Fire operation environment monitoring method and monitoring system based on safety helmet

Abstract

The invention discloses a fire operation environment monitoring method and monitoring system based on safety helmets, and relates to the field of fire operation safety monitoring, and the technical scheme is that the method comprises the steps: constructing a cooperative sensing network composed of a plurality of safety helmet terminals; multi-source sensing data such as gas, temperature, dust, oxygen content and light / heat radiation and spatial position information are synchronously collected; fusing and generating local risk parameters under a unified coordinate system, constructing a three-dimensional environment risk field covering the operation area, and analyzing a risk spatio-temporal evolution law; therefore, the safety state is comprehensively evaluated, the diffusion trend of a high-risk area is predicted, and grading early warning, equipment linkage and sensing resource self-adaptive adjustment strategies are dynamically generated. The method has the advantages that pre-combustion or abnormal energy release can be recognized before open fire occurs, the early warning advance and recognition accuracy are remarkably improved, the high-risk area space coverage rate is increased, the false alarm rate is reduced, and the personnel evacuation completion rate is increased.

Description

Technical Field

[0001] This invention relates to the field of hot work safety monitoring, and in particular to a hot work environment monitoring method and system based on a safety helmet. Background Technology

[0002] Hot work is widely practiced in fields such as petrochemicals, metallurgy, power generation, shipbuilding, and municipal engineering, encompassing various operations involving open flames, high temperatures, and molten slag splashes, including welding, cutting, grinding, and baking. In environments containing flammable gases, liquid vapors, or dust, hot work can easily become an ignition source. If it forms an explosive mixture with leaked flammable materials, it can easily lead to major safety accidents such as fires, explosions, and poisoning / asphyxiation. Therefore, real-time monitoring and risk warning of the environmental conditions at hot work sites has always been an important research direction in the field of safety engineering.

[0003] In existing technologies, monitoring of hot work environments mainly relies on equipment such as fixed combustible gas detectors, area infrared flame detectors, and portable gas detectors. Fixed combustible gas detectors are typically installed at specific locations in plant areas, pipe racks, or tank areas, triggering alarms by measuring the concentration of combustible gas in a localized space. However, these devices are point-based monitoring systems, which are greatly limited by their installation location and number. For hot work scenarios where the leak source is unclear or the leak pattern is variable, there are often monitoring blind spots, and they are difficult to reflect the spatial diffusion and distribution of risks. While area infrared flame detectors can respond quickly to open flames, their ability to identify pre-ignition, smoldering, or abnormal energy release events without obvious flames is limited. They can only issue alarms after the fire has clearly developed, resulting in insufficient early warning timeliness.

[0004] Portable gas detectors are currently one of the most commonly used environmental monitoring tools for field workers. They are typically worn on the chest or waist to detect the concentration of flammable or toxic gases at a specific point in the surrounding environment. However, these devices are significantly affected by changes in sampling location and posture, providing only localized numerical information and failing to provide a comprehensive risk assessment of the entire work area. Furthermore, most portable detectors rely on instantaneous concentration exceeding limits for alarms, making it difficult to combine multiple parameters such as temperature, oxygen content, and dust levels to comprehensively determine "pre-combustion" or "energy accumulation" states. This can easily lead to delayed warnings or failure to identify risk evolution trends. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method and system for monitoring hot work environments based on safety helmets.

[0006] The technical solution includes the following steps: Construct a collaborative sensing network consisting of multiple safety helmet terminals worn by workers to synchronously collect multi-source sensor data and corresponding spatial location information of the hot work area; The system monitors the dynamic changes in the multi-source sensor data, analyzes and identifies potential hazardous features related to hot work risks, and generates local risk parameters associated with the corresponding spatial location information. By aggregating the local risk parameters of each safety helmet terminal under a unified work space coordinate system, a three-dimensional environmental risk field covering the hot work area is established, and the spatiotemporal evolution law of environmental risk is analyzed. Based on the aforementioned three-dimensional environmental risk field and its spatiotemporal evolution, a comprehensive assessment of the safety status of the hot work environment is conducted, and the diffusion trend of high-risk areas is predicted. Based on the assessment results and predicted trends, a graded early warning strategy for personnel and equipment and an adaptive adjustment scheme for the perception resources of each safety helmet terminal are generated, and dynamic early warning information and control instructions are output.

