / or cameras.

[0030] The on-site operation conditions of the present invention are monitored in real time through sensors and / or cameras.

[0031] Furthermore, problematic behaviors in field operations include warning behaviors and / or illegal behaviors.

[0032] The problematic behaviors in the field operation of the present invention include warning behaviors and / or illegal behaviors. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flow chart for distinguishing illegal operations of the present invention;

[0034] Figure 2 It is a flow chart of the present invention for performing text recognition on a picture of a work ticket. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0036] An embodiment of a method for distinguishing illegal operations:

[0037] When working on site, first upload the work ticket image, use deep learning text recognition technology to detect and identify the text on the work ticket image, extract key text information such as personnel, equipment, permission time, work area, etc. in the work ticket, and combine the current work situation to determine whether this work is a work without a ticket. The accuracy of key information extraction and recognition is directly related to the correctness of the control unit's judgment of illegal operations. Use the weather website interface to obtain weather information in the current work area, match and compare the acquired weather data with the key work information extracted from the work ticket, and establish a correlation model. For example: outdoor hoisting operations are strictly prohibited in strong winds when the wind speed reaches level 5; outdoor live work is strictly prohibited on rainy days; cable trenches and other confined space operations are strictly prohibited on rainy days, etc., so as to accurately judge whether the current weather conditions meet the needs of on-site operations. If the weather forecast data obviously does not meet the requirements of the current on-site operations, the operation should be stopped immediately to avoid accidental injuries and losses. During the on-site operation, a variety of sensors are used for multi-angle and all-round monitoring and cameras for real-time video monitoring to monitor the on-site operation situation in real time, comprehensively analyze and judge the problematic behaviors during the on-site operation (problematic behaviors refer to alarm problems and violations), and issue early warnings for illegal construction operations to ensure the safety of on-site operations and effectively improve the intelligent management level and operation efficiency of on-site operations. Among them, the alarm behaviors (alarm problems) during the operation process include sensor detection data exceeding the threshold, operator vital sign monitoring data exceeding the threshold, outsider intrusion identification, and staff leaving the work area; illegal behaviors (violation problems) include not wearing a safety helmet, not dressing correctly, key operators are absent, not wearing a safety rope, smoking, making phone calls, etc. The illegal operation identification process based on the operation ticket information and extreme weather is as follows: Figure 1 shown.

[0038] The text recognition process of the work ticket image is as follows Figure 2 As shown in the figure, the uploaded work ticket image is first preprocessed by grayscale processing and image scaling. Then, the preprocessed wiring image is input into the text detection model, which is obtained through data enhancement and data set training to obtain the position information of the text in the wiring image. Finally, the image of the area where text information is detected is input into the text recognition model for recognition. After processing such as text fusion and removal of redundant text, the final text recognition result is obtained.

[0039] The present invention provides a method for distinguishing illegal operations in which the key information of a work ticket is associated with weather data. Based on the text recognition technology, the uploaded work ticket image is intelligently identified to obtain the key information of the work ticket. Combined with the extreme weather information obtained from the weather website, an intelligent judgment is made on whether there are illegal behaviors such as unticketed operations and operations in extreme weather in the on-site operations, thereby realizing unticketed operations, visual monitoring of operations, and early warning of illegal construction operations in extreme weather conditions, enhancing the safety of personnel operations, improving the level of intelligent management and the quality and efficiency of operations, completing the safety supervision of power grid infrastructure sites, and ensuring the important means of safe production of electricity, and ultimately achieving the purpose of avoiding dangers and harmful factors in the operation process and preventing problems before they occur. Deep empowerment and innovation-driven, to create safer, higher-quality, and smarter infrastructure construction sites.

[0040] An embodiment of a system for distinguishing illegal operations:

[0041] A system for distinguishing illegal operations includes a processor, which is used to execute instructions to implement the illegal operation distinguishing method as described above, wherein the illegal operation distinguishing method has been described in detail in an embodiment of the illegal operation distinguishing method and will not be repeated here.

Claims

1. A method for distinguishing illegal operations, characterized in that: The steps include: The operation information is obtained by extracting the text in the operation ticket. If there is no operation information in the current operation, the operation is judged as an unticketed operation in the illegal operation. The judgment is completed and an illegal operation warning is issued. Otherwise, it is a ticketed operation and the next step of judgment is carried out; Obtain weather information of the current operation area, and determine whether the current weather meets the requirements of the current on-site operation through the correlation model of weather and operation information. If not, the operation is judged as illegal operation, the judgment is completed and an illegal operation warning is issued. If it meets the requirements, the next step of judgment is carried out; By real-time monitoring of on-site operation conditions, problematic behaviors in on-site operations can be identified in real time, and the judgment is completed and an early warning of illegal operations is issued.

2. The method for distinguishing illegal operations according to claim 1, characterized in that: The current operation information is obtained in the following manner: after preprocessing the image of the current operation ticket, a region image containing text position information is obtained through a pre-acquired text detection model, and finally the region image is recognized through a pre-acquired text recognition model.

3. The method for distinguishing illegal operations according to claim 2, characterized in that: After the pre-acquired text recognition model recognizes the regional image, it also performs text fusion and residue removal processing.

4. The method for distinguishing illegal operations according to claim 1, characterized in that: The on-site operation conditions are monitored in real time through sensors and / or cameras.

5. The method for distinguishing illegal operations according to claim 1, characterized in that: The problematic behaviors in the field operation include warning behaviors and / or illegal behaviors.

6. A system for distinguishing illegal operations, characterized in that: The invention comprises a processor, wherein the processor is used to execute instructions to implement the following method: The operation information is obtained by extracting the text in the operation ticket. If there is no operation information in the current operation, the operation is judged as an unticketed operation in the illegal operation. The judgment is completed and an illegal operation warning is issued. Otherwise, it is a ticketed operation and the next step of judgment is carried out; Obtain weather information of the current operation area, and determine whether the current weather meets the requirements of the current on-site operation through the correlation model of weather and operation information. If not, the operation is judged as illegal operation, the judgment is completed and an illegal operation warning is issued. If it meets the requirements, the next step of judgment is carried out; By real-time monitoring of on-site operation conditions, problematic behaviors in on-site operations can be identified in real time, and the judgment is completed and an early warning of illegal operations is issued.

7. The system for distinguishing illegal operations according to claim 6, characterized in that: The current operation information is obtained in the following manner: after preprocessing the image of the current operation ticket, a region image containing text position information is obtained through a pre-acquired text detection model, and finally the region image is recognized through a pre-acquired text recognition model.

8. The system for distinguishing illegal operations according to claim 7, characterized in that: After the pre-acquired text recognition model recognizes the regional image, it also performs text fusion and residue removal processing.

9. The system for distinguishing illegal operations according to claim 6, characterized in that: The on-site operation conditions are monitored in real time through sensors and / or cameras.

10. The system for distinguishing illegal operations according to claim 6, characterized in that: The problematic behaviors in the field operation include warning behaviors and / or illegal behaviors.

Images