Special operation safety monitoring method and system

Abstract

The present application relates to the technical field of industrial safety production monitoring, and provides a special operation safety monitoring method and system, which comprises the following steps: obtaining permission constraint conditions from an electronic operation permit, generating a three-dimensional dynamic monitoring space by risk expansion of a permission space range according to an operation process type, determining a patrol observation task of a mobile monitoring device, controlling the mobile monitoring device to perform the patrol observation task to collect on-site sensing data, and verifying the on-site sensing data in combination with the permission constraint conditions to obtain a verification result containing an operation consistency verification result, and outputting a safety control instruction to a safety control device according to a mapping relationship between a permission identifier and an associated device identifier when the verification result meets a preset condition. The present application realizes accurate alignment of a monitoring range and permission semantics and physical risks, and forms a closed loop from risk space reconstruction to compliance verification and then to safety hard control.

Description

Technical Field

[0001] This invention relates to the field of industrial safety production monitoring technology, specifically to a method and system for special operation safety monitoring. Background Technology

[0002] In heavy industries such as shipbuilding, metallurgy, and petrochemicals, special operations are high-frequency and high-risk operations, typically requiring dedicated personnel or auxiliary equipment for on-site monitoring. Existing special operation monitoring solutions mostly employ fixed electronic fences or general visual detection algorithms to monitor the work area. However, this monitoring method usually relies solely on static geographical boundaries or general flame and smoke characteristics for judgment, failing to dynamically correlate the multi-dimensional semantic information contained in the electronic work permit, such as process type, height level, and associated equipment, with the actual physical risk distribution on-site. This results in discrepancies between the sensing range of the monitoring equipment and the authorized scope and actual hazardous areas, easily leading to false alarms or missed alarms. Furthermore, it is difficult to identify deeper violations such as unauthorized personnel working on-site or process mismatches, failing to form a precise safety closed loop deeply coupled with the semantics of the permit. Summary of the Invention

[0003] To address the problem of existing technologies where the scope of monitoring perception is disconnected from the semantics of permits and physical risks, this invention provides a special operation safety monitoring method, system, and medium that enables dynamic spatial construction, operation consistency verification, and a closed-loop safety hardware control based on the semantics of electronic work permits.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: A method for safety monitoring of special operations includes: obtaining permit constraints from an electronic work permit, the permit constraints including at least a permitted spatial range, a work process type, a permit validity period window, an associated equipment identifier, a permit identifier, and a mapping relationship between the permit identifier and the associated equipment identifier; expanding the permitted spatial range for risk based on the work process type to generate a three-dimensional dynamic monitoring space, and determining a patrol observation task for a mobile monitoring device in the three-dimensional dynamic monitoring space; controlling the mobile monitoring device to execute the patrol observation task to collect on-site perception data, and verifying the data by combining the permit constraints and the on-site perception data to obtain a verification result that includes at least a work consistency verification result; and when the verification result meets preset conditions, outputting a safety control command to a safety control device according to the mapping relationship between the permit identifier and the associated equipment identifier.

[0005] Optionally, the special operation includes hot work, and the operation process type includes gas cutting, electric welding, or grinding; the step of expanding the permitted space range according to the operation process type to generate a three-dimensional dynamic monitoring space includes: determining a horizontal risk expansion area according to the operation process type, wherein the planar dimension of the permitted space range is the permitted planar range; determining a vertical risk expansion area according to the height dimension of the permitted space range, wherein the height dimension is the permitted height range; and spatially enclosing the permitted planar range, the horizontal risk expansion area, and the vertical risk expansion area to generate the three-dimensional dynamic monitoring space.

[0006] By introducing process type and height dimension as the basis for risk expansion, the original static two-dimensional permit plane is transformed into a three-dimensional dynamic monitoring space containing horizontal and vertical risk dimensions. This allows the patrol and observation range of mobile monitoring equipment to be adaptively adjusted according to the physical characteristics of the operation process (such as splash radius and flue gas rising height), thereby solving the technical problem that fixed fences cannot match the actual physical risk distribution and improving the spatial accuracy of monitoring.

[0007] Optionally, the verification result includes a job consistency verification result, and the method for generating the job consistency verification result includes: extracting the on-site job object identifier and job process features from the on-site perception data; comparing the on-site job object identifier and job process features with the authorized job information in the permission constraints to generate the job consistency verification result.

[0008] By technically comparing the object identifiers and process characteristics collected on-site with the authorization information in the licensing constraints, a dimension for verifying operational compliance that is independent of simple environmental perception is constructed. This dimension can effectively identify hidden violations such as personnel substitution and unauthorized process changes, eliminating the blind spot of traditional monitoring that can only detect environmental anomalies but cannot verify the compliance of operational behaviors.

