CN-122002006-A - On-spot intelligence inspection and fortune maintenance system

CN122002006ACN 122002006 ACN122002006 ACN 122002006ACN-122002006-A

Abstract

The invention relates to the technical field of industrial safety inspection and discloses a field intelligent inspection and operation and maintenance system, which adopts a three-layer architecture of a front-end intelligent wearing sensing layer, a middle-end data transmission layer and a rear-end intelligent operation and maintenance platform layer, wherein the front-end wearing layer adopts waterproof and wear-resistant materials, an information interaction module, an AI analysis module, an acousto-optic early warning module, a safety protection module and an image acquisition module can be integrated to complete field data acquisition, abnormal identification and personnel safety protection, the middle-end transmission layer comprises a communication module, an edge calculation gateway and a data cache module to realize real-time safe data transmission and broken network repair transmission, and the rear-end platform layer is provided with a problem visualization module, a correction task management module, a guide and pin item tracking module to realize problem visualization presentation, intelligent dispatch, guide issuing and closed-loop pin item. The intellectualization and safety of the inspection operation and maintenance are greatly improved by constructing a complete closed loop of acquisition, analysis, early warning, rectification and marketing.

Inventors

LEI QIAN
Ya Jishuai
ZHANG YU
ZHAO GUANGMING
DING SHENGXIAO
ZHANG HUASONG
LIAO PENGZHAN
ZHOU SHENDONG
DUANMU QINGWEN
ZHAO XINRUI

Assignees

华电科工股份有限公司

Dates

Publication Date: 20260508
Application Date: 20260109

Claims (10)