[0007] Preferably, the helmet terminal includes: A multimodal environmental sensing unit is used to collect data on gas concentration, temperature, dust concentration, oxygen content, as well as light radiation and thermal radiation. A spatial positioning unit is used to acquire the three-dimensional position and / or attitude information of the safety helmet terminal in the work space; The edge computing unit is configured to preprocess and extract features from the collected multi-source sensor data, and generate local risk parameters based on a preset model. The wireless communication unit is used to upload the local risk parameters and multi-source sensor data to the collaborative sensing network; A power supply unit is used to supply power to the various components of the helmet terminal.

[0008] Preferably, the process involves monitoring the dynamic changes in the multi-source sensor data, analyzing and identifying potential hazards related to hot work risks, and generating local risk parameters associated with corresponding spatial location information. This specifically includes the following steps: Time series analysis was performed on multi-source sensor data to obtain the rate of change, gradient of change, and fluctuation characteristics of gas concentration, temperature, dust concentration, oxygen content, light radiation data, and thermal radiation data. Based on the changing characteristics of optical and thermal radiation data, identify signs of abnormal energy release and / or pre-ignition; By fusing the abnormal energy release signs and / or pre-combustion signs with environmental parameters such as gas concentration, temperature, dust concentration, and oxygen content at the corresponding time, local risk parameters characterizing the environmental risk level of the corresponding spatial location are obtained.

[0009] Preferably, identifying signs of abnormal energy release and / or pre-ignition specifically includes the following steps: When the optical radiation data shows a transient strong light pulse and the thermal radiation data temperature rise rate exceeds a threshold, it is identified as an abnormal energy release event. When an increase in combustible gas concentration, a decrease in oxygen content, and abnormal thermal radiation data are detected, but there are no continuous open flame characteristics, it is identified as a pre-ignition event.

[0010] Preferably, the local risk parameters of each safety helmet terminal are aggregated under a unified work space coordinate system to establish a three-dimensional environmental risk field covering the hot work area, and the spatiotemporal evolution law of environmental risk is analyzed. This specifically includes the following steps: Map the local risk parameters and their corresponding spatial location information uploaded by each safety helmet terminal to three-dimensional mesh nodes under a unified work space coordinate system; Based on the risk parameters of the three-dimensional mesh nodes, a continuous three-dimensional environmental risk field is generated using a spatial interpolation algorithm; The three-dimensional environmental risk field is time-series aligned and differentially calculated for multiple consecutive time windows to obtain the dynamic change vector of the risk field. The dynamic change vector includes the risk intensity growth rate, the migration direction of high-risk areas, and the risk diffusion speed. Based on the dynamic change vector analysis, the evolution of environmental risks in spatial and temporal dimensions can be analyzed to identify potential risk clusters or rapidly deteriorating areas.

[0011] Preferably, based on the three-dimensional environmental risk field and its spatiotemporal evolution, a comprehensive assessment of the safety status of the hot work environment is conducted, and the diffusion trend of high-risk areas is predicted. This specifically includes the following steps: The risk parameters of each grid node in the current three-dimensional environmental risk field and their corresponding dynamic change vectors are weighted and fused to generate a risk index that represents the comprehensive risk level. The risk index is compared with a multi-level preset risk threshold, and each grid node is divided into a safe zone, an early warning zone, or a high-risk zone. Based on the migration direction and diffusion speed of grid nodes in high-risk areas, a spatiotemporal extrapolation algorithm is used to predict the spatial distribution range of high-risk areas within a future preset time window. By combining the relative relationship between the predicted high-risk areas and the current location of workers and the distribution of key equipment, it is determined whether there is an unacceptable risk in the hot work environment within the stated time window, and the determination result serves as the core basis for the comprehensive safety status assessment and the generation of subsequent early warning strategies.