[0009] Optionally, when the verification result meets the preset conditions, outputting a security control command to the security control device according to the mapping relationship between the license identifier and the associated device identifier includes: when the verification result meets the preset conditions, generating a security control command carrying the mapping relationship and outputting it to the security control device, so that the security control device performs a cut-off operation or a restriction operation on the device corresponding to the associated device identifier according to the mapping relationship, and maintains the locked state of the device corresponding to the associated device identifier until the preset reset condition is met.

[0010] Optionally, determining the horizontal risk expansion area based on the operation process type includes at least one of the following: determining the spark splash prediction area outside the permissible plane range as the horizontal risk expansion area based on the spark splash parameters corresponding to the operation process type and the permissible height range; acquiring real-time micro-meteorological data and determining the downwind hazard diffusion prediction area as the horizontal risk expansion area by combining the real-time micro-meteorological data and the combustible material diffusion parameters corresponding to the operation process type.

[0011] Optionally, the authorized operation information includes an authorized personnel identifier and an authorized process type; the on-site operation object identifier includes a personnel identity identifier; the operation process characteristics include acoustic features or arc light features; the step of comparing the on-site operation object identifier and the operation process characteristics with the authorized operation information in the permission constraints to generate the operation consistency verification result includes: comparing the personnel identity identifier with the authorized personnel identifier to generate a personnel comparison result; when the operation process characteristics include acoustic features and / or arc light features, comparing the corresponding features with the reference features corresponding to the authorized process type to generate a process comparison result; and generating the operation consistency verification result based on the personnel comparison result and / or the process comparison result.

[0012] Optionally, the safety control device includes a safety programmable logic controller or a hardwired safety loop; the preset reset conditions include receiving manual reset confirmation information and the electronic work permit being updated to an effective status.

[0013] Optionally, determining the patrol observation task of the mobile monitoring device for the three-dimensional dynamic monitoring space includes: projecting the three-dimensional dynamic monitoring space onto the ground layer to determine the projected monitoring area, and generating a candidate patrol point sequence based on the boundary of the projected monitoring area; obtaining the drivability cost map of the mobile monitoring device, and using the drivability cost map to filter out the patrol point sequence that meets preset drivability conditions from the candidate patrol point sequence; and combining the horizontal risk expansion area and the vertical risk expansion area to determine the observation attitude constraints and dwell time of each patrol point in the patrol point sequence.

[0014] This invention also provides a special operation safety monitoring system, comprising: a permit parsing module, used to obtain permit constraints from an electronic work permit, the permit constraints including at least a permitted spatial range, a work process type, a permit validity time window, an associated equipment identifier, a permit identifier, and a mapping relationship between the permit identifier and the associated equipment identifier; a dynamic space generation module, used to extend the permitted spatial range according to the work process type to generate a three-dimensional dynamic monitoring space, and determine the patrol observation task of a mobile monitoring device for the three-dimensional dynamic monitoring space; a data acquisition and verification module, used to control the mobile monitoring device to perform the patrol observation task to acquire on-site perception data, and to perform verification in combination with the permit constraints and the on-site perception data to obtain a verification result that includes at least the operation consistency verification result; and a safety response module, used to output a safety control command to a safety control device according to the mapping relationship between the permit identifier and the associated equipment identifier when the verification result meets preset conditions.

[0015] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the special operation safety monitoring method as described above.

[0016] Beneficial effects: The special operation safety monitoring method and system provided by this invention parses the permit constraints from the electronic work permit and generates a three-dimensional dynamic monitoring space by risk expansion of the permit space range according to the type of operation process. This enables the patrol and observation tasks of mobile monitoring equipment to no longer be limited to static geofences, but to dynamically adapt to real physical risk areas according to process characteristics. On this basis, multi-dimensional verification including operation consistency check is performed by combining on-site perception data and permit constraints. In case of anomalies, safety control instructions are output based on the mapping relationship between permit identifiers and equipment identifiers. This achieves accurate alignment between the monitoring range and permit semantics and physical risks, effectively solving the problems of false alarms, missed alarms and weak violation identification capabilities caused by the disconnect between perception and permits in the prior art. It forms a complete technical closed loop from risk space reconstruction to compliance verification and then to safety hardware control. Attached Figure Description

[0017] Figure 1 This is a flowchart of a special operation safety monitoring method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the constituent modules of the special operation safety monitoring system according to an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0020] like Figure 1 As shown, this embodiment provides a special operation safety monitoring method. The executing entity of this method can be a processor or control system deployed on an edge computing node, cloud server, or local mobile monitoring device. The application object is a mobile monitoring device, which can be a legged robot, wheeled robot, tracked robot, or drone, or other intelligent terminal with autonomous movement and perception capabilities. It should be understood that although this embodiment uses a legged robot as the main example for illustration, this is not a limitation on the configuration of the mobile monitoring device; any mobile carrier capable of on-site data collection and patrol observation functions is applicable. In this embodiment, the special operation is a hot work operation, and the special safety monitoring method includes the following steps: Step S100: Obtain the license constraints from the electronic work permit. The license constraints include at least the licensed spatial range, work process type, license validity period window, associated equipment identifier, license identifier, and the mapping relationship between the license identifier and the associated equipment identifier.