1. The on-site intelligent inspection and operation and maintenance system is characterized by comprising a front-end intelligent wearing sensing layer, a middle-end data transmission layer and a rear-end intelligent operation and maintenance platform layer, wherein: the front-end intelligent wearing sensing layer is made of waterproof and wear-resistant materials, and is integrated with an information interaction integration module, an AI analysis module, an acousto-optic early warning module, a safety protection module and an image acquisition module, and is used for acquiring on-site inspection data in real time, identifying abnormality and carrying out personnel safety protection early warning; The middle-end data transmission layer comprises a communication module, an edge computing gateway and a data buffer, and is used for transmitting data acquired by the front-end intelligent wearing sensing layer, an analysis result and early warning information to the rear-end intelligent operation and maintenance platform layer in real time; The back-end intelligent operation and maintenance platform layer comprises a problem visualization display module, a correction task management module and a solution guidance and marketing item tracking module, wherein the solution guidance and marketing item tracking module is used for realizing visual presentation of inspection finding problems, intelligent distribution of correction tasks, issuing of solution guidance and tracking and closed-loop marketing items of the correction whole flow.
2. The system of claim 1, wherein the information interaction integration module comprises: The intercom unit is used for realizing voice communication; The gyroscope unit is used for monitoring the gesture of the patrol personnel; the altimeter unit is used for monitoring the space height of the patrol personnel; A network data transmission unit for supporting multi-mode communication; the positioning unit adopts a satellite and ground enhanced composite positioning technology, is used for acquiring and uploading the three-dimensional space position and the movement track of the patrol personnel in real time, and is provided with an electronic fence function; the identity recognition unit is used for recognizing and verifying the identity of the wearer and calling the corresponding background authority information; The interactive display screen is used for receiving background instructions, displaying on-site early warning information, presenting a correction instruction scheme and feeding back task processing progress.
3. The system of claim 1, wherein the AI analysis module is built-in with an edge computing chip and is configured to perform the following image analysis procedure: The method comprises the steps of extracting key frame images from a real-time video stream provided by an image acquisition module at a preset frame rate, preprocessing the key frame images, synchronously executing the following analysis on the same batch of key frame images after preprocessing the same batch of key frame images based on multithreading or parallel calculation: a first analysis task of invoking an irregular operation recognition model to recognize whether at least one of the actions of wearing protective equipment without regulation and illegal crossing of guardrails and unauthorized operation of unauthorized equipment exists; A second analysis task, namely calling a site dangerous case detection model to identify whether at least one abnormal state of flame, smoke, liquid leakage, equipment shell damage and instrument pointer abnormality exists; And executing a decision according to the analysis result, namely triggering the highest-level emergency response if the confidence level of the dangerous situation detected by the second analysis task exceeds a first threshold value, triggering an alarm associated with the position and the evidence if the illegal action is detected by the first analysis task, and otherwise, sending a conventional status report.
4. The system of claim 3, wherein the nonstandard operation recognition model and the on-site dangerous situation detection model are integrated in the same AI analysis framework, the framework shares the same backbone feature extraction network, and features extracted by the backbone feature extraction network are processed by the feature enhancement module and then respectively input to the special nonstandard operation recognition head and the on-site dangerous situation detection head to output detection results in parallel.
5. The system of claim 2, wherein the security module comprises: The land anti-falling unit is connected with the gyroscope unit and the image acquisition module and is configured to trigger the airbag to be inflated and ejected in a set time after judging that the dumping acceleration and the inclination angle of the human body exceed the preset threshold according to the data acquired by the gyroscope and the image data acquired by the image acquisition module are confirmed in an auxiliary mode; A water area lifesaving unit comprising a water logging detection circuit, a water logging locking circuit, a pressure sensor and an air bag assembly configured to: And when the water resistance of the water logging detection circuit is smaller than a first preset resistance value, starting the pressure sensor to verify the water depth, when the water depth is larger than a preset water depth threshold value, closing the water logging locking circuit, activating the air bag assembly to inflate and expand, and simultaneously automatically activating the water surface positioning beacon, and if the water resistance is larger than or equal to the first preset resistance value, directly returning to a conventional monitoring state.
6. The system according to claim 1, wherein the audible and visual early warning module comprises a buzzer and an LED warning lamp, and drives the buzzer and the LED warning lamp to emit audible and visual warning signals in different modes according to different trigger scenes; The image acquisition module is used for shooting images/videos of the field environment and the equipment state of the patrol personnel according to a preset triggering mode.
7. The system of claim 1, wherein the communication module supports a plurality of wireless communication schemes and automatically switches according to signal quality; the edge computing gateway is deployed on the site side and is configured to preprocess and encrypt the data transmitted by the front-end intelligent wearing sensing layer, and the preprocessing comprises image compression and key early warning information screening; and the data buffer is used for automatically buffering the data to be transmitted by the front-end intelligent wearing sensing layer when the network signal of the communication module is interrupted, and automatically completing data retransmission after the network signal is recovered.
8. The system of claim 1, wherein the back-end intelligent operation and maintenance platform layer comprises: the problem visual display module is used for displaying the position, the track and the problem early warning points of the patrol personnel on the electronic map in real time and generating a patrol statistics billboard; the correction task management module is used for automatically distributing the correction task to the responsible person terminal through an intelligent dispatch algorithm according to the type, the grade and the occurrence position of the early warning information and combining the skill matching degree and the position proximity of the operation and maintenance personnel; The solution guiding module is used for internally arranging a standardized guiding knowledge base classified according to industry and problem types and automatically pushing a matched solution guiding scheme to a responsible person terminal; The sales item tracking module is used for tracking the state of the rectification task, receiving and verifying rectification completion evidence and overdue reminding statistics uploaded by a responsible person, and realizing closed-loop management of the task.
9. The system of claim 8, wherein the intelligent dispatch algorithm calculates dispatch priority scores for candidate responsible persons based on a composite scoring model and assigns tasks based on the scores, the composite scoring model being: W=α·S k +β·L o +γ·E m , The system comprises a task generating position, a task matching degree S k , a task emergency degree E m , a task matching degree S o , a task matching degree S k , a task matching degree S3832 and a task matching degree S3832, wherein W is a comprehensive score, S k is the skill matching degree of the candidate responsible person, L o is the position adjacent degree of the candidate responsible person and the task generating position, E m is the task emergency degree, alpha, beta and gamma are preset weight coefficients; The task allocation flow of the intelligent dispatch algorithm comprises the steps of determining the geographical screening range of initial candidate persons according to the task emergency degree, calculating the comprehensive scores W of the candidate persons in the range and sequencing, dispatching the task to the candidate person with the highest score, and automatically dispatching the task to the next downstream person in a downstream mode when the task is not responded within the preset time.
10. The system of any one of claims 1-9, wherein the back-end intelligent operation and maintenance platform layer is further configured to, when receiving an emergency call signal or a safety protection trigger signal triggered by the front-end intelligent wearing sensing layer, perform a protruding alarm at a platform interface and initiate an emergency linkage procedure.