[0012] Preferably, based on the assessment results and predicted trends, a tiered early warning strategy for personnel and equipment, as well as an adaptive adjustment scheme for the sensing resources of each safety helmet terminal, are generated, and dynamic early warning information and control instructions are output. This specifically includes the following steps: Based on the risk level classification results of each grid node and the predicted spatial distribution range of high-risk areas within a preset time window, determine the personnel early warning level and equipment control level corresponding to different risk levels. Based on the personnel warning level, corresponding warning strategies are generated for workers who are in the warning zone, high-risk zone, or are predicted to enter the high-risk zone, including at least one of the following: triggering the sound and light alarm or vibration reminder of the safety helmet terminal; Based on the equipment control level, instructions to restrict start-up or reduce power operation are generated for equipment performing hot work, and emergency shutdown or isolation instructions are generated for nearby critical non-hot work equipment. Based on the predicted range of high-risk areas, risk change trends, and the remaining battery power and communication load status of each safety helmet terminal, the sensing resource configuration of the safety helmet terminal is dynamically adjusted, including at least one of the following: increasing or decreasing the sampling frequency of multi-source sensor data, adjusting the data upload cycle, or enabling enhanced monitoring mode of specific types of sensors. The generated graded early warning information and equipment control instructions are sent to the corresponding safety helmet terminal and / or hot work equipment control system through the wireless communication module, and execution status feedback is received to confirm the implementation of the early warning prompts and control actions.

[0013] A hot work environment monitoring system based on a safety helmet, characterized in that: Collaborative sensing network module: Composed of multiple terminals worn by workers on safety helmets, used to synchronously collect data on gas concentration, temperature, dust concentration, oxygen content, light radiation, heat radiation, and corresponding spatial location information in the hot work area; Local risk generation module: Configured in the edge computing unit of each safety helmet terminal, it is used to perform time series analysis on multi-source sensor data, identify abnormal energy release events or pre-ignition events, and integrate environmental parameters to generate local risk parameters associated with spatial location; 3D Risk Field Modeling Module: Used to aggregate local risk parameters uploaded by various safety helmet terminals under a unified work space coordinate system, construct a 3D environmental risk field covering the hot work area through spatial interpolation algorithm, and calculate the dynamic change vector of the risk field to analyze the spatiotemporal evolution law; Safety Status Assessment and Prediction Module: Used to perform weighted fusion of three-dimensional environmental risk field and dynamic change vector, divide into safe zone, early warning zone and high-risk zone, predict the spread range of high-risk area within future time window, and determine whether there is unacceptable risk by combining personnel location and distribution of key equipment. The graded early warning and resource regulation module is used to generate sound, light and vibration early warning strategies for personnel and control instructions for equipment based on the assessment and prediction results. It dynamically adjusts the sampling frequency, upload cycle and sensor working mode of each safety helmet terminal, and sends early warning information, evacuation guidance and sensing resource configuration instructions through the wireless communication network, while receiving execution status feedback.

[0014] The beneficial effects of the technical solution provided by the embodiments of the present invention are as follows: By integrating multimodal environmental perception and spatial positioning into the safety helmet terminal, joint monitoring of combustible gas, temperature, dust, oxygen content, and light / thermal radiation can be achieved, enabling the identification of pre-combustion and abnormal energy release states before the appearance of open flames, significantly advancing the warning time; by constructing a three-dimensional environmental risk field under a unified coordinate system, the spatial coverage and modeling accuracy of high-risk areas are significantly improved, providing the command end with an intuitive risk distribution and evolution trend; combined with risk assessment and prediction results, graded early warning, equipment linkage, and adaptive scheduling of perception resources are implemented, improving the personnel evacuation completion rate and reducing the false alarm rate, and overall realizing the technical upgrade of the hot work environment from "point-based passive detection" to "spatialized, predictive active protection". Attached Figure Description

[0015] Figure 1 The first embodiment of the present invention provides an overall flowchart of a hot work environment monitoring method based on a safety helmet. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. Of course, the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0017] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0018] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on this invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0019] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0020] Example 1 See Figure 1 This invention provides a method for monitoring hot work environments based on safety helmets, comprising: S1. Construct a collaborative sensing network consisting of multiple safety helmet terminals worn by workers to synchronously collect multi-source sensor data and corresponding spatial location information of the hot work area. The safety helmet terminal includes: A multimodal environmental sensing unit is used to collect data on gas concentration, temperature, dust concentration, oxygen content, as well as light radiation and thermal radiation. A spatial positioning unit is used to acquire the three-dimensional position and / or attitude information of the safety helmet terminal in the work space; The edge computing unit is configured to preprocess and extract features from the collected multi-source sensor data, and generate local risk parameters based on a preset model. The wireless communication unit is used to upload the local risk parameters and multi-source sensor data to the collaborative sensing network; A power supply unit is used to supply power to the various components of the helmet terminal.