[0021] Specifically, the electronic work permit, as an electronic credential for administrative approval, provides structured constraint inputs for all subsequent monitoring activities. The processor retrieves structured permit data in real time from the work permit management system via API interface or message bus. The permit spatial range defines the legal geographical boundaries of the work, typically represented by a two-dimensional polygon coordinate sequence and height range; the work process type specifies the specific nature of the work, such as gas cutting, electric welding, grinding, etc., with different processes corresponding to different physical risk characteristics; the permit validity time window specifies the start and end time interval of the hot work, determining whether the current time is within the valid work period; work exceeding the time window is considered an invalid permit and triggers corresponding verification results; the associated equipment identifier lists the unique codes of equipment such as welding machines, gas cylinder valves, and power supply circuits bound to this permit; and the permit identifier is a globally unique index for this work. Furthermore, the permit constraints also include a fence version number, used to identify the current version of the permit spatial range. When the permit spatial range changes, the fence version number is updated accordingly to ensure that subsequent monitoring logs can be accurately associated with the specific version of the permit range. By structuring and parsing the above multidimensional information, abstract management systems can be transformed into machine-understandable digital constraints, providing accurate input benchmarks for subsequent dynamic space generation and consistency verification. This solves the problem of false alarms and missed alarms caused by the disconnect between the perception system and permission information in traditional monitoring solutions.

[0022] Step S200: Expand the permitted space range according to the type of operation process to generate a three-dimensional dynamic monitoring space, and determine the patrol and observation tasks of the mobile monitoring equipment for the three-dimensional dynamic monitoring space.

[0023] Specifically, the process of expanding the permitted space range based on the type of work process to generate a three-dimensional dynamic monitoring space includes: determining the horizontal risk expansion area based on the type of work process, where the planar dimension of the permitted space range is the permitted plane range; determining the vertical risk expansion area based on the height dimension of the permitted space range, where the height dimension is the permitted height range; and spatially enclosing the permitted plane range, the horizontal risk expansion area, and the vertical risk expansion area to generate the three-dimensional dynamic monitoring space. This risk modeling method based on the physical characteristics of the process is not a simple geometric stretching. For example, for high-altitude gas cutting operations, the vertical risk expansion area needs to cover the vertical depth to which molten slag may fall, while the horizontal risk expansion area needs to cover the spatter radius.

[0024] In a specific example, determining the horizontal risk expansion area based on the type of work process can be done in at least one of the following ways: determining the predicted area of ​​spark fall outside the permissible plane range as the horizontal risk expansion area based on the spark splash parameters and permissible height range corresponding to the type of work process; or, obtaining real-time micro-meteorological data and combining the real-time micro-meteorological data with the combustible material diffusion parameters corresponding to the type of work process to determine the downwind hazard diffusion prediction area as the horizontal risk expansion area.

[0025] Specifically, the generation of the spark splash prediction zone is based on a projectile motion model. This model uses the work point as the launch origin and comprehensively considers the work point height (taken from the upper limit of the permitted height range), the initial velocity distribution of the sparks (the initial velocity ranges for gas cutting, electric welding, and grinding are approximately 3-8 m / s, 1-4 m / s, and 5-12 m / s, respectively), the splash angle distribution (usually concentrated in the horizontal range of ±30° to ±60°), and the air resistance coefficient. The model calculates the probability density distribution of the molten slag's impact point on the ground through Monte Carlo simulation or analytical solution. Areas where the impact probability density exceeds a preset threshold (such as the contour line with a cumulative probability of 90%) are identified as the spark splash prediction zone, thus including high-risk impact areas within the monitoring scope. Different work process types correspond to different splash parameter sets, and the system automatically matches the corresponding parameter configuration according to the work process type in the permitted constraints.

[0026] The downwind hazard diffusion prediction zone is based on a Gaussian plume diffusion model, using on-site wind speed and direction sensor data to predict the drift path and concentration field of combustible gases or smoke. This model uses the leak source or work site as the origin and combines real-time wind speed (affecting the diffusion rate), wind direction (determining the drift direction), atmospheric stability level, and other micrometeorological parameters to calculate the predicted concentration of combustible gases at various downwind locations. Areas where the predicted concentration exceeds a preset safety threshold (e.g., 10% of the lower explosive limit of combustible gases) are designated as the downwind hazard diffusion prediction zone.