Description

On-spot intelligence inspection and fortune maintenance system Technical Field The invention relates to the technical field of industrial safety inspection, in particular to a method and a system for intelligent on-site inspection and operation and maintenance. Background In the industrial fields of electric power, chemical industry, construction and the like, on-site inspection is a core link for guaranteeing stable operation of equipment and avoiding safety production accidents. According to the related statistics, most field safety accidents are rooted in that inspection is not in place, irregular operation is not corrected in time, and the limitation of the traditional inspection mode becomes a prominent pain point of industry safety management. The traditional inspection highly depends on manual recording and paper machine account, so that the defects of lag in data input (average lag feedback of the problem exceeds 2 hours), easiness in information omission and the like exist, and real-time management and control on the operation behaviors of inspection personnel are difficult to realize, so that illegal operation is difficult to stop in time. Meanwhile, each link of the problem discovery-correction-marketing item in the existing operation and maintenance flow is disjointed, correction instructions are not timely transmitted, correction progress is not tracked, correction effects are difficult to check, and safety risks are further amplified. The intelligent inspection technology in the current industry is multi-focus single-function optimization, part of equipment only has positioning or image acquisition capability, full-link response from abnormal recognition to early warning intervention cannot be completed, the front-end inspection equipment and the rear-end operation and maintenance system are poor in linkage, a background platform is a static data statistics tool, on-site dynamic and issuing accurate correction guidance cannot be obtained in real time, and a closed-loop management system cannot be formed. Disclosure of Invention In order to solve the technical problems of hysteresis and fragmentation of traditional inspection, the invention provides an on-site intelligent inspection and operation and maintenance system, which can realize linkage of front-end equipment and a back-end operation and maintenance system in real time and realize full-flow closed-loop management of 'problem identification-early warning-rectification-marketing', thereby improving the safety of on-site operation and the intelligent level of operation and maintenance management. In a first aspect, the present invention provides a field intelligent inspection and operation system, including: the perception layer, middle-end data transmission layer and rear end intelligence fortune dimension platform layer are dressed to front end intelligence, wherein: the front-end intelligent wearing sensing layer is made of waterproof and wear-resistant materials, and is integrated with an information interaction integration module, an AI analysis module, an acousto-optic early warning module, a safety protection module and an image acquisition module, and is used for acquiring on-site inspection data in real time, identifying abnormality and carrying out personnel safety protection early warning; The middle-end data transmission layer comprises a communication module, an edge computing gateway and a data buffer, and is used for transmitting data acquired by the front-end intelligent wearing sensing layer, an analysis result and early warning information to the rear-end intelligent operation and maintenance platform layer in real time; The back-end intelligent operation and maintenance platform layer comprises a problem visualization display module, a correction task management module and a solution guidance and marketing item tracking module, wherein the solution guidance and marketing item tracking module is used for realizing visual presentation of inspection finding problems, intelligent distribution of correction tasks, issuing of solution guidance and tracking and closed-loop marketing items of the correction whole flow. The on-site intelligent inspection and operation and maintenance system provided by the embodiment of the invention realizes the full-link management and control from front-end data acquisition to back-end operation and maintenance closed loop of on-site inspection by constructing a three-layer architecture system. The front-end integrated multi-module can synchronously complete data acquisition, abnormal identification and personnel protection, breaks the limitation of single function of the traditional inspection equipment, ensures real-time and safe circulation of data by a middle-end transmission layer, avoids data loss or leakage, realizes integrated management of problem visualization, intelligent dispatch, instruction issuing and sales item tracking by a rear-end platform, and solves the problems in the tr