[0021] Furthermore, the cooperative sensing network is constructed and operated in the following manner: Before entering the hot work area, each safety helmet terminal automatically connects to the same local area network or dedicated industrial Internet of Things (IIoT) network via wireless communication unit, forming a self-organized collaborative sensing network. The master control node or cloud coordinator in the network issues a unified timestamp synchronization command, enabling each terminal to synchronously collect environmental data and location information with the same sampling period (e.g., 1 Hz base frequency, which can be dynamically increased to 10 Hz in high-risk scenarios), ensuring that multi-source data are aligned in the time dimension.

[0022] The multimodal environment sensing unit includes: An electrochemical gas sensor array is used for real-time detection of combustible gases (such as methane and propane), toxic gases (such as CO and H2S), and oxygen content. Digital temperature and humidity sensors and infrared thermopile sensors are used to measure ambient temperature and thermal radiation intensity, respectively. Photodiodes or broadband photosensitive sensors are used to capture light radiation signals in the visible to near-infrared bands and support the detection of instantaneous strong light pulses; Laser scattering dust sensor is used to monitor the concentration of combustible dust in the air; The spatial positioning unit adopts a fusion positioning scheme, including a UWB (Ultra-Wideband) indoor positioning module and an IMU (Inertial Measurement Unit), to achieve centimeter-level three-dimensional position calculation in an industrial plant without GPS signal, and combines the SLAM algorithm to compensate for position drift caused by short-term communication interruption; Each unit is connected to the edge computing unit via an internal bus (such as I²C or SPI). The raw sensing data is filtered for noise, outliers are removed and timestamped by the edge computing unit, and then bound to the location information to form a structured data packet. This data packet is then uploaded by the wireless communication unit to the edge server or remote monitoring platform of the collaborative sensing network via Wi-Fi 6 or LoRaWAN protocol.

[0023] S2. Monitor the dynamic change characteristics of the multi-source sensor data, analyze and identify potential hazard characteristics related to hot work risks, and generate local risk parameters associated with the corresponding spatial location information; During system operation, each safety helmet terminal periodically uploads multi-source sensor data, including physical quantities such as combustible gas concentration, ambient temperature, dust concentration, oxygen content, light radiation intensity, and heat radiation intensity.

[0024] To analyze the changing trend of the above data over time, we first perform time series processing on it. Let the environmental parameter vector collected at time t be...

[0025] in For combustible gas concentration, For ambient temperature, For dust concentration, For oxygen content, For light radiation intensity, This refers to thermal radiation intensity or equivalent surface temperature.

[0026] Using the above data within a time window Based on the sampling sequence, the dynamic change rate, first-order gradient change intensity, and short-term fluctuation characteristics of each parameter are calculated.

[0027] For environmental parameter components (in Its rate of change can be expressed as

[0028] Its local gradient change can be expressed as

[0029] Its short-term fluctuation characteristics can be expressed as follows:

[0030] Based on the dynamic analysis of light and thermal radiation, signs of abnormal energy release or pre-ignition can be further identified. When the light radiation signal exhibits a significant pulse peak within a very short time, and the pulse duration is less than 100 ms, while the heating rate of thermal radiation meets the following criteria...

[0031] And the light radiation intensity is higher than the preset threshold. When this combination of features is used, it can be identified as an abnormal energy release event, which can be used to characterize high-energy transient events such as electric arcs and metal sparks.

[0032] In the absence of persistent open flame characteristics, if an increase in combustible gas concentration is detected and the following conditions are met...

[0033] At the same time, the oxygen content decreases and meets the requirements.

[0034] And the thermal radiation level was observed to be higher than the reference thermal radiation intensity. And the offset exceeds the threshold

[0035] Furthermore, the duration of the aforementioned anomaly exceeded the preset time. When this occurs, it can be identified as a pre-ignition event. This event is typically used to characterize the state in which combustible gases or dust form latent combustion conditions in a localized area.