[0027] Furthermore, when the generated downwind hazard diffusion prediction zone and the preset restricted area spatially overlap, regardless of whether the current on-site sensing data shows that the flammable gas concentration has reached the alarm threshold, a hazard status verification result is generated and incorporated into the verification result to trigger an early warning or safety control action. The preset restricted area here can be a work area designated by other electronic work permits, an area containing fixed hazard sources (such as oil depots or gas cylinder storage areas), or a densely populated area. This prediction-based rather than measurement-based early triggering mechanism can proactively take measures to drive away or cut off hazard when the physical diffusion trend clearly points to a sensitive area before the hazard actually occurs, effectively compensating for the lag in traditional solutions that only respond after the concentration exceeds the limit.

[0028] After determining the three-dimensional dynamic monitoring space, specific patrol strategies need to be planned. Specifically, the patrol observation tasks of the mobile monitoring equipment for the three-dimensional dynamic monitoring space include: projecting the three-dimensional dynamic monitoring space onto the ground layer to determine the projected monitoring area, and generating a sequence of candidate patrol points based on the boundaries of the projected monitoring area; obtaining the drivability cost map of the mobile monitoring equipment, and using the drivability cost map to select a sequence of patrol points from the candidate patrol point sequences that meet the preset drivability conditions; and determining the observation attitude constraints and dwell time of each patrol point in the patrol point sequence by combining the horizontal risk expansion area and the vertical risk expansion area.

[0029] The candidate patrol point sequence is generated as follows: the three-dimensional dynamic monitoring space is projected onto the ground layer to obtain the projected monitoring area. A preset distance (e.g., 0.5-1.5m) is offset outwards along the boundary polygon of this projected monitoring area to ensure that the risk area outside the guardrail is also covered. Then, samples are taken at equal intervals according to arc length (e.g., one sampling point is set every 2-3m) to generate the candidate patrol point sequence. For patrol points falling into the predicted Martian splashdown area or the predicted downwind hazard diffusion area, the sampling density can be increased to improve monitoring coverage.

[0030] The accessibility cost map is generated by mobile monitoring equipment at the work site, reflecting the equipment's adaptability to the terrain. For legged robots, the map at least distinguishes between areas with treadable steel plates (low cost, priority access), ditch no-go zones (infinite cost, impassable), and oily / mud-affected zones (medium cost, cautious access). For wheeled robots, it distinguishes between flat surfaces and step obstacles. The mapping can be generated in real-time by a depth camera or LiDAR mounted on the mobile monitoring equipment, or it can be loaded from a pre-built digital twin model of the factory area. When selecting patrol point sequences, candidate points with a cost exceeding a preset access threshold are removed, and the remaining points are replaced with alternative points on the detour paths between adjacent reachable points to ensure the continuity and integrity of the patrol paths.

[0031] The observation attitude constraints ensure that the sensor's field of view covers high-risk areas. For example, when the patrol point is located at the edge of the predicted Martian splashdown zone, the gimbal must maintain a specific elevation angle to monitor falling objects above; when the patrol point is located in the downwind danger spread prediction zone, the gimbal must be oriented towards the spread source. The dwell time is dynamically adjusted according to the risk level of the area where the patrol point is located. The dwell time is longer in high-risk areas (such as the center of the predicted Martian splashdown zone) (e.g., 30-60 seconds), and shorter in low-risk areas (such as the edge of the permissible plane) (e.g., 10-20 seconds). In this way, the patrol mission is no longer a blind path traversal, but a directional observation deeply coupled with the risk distribution, improving monitoring efficiency and coverage.

[0032] Step S300: Control the mobile monitoring device to perform patrol observation tasks to collect on-site perception data, and perform verification by combining the permission constraints and on-site perception data to obtain a verification result that includes at least the operation consistency verification result.

[0033] The acquisition of on-site sensing data relies on various sensors mounted on mobile monitoring equipment, including but not limited to infrared thermal imaging sensors, visible light cameras, combustible gas sensors, acoustic sensors, and micro-meteorological sensors. In a specific example, the infrared thermal imaging sensor and the visible light camera are configured in a dual-beam coaxial configuration, enabling pixel-level registration between the infrared thermal image and the visible light image. Using this registration result, hotspots detected in the infrared thermal image can be back-projected onto the world coordinate system and bound to the corresponding grid cells in the three-dimensional dynamic monitoring space, thereby achieving precise location and spatial attribution of temperature anomalies. This dual-beam fusion mechanism avoids the limitation of a single infrared sensor failing to accurately locate the spatial position of hotspots, and also avoids the limitation of a single visible light camera being unable to penetrate smoke or identify hidden heat sources.

[0034] In addition to the environmental perception data mentioned above, the on-site perception data also includes the status data of the mobile monitoring device itself. For legged robots, the status data includes IMU vibration data and foot contact force distribution data. The system determines whether the robot is currently in a stable standing state based on the IMU vibration data, and evaluates the uniformity of force on each supporting leg in combination with the foot contact force distribution data. When the IMU vibration amplitude exceeds the preset stability threshold or the foot force distribution is severely uneven, the robot is determined to be in an unstable standing state. At this time, the confidence weight of the current infrared thermal imaging data and combustible gas concentration data is reduced to prevent false alarms caused by body shaking; at the same time, the dynamic replanning of the patrol observation task is triggered, causing the mobile monitoring device to retreat to a nearby rigid stable standing point (such as a steel plate area) to re-establish a stable monitoring posture, and report the temporary monitoring blind spot to the platform.

[0035] The verification results include job consistency verification results. These results are generated by: extracting the on-site job object identifier and job process characteristics from the on-site sensing data; comparing the on-site job object identifier and job process characteristics with the authorized job information in the permission constraints to generate the job consistency verification results. This avoids situations where there is work on-site but it is not a permitted or authorized job.

[0036] In a specific example, authorized work information includes authorized personnel identification and authorized process type; on-site work object identification includes personnel identity identification; work process characteristics include acoustic features or arc light features. The comparison process specifically includes: comparing personnel identity identification with authorized personnel identification to generate personnel comparison results; when work process characteristics include acoustic features and / or arc light features, comparing the corresponding features with reference features corresponding to the authorized process type to generate process comparison results; and generating work consistency verification results based on personnel comparison results and / or process comparison results. For example, the identity ID is obtained by reading the RFID tag on the worker's safety helmet or through facial recognition and compared with the authorized list on the ticket; simultaneously, the work sound is collected through a microphone array, and the frequency domain energy distribution characteristics are extracted, or the arc light flicker frequency and spectral characteristics are captured by a camera and matched with a pre-stored process fingerprint database such as "MMC welding," "argon arc welding," and "angle grinding." If unauthorized personnel or mismatched process characteristics are detected (e.g., the ticket indicates welding but grinding sounds are heard), the consistency verification is deemed to have failed.

[0037] Step S400: When the verification result meets the preset conditions, a security control command is output to the security control device according to the mapping relationship between the license identifier and the associated device identifier.

[0038] This embodiment employs a hierarchical response system, executing different levels of response actions based on the severity of the anomaly reflected in the verification results. The hierarchical response system comprises three levels: Level 1 is the predictive warning level, Level 2 is the hard-control cutoff level, and Level 3 is the audit and evidence preservation level. Each level of response can be triggered independently or sequentially according to the anomaly escalation conditions.

[0039] Level 1 response is used to handle early warning anomalies, such as situations where the predicted diffusion zone intersects with the restricted area but the on-site concentration has not yet exceeded the limit, or unauthorized personnel are detected entering the monitored area but have not yet started work. Level 1 response actions include local audible and visual deterrence (issuing alarms through speakers and warning lights on mobile monitoring equipment) and platform alarm push notifications (reporting abnormal information to the monitoring center). Level 1 response aims to eliminate risks in their nascent stage and prevent escalation.

[0040] The second-level response handles severe anomalies requiring immediate hazard isolation, such as failed work consistency checks (unauthorized personnel operation or process mismatch), detection of open flames or smoke, or excessive flammable gas concentrations. Specifically, when the check result meets preset conditions, a safety control command carrying a mapping relationship is generated and output to the safety control device. This allows the safety control device to perform isolation or restriction operations on the device corresponding to the associated device identifier based on the mapping relationship, and maintain the locked state of the device corresponding to the associated device identifier until the preset reset conditions are met. The mapping relationship ensures the accuracy of control actions and avoids erroneous isolation due to device identifier confusion. Preferably, the safety control device includes a safety programmable logic controller or a hard-wired safety loop; the preset reset conditions include receiving manual reset confirmation information and the electronic work permit being updated to an effective state. This means that safety isolation is performed at the physical layer or independent safety controller layer, independent of the stability of upper-layer business software, and is unbypassable. The dual reset conditions (manual confirmation + permit status update) prevent automatic resumption of equipment operation if the hazard has not been eliminated or the permit has expired.

[0041] The third-level response is used for tamper-proof auditing and evidence preservation throughout the monitoring process. Regardless of whether a first- or second-level response is triggered, the system writes key monitoring events to the monitoring log. Each log entry records the event timestamp, the location of the moving monitoring device, the permission identifier, the fence version number, the verification result summary, the executed response action, and the hash value of the previous log entry. The system calculates the current hash value for each log entry, with the input including the above fields and the hash value of the previous log entry, thus forming a linked hash chain. Tampering with any log entry will cause inconsistencies in the hash verification of all subsequent entries, making it detectable. The introduction of the fence version number ensures that log entries are bound to a specific version's permission space range. When the permission range changes, the old and new logs can be distinguished by the version number, avoiding version confusion during subsequent tracing.