[0036] To achieve a quantitative representation of the current environmental risk status, the event identification results are fused with other environmental parameters to generate local risk parameters. .

[0037] Risk parameters can be constructed using a weighted fusion model:

[0038] in This indicates that an abnormal energy release event or pre-ignition event has been detected; otherwise, it is 0. and These represent the event weights and environmental factor weights, respectively. Risk contribution coefficients for different environmental factors; Characterizes oxygen deficit.

[0039] Finally, the obtained local risk parameters are bound to the spatial location information at the corresponding time to form a structured risk parameter tuple.

[0040] in This indicates the position of the safety helmet terminal in a unified workspace coordinate system. The tuple is uploaded to the collaborative sensing network via a wireless communication module for subsequent steps involving the construction of a three-dimensional environmental risk field and analysis of risk evolution patterns.

[0041] S3. Under a unified work space coordinate system, aggregate the local risk parameters of each safety helmet terminal, establish a three-dimensional environmental risk field covering the hot work area, and analyze the spatiotemporal evolution of environmental risks. To achieve a spatial description of the risks of hot work environments, the terminals of each safety helmet are positioned at all times. Uploaded local risk parameters and their corresponding spatial locations Mapped to a unified workspace coordinate system Inside.

[0042] The mapped data is represented as

[0043] in Indicates the first One valid sampling location, To construct a continuous three-dimensional environmental risk field, the discrete risk data is embedded into a preset three-dimensional grid structure, corresponding to the local risk parameters.

[0044] Let the coordinates of the grid nodes be... The goal is to obtain the risk value of each node. Based on the spatial distance relationship between each node and the sampling point, a spatial interpolation algorithm can be used to estimate the risk value.

[0045] Interpolation methods may include, but are not limited to, inverse distance weighted interpolation, kriging interpolation, or radial basis function interpolation. Taking inverse distance weighted interpolation as an example, its interpolation expression is:

[0046] in This represents the distance attenuation coefficient (typically 1.5~3). If radial basis function (RBF) interpolation is used, a risk field of the following form can be constructed:

[0047] in Basis functions, such as Gaussian functions

[0048] The risk distribution obtained through the above interpolation That is, the hot work area is defined at any time. A three-dimensional environmental risk field was constructed. To characterize the dynamic evolution of risk over time, multiple consecutive time windows were selected to form a risk field sequence.

[0049] Perform time-series alignment and differencing on the risk fields of adjacent time periods to obtain the risk rate of change field:

[0050] Based on this, the risk gradient and diffusion direction can be further calculated. The risk gradient vector is represented as:

[0051] This vector can be used to characterize the migration direction of high-risk areas; the spatial diffusion rate of risk can be expressed as...

[0052] By analyzing the rate of change, migration direction, and diffusion speed of risks, we can identify areas where potential risks accumulate, areas that are continuously increasing, or areas that are rapidly deteriorating, thus providing a basis for subsequent risk prediction, early warning strategy formulation, and resource allocation.

[0053] S4. Based on the three-dimensional environmental risk field and its spatiotemporal evolution, comprehensively assess the safety status of the hot work environment and predict the diffusion trend of high-risk areas. After constructing the three-dimensional environmental risk field, it is necessary to further analyze its spatial distribution characteristics and temporal changes to form a quantitative assessment of the safety status of the current hot work area. To this end, the risk parameters of each grid node in the risk field and their corresponding dynamic change vectors are first calculated and integrated. Let the grid node at time t be... Its risk value is The dynamically changing vector is

[0054] in For the rate of change of risk, For the risk gradient vector, This refers to the speed at which the risk spreads.

[0055] To integrate the aforementioned multiple risk factors into quantifiable risk assessment indicators, a weighted fusion model is introduced to construct a comprehensive risk index. Its form can be expressed as

[0056] α1, α2, and α3 are weighting parameters configured according to the types of hazards in different hot work environments. The comprehensive risk index can simultaneously characterize the current risk level, risk growth trend, and risk diffusion capability, and is an important basis for assessing safety status.

[0057] To facilitate regional division, the comprehensive risk index is compared with a preset multi-level risk threshold system. Let the safety threshold be θ1 and the warning threshold be θ2, where θ1 < θ2. When...