[0042] In real-world industrial scenarios, multiple valid electronic work permits may exist simultaneously in the same work area. To address this concurrent multi-permit situation, this embodiment also includes a multi-permit arbitration and dynamic replanning mechanism. The system monitors the event bus of the work permit management system to detect changes in permit status in real time. When it detects that an electronic work permit has been revoked or its permitted spatial range has changed, the system re-parses the changed permit constraints and regenerates the corresponding three-dimensional dynamic monitoring space and patrol observation tasks. When the same mobile monitoring device receives multiple valid electronic work permits simultaneously, the system calculates the spatial intersection area between the permitted plane ranges of each permit, prioritizes arbitration based on the work level of each permit, and increases the patrol point density and / or extends the dwell time in the spatial intersection area when generating patrol observation tasks to ensure denser monitoring coverage of high-risk intersection areas.

[0043] The following section uses three typical working scenarios to further illustrate the specific application of the method in this embodiment.

[0044] Operating Condition 1: Gas cutting operation on the pipe section beside the dock wall. The permitted plane area of ​​the electronic work permit is a 5m polygon outside the anchor point, and the permitted height range is from the ground to 4m. The operation process type is level two gas cutting, and the associated equipment is an acetylene valve. Based on the gas cutting process type and the 4m height range, the system generates a parabolic spark splash prediction zone (2-5m outside the permitted plane area, and 1m from the ground). At this time, the on-site micro-meteorological data shows that the wind direction is towards the adjacent tarpaulin pile. The system generates a downwind hazard diffusion prediction zone and detects that this prediction zone has a spatial intersection with the area where the tarpaulin pile is located (preset restricted area). When the mobile monitoring equipment reaches the downwind monitoring point, the combustible gas concentration is only 8% (not exceeding the limit), but the system triggers the first level response because the diffusion prediction zone intersects with the restricted area, and executes an early warning broadcast to drive away the flammable gas. Subsequently, a spark fell into the tarpaulin pile and ignited, causing the flammable gas concentration to surge to 20%. The hazard status verification result met the preset conditions, and the system triggered a second-level response, issuing a cut-off command carrying the mapping relationship between the permission identifier and the acetylene valve identifier to the safety programmable logic controller, and executing a physical trip to cut off the gas source.

[0045] Scenario 2: Interception of Substitute Workers and Unauthorized Process Modifications in Confined Spaces. The electronic work permit specifies the work process as MIG / MAG welding, and the authorized personnel is identified as Zhang San (safety helmet RFID number ZS001). During patrol, the mobile monitoring equipment, through the RFID reader, identified the on-site worker's safety helmet number as Li Si (LS002). The personnel comparison results are inconsistent, triggering the first-level response alarm "Unauthorized Personnel Working." Simultaneously, the acoustic sensor extracts high-frequency angle grinding features instead of welding features, resulting in inconsistent process comparison results. The first-level response alarm is upgraded to "Unauthorized Process Changes," and the platform forcibly suspends the electronic work permit.

[0046] Scenario 3: Self-Safety Reduction and Retreat on the Grid Scaffolding Board. A mobile monitoring device (legged robot) is performing a monitoring task on the deck's grid scaffolding board when a nearby sledgehammer strikes, causing severe deck vibration. The IMU vibration data exceeds the preset stability threshold, and the foot contact force distribution data indicates uneven force distribution. The system determines that the robot is in an unstable stationary state. The system automatically reduces the confidence weight of the current infrared thermal imaging data and combustible gas concentration data to prevent false alarms. Simultaneously, it triggers dynamic replanning of the patrol observation task. The mobile monitoring device retreats to a rigid steel plate area 1.5m away to re-establish a stable monitoring posture and reports the temporary monitoring blind spot to the platform.

[0047] It should be understood that the above-described working scenarios are merely examples, and the method of this invention is equally applicable to other special operation scenarios. For example, in a high-altitude operation monitoring scenario, where the operation type is high-altitude work, the risk expansion area can be determined based on the working height and the radius of the falling object, and the vertical risk expansion area can be determined based on the fall trajectory; in a confined space operation monitoring scenario, the risk expansion area can be determined based on the space ventilation conditions and the diffusion characteristics of harmful gases. Those skilled in the art can adapt the corresponding risk expansion parameters and verification rules according to the specific type of special operation, and these modifications all fall within the protection scope of this invention.

[0048] The present invention also provides a special operation safety monitoring system, which includes a permit parsing module 21, a dynamic space generation module 22, a data acquisition and verification module 23, and a safety response module 24, which are described in detail below: The license parsing module 21 is used to obtain license constraints from the electronic work permit. These constraints include at least the permitted spatial range, work process type, permit validity period window, associated equipment identifier, permit identifier, and the mapping relationship between the permit identifier and the associated equipment identifier. Specifically, the license parsing module communicates with the work permit management system via an API interface or message bus, pulls structured permit data in real time, and parses the two-dimensional polygon coordinate sequence, height range, process code, equipment code, and mapping relationship into a data structure that can be processed internally by the system. When the permitted spatial range changes, the license parsing module synchronously obtains the updated fence version number to ensure that the spatial data used by downstream modules is consistent with the current permit version.

[0049] The dynamic space generation module 22 is used to extend the permitted space range according to the operation process type, generate a three-dimensional dynamic monitoring space, and determine the patrol observation tasks of the mobile monitoring equipment for the three-dimensional dynamic monitoring space. Specifically, the dynamic space generation module determines the horizontal risk extension area and the vertical risk extension area according to the operation process type. The horizontal risk extension area may include the spark splash prediction area generated based on spark splash parameters and / or the downwind hazard diffusion prediction area generated based on real-time micro-meteorological data. The vertical risk extension area is determined by the permitted height range. The permitted plane range, the horizontal risk extension area, and the vertical risk extension area are spatially enclosed to generate the three-dimensional dynamic monitoring space. Based on this, the dynamic space generation module projects the three-dimensional dynamic monitoring space onto the ground layer to determine the projected monitoring area, generates a candidate patrol point sequence along the boundary, and selects the patrol point sequence that meets the preset passage conditions by combining the accessibility cost map of the mobile monitoring equipment. The observation attitude constraints and dwell time are determined according to the risk level of the area where each patrol point is located.

[0050] The data acquisition and verification module 23 controls the mobile monitoring equipment to perform patrol observation tasks to collect on-site perception data. It then verifies the data against permitted constraints and on-site perception data to obtain a verification result that includes at least the operational consistency verification result. Specifically, the module drives the mobile monitoring equipment to proceed sequentially according to the patrol point sequence, observation attitude constraints, and dwell time determined in the patrol observation task, collecting on-site perception data, including infrared thermal imaging data, visible light images, combustible gas concentration, acoustic signals, and micro-meteorological parameters. The module extracts the on-site work object identifier and operational process characteristics from the on-site perception data and compares them with the authorized operation information in the permitted constraints to generate an operational consistency verification result. This comparison process includes comparing personnel identification with authorized personnel identification to generate a personnel comparison result, and comparing acoustic features or arc light features with reference features corresponding to the authorized process type to generate a process comparison result.

[0051] The safety response module 24 is used to output safety control commands to the safety control device according to the mapping relationship between the license identifier and the associated device identifier when the verification result meets the preset conditions. Specifically, the safety response module performs graded responses according to the severity of the anomaly reflected by the verification result: for early warning anomalies, it performs local audible and visual eviction and pushes platform alarms; for severe anomalies that require the hazard source to be cut off, it generates a safety control command carrying the mapping relationship between the license identifier and the associated device identifier and outputs it to the safety control device, so that the safety control device performs cut-off or restriction operations on the device corresponding to the associated device identifier according to the mapping relationship, and maintains the locked state until it receives manual reset confirmation information and the electronic work permit status is updated to the effective state; at the same time, the safety response module writes key monitoring events into the monitoring log and calculates the hash value of each log entry to form a hash chain linking the previous and subsequent entries, ensuring the non-repudiation and traceability of the monitoring process.

[0052] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the special operation safety monitoring method as described in any of the above method embodiments.

[0053] It should be understood that the sensor types, communication protocols, robot configurations, and safety control device forms in the above embodiments can all be replaced according to actual needs. For example, personnel identification can also use UWB positioning tags or iris recognition; process feature comparison can also introduce thermal imaging texture analysis; and the safety control device can also be a relay array or FPGA security module with a security certification level. These variations are all within the protection scope of this invention.

[0054] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for safety monitoring of special operations, characterized in that, include: Obtain the licensing constraints from the electronic work permit. The licensing constraints include at least the licensed spatial range, work process type, license validity period window, associated equipment identifier, license identifier, and the mapping relationship between the license identifier and the associated equipment identifier. Based on the type of work process, the permitted space range is extended for risk control to generate a three-dimensional dynamic monitoring space, and the patrol and observation tasks of the mobile monitoring equipment for the three-dimensional dynamic monitoring space are determined. The mobile monitoring device is controlled to perform the patrol observation task to collect on-site perception data, and the data is verified in combination with the permission constraints to obtain a verification result that includes at least the operation consistency verification result. When the verification result meets the preset conditions, a security control command is output to the security control device according to the mapping relationship between the license identifier and the associated device identifier.

2. The special operation safety monitoring method according to claim 1, characterized in that, The special operations include hot work, and the types of operations include gas cutting, electric welding, or grinding. The step of extending the permitted space range according to the work process type to generate a three-dimensional dynamic monitoring space includes: The horizontal risk extension area is determined according to the operation process type, wherein the plane dimension in the permitted space range is the permitted plane range; The vertical risk expansion area is determined based on the height dimension within the permitted space range, where the height dimension is the permitted height range. The permitted plane range, the horizontal risk expansion area, and the vertical risk expansion area are spatially enclosed to generate the three-dimensional dynamic monitoring space.

3. The special operation safety monitoring method according to claim 1, characterized in that, The verification result includes a job consistency verification result, and the method for generating the job consistency verification result includes: Extract the on-site operation object identifier and operation process characteristics from the on-site perception data; The on-site work object identifier and the work process characteristics are compared with the authorized work information in the permission constraints to generate the work consistency verification result.

4. The special operation safety monitoring method according to claim 1, characterized in that, When the verification result meets the preset conditions, a security control command is output to the security control device according to the mapping relationship between the license identifier and the associated device identifier, including: When the verification result meets the preset conditions, a security control command carrying the mapping relationship is generated and output to the security control device, so that the security control device performs a cut-off operation or a restriction operation on the device corresponding to the associated device identifier according to the mapping relationship, and maintains the locked state of the device corresponding to the associated device identifier until the preset reset condition is met.

5. The special operation safety monitoring method according to claim 2, characterized in that, The determination of the horizontal risk extension area based on the operation process type includes at least one of the following: Based on the spark splash parameters corresponding to the operation process type and the permissible height range, the spark splash prediction area outside the permissible plane range is determined as the horizontal risk extension area; Real-time micro-meteorological data is acquired, and the downwind hazard diffusion prediction area is determined as the horizontal risk expansion area by combining the real-time micro-meteorological data with the combustible material diffusion parameters corresponding to the operation process type.

6. The special operation safety monitoring method according to claim 3, characterized in that, The authorized operation information includes the authorized personnel identifier and the authorized process type; the on-site operation object identifier includes the personnel identity identifier; the operation process characteristics include acoustic characteristics or arc light characteristics; The step of comparing the on-site work object identifier and the work process characteristics with the authorized work information in the permission constraints to generate the work consistency verification result includes: The personnel identification identifier is compared with the authorized personnel identifier to generate a personnel comparison result; When the operation process features include acoustic features and / or arc light features, the corresponding features are compared with the reference features corresponding to the authorized process type to generate process comparison results; Based on the personnel comparison results and / or the process comparison results, the operation consistency verification results are generated.

7. The special operation safety monitoring method according to claim 4, characterized in that, The safety control device includes a secure programmable logic controller or a hardwired safety loop. The preset reset conditions include receiving manual reset confirmation information and the electronic work permit being updated to an effective status.

8. The special operation safety monitoring method according to claim 2, characterized in that, The determination of the patrol and observation tasks of the mobile monitoring device for the three-dimensional dynamic monitoring space includes: The three-dimensional dynamic monitoring space is projected onto the ground layer to determine the projected monitoring area, and a sequence of candidate patrol points is generated based on the boundary of the projected monitoring area. Obtain the accessibility cost map of the mobile monitoring device, and use the accessibility cost map to filter out the patrol point sequence that meets the preset access conditions from the candidate patrol point sequence; By combining the horizontal risk expansion area and the vertical risk expansion area, the observation attitude constraints and dwell time of each patrol point in the patrol point sequence are determined.

9. A special operation safety monitoring system, characterized in that, include: The license parsing module is used to obtain license constraints from the electronic work permit. The license constraints include at least the license space range, work process type, license validity time window, associated equipment identifier, license identifier, and the mapping relationship between the license identifier and the associated equipment identifier. The dynamic space generation module is used to extend the permitted space range according to the operation process type, generate a three-dimensional dynamic monitoring space, and determine the patrol and observation tasks of the mobile monitoring equipment for the three-dimensional dynamic monitoring space. The data acquisition and verification module is used to control the mobile monitoring device to perform the patrol observation task to acquire field perception data, and to perform verification in combination with the permission constraints and the field perception data to obtain a verification result that includes at least the operation consistency verification result. The security response module is used to output a security control command to the security control device according to the mapping relationship between the license identifier and the associated device identifier when the verification result meets the preset conditions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the special operation safety monitoring method as described in any one of claims 1 to 8.

Images