[0058] Then the grid node is designated as a safe zone; when

[0059] Then it is determined to be a warning zone; when

[0060] This area is then classified as a high-risk zone. After identifying the high-risk zone, further analysis is conducted based on the node risk gradient direction and diffusion rate to determine future time windows. The spatial distribution range of high-risk areas within the region is predicted. The prediction model can employ a spatiotemporal diffusion method based on local linear extrapolation, which can be expressed as follows:

[0061] This formula indicates that the center of a high-risk area will migrate along the risk gradient, and the migration distance is determined by both the diffusion rate and the prediction time window.

[0062] After predicting high-risk areas for the future, the system further considers the current location distribution of workers, the location of key equipment, and the layout of combustible materials to determine whether the high-risk areas will cover key areas or be close to personnel work sites in the future. If the minimum safe distance between the predicted area and personnel or equipment is less than a preset threshold, the hot work environment is determined to have an unacceptable risk within that time window. This determination will serve as the basis for generating subsequent graded early warning strategies and triggering resource scheduling mechanisms.

[0063] S5. Based on the evaluation results and predicted trends, generate a graded early warning strategy for personnel and equipment and an adaptive adjustment scheme for the perception resources of each safety helmet terminal, and output dynamic early warning information and control instructions.

[0064] After completing a comprehensive risk assessment of the hot work environment and predicting the spread of high-risk areas, corresponding early warning measures and equipment control strategies need to be formulated based on the risk levels of different areas. First, based on the risk index of each grid node and its comparison with preset thresholds, the hot work area is divided into safe zones, early warning zones, and high-risk zones. Combining this with the predicted spatial distribution of future high-risk areas, different early warning levels and equipment control levels can be configured for different risk levels. For example, basic monitoring is implemented in safe zones, enhanced monitoring and mild warnings are triggered in early warning zones, and immediate intervention measures are implemented in high-risk zones.

[0065] The early warning strategy for personnel is determined based on their location and the risk areas they may enter in the future. When a worker is located in a warning zone, a high-risk zone, or when their location is predicted to intersect with a high-risk zone in the future, the system automatically generates a corresponding early warning strategy. The early warning strategy may include triggering audible and visual alarms or vibration alerts on the safety helmet terminal, or displaying risk direction prompts on the terminal display interface, so that workers can take timely retreat and avoidance measures.

[0066] The control strategy for hot work equipment and adjacent critical equipment is determined by both the risk level and the forecast results. When equipment performing hot work is located in a high-risk area or is expected to be covered by a high-risk area in the future, the system issues control instructions to the equipment control system to restrict start-up and shutdown, reduce power operation, or conduct emergency shutdown. For adjacent critical equipment that is not hot work, when it is predicted that it may be affected by risks, instructions to isolate, cut off power, or switch to backup links can be issued to reduce the risk of accident coupling.

[0067] While formulating early warning and control strategies, it is also necessary to adaptively adjust the sensing resources of the safety helmet terminals. Based on the risk level of the area where each terminal is located, the predicted risk trend, its remaining battery power, and its current communication load status, the sensor sampling frequency, data upload cycle, or edge computing mode should be dynamically adjusted. For example, terminals in safe zones can reduce sampling frequency and upload rate to reduce energy consumption; terminals in early warning zones can increase the sampling rate of key sensors (such as thermal radiation and combustible gas sensors); terminals located in high-risk areas or along risk diffusion paths can enable full-band high-precision data acquisition and execute local event recognition algorithms to improve risk detection sensitivity.

[0068] Ultimately, the generated tiered early warning information, evacuation guidance information, and equipment control instructions are transmitted in real time to the corresponding safety helmet terminals and / or hot work equipment control systems via the collaborative sensing network. Simultaneously, the system receives execution status feedback from the terminals and equipment, confirming the implementation of early warning and control measures, thus forming a closed-loop safety control process of "risk identification—strategy generation—execution feedback."

[0069] Example 2, an embodiment of the present invention, provides a method for monitoring hot work environments based on safety helmets. To verify the effectiveness of the proposed method, a set of simulation tests was designed for actual working conditions. To verify the effectiveness of the method in a real industrial scenario, a typical hot work area in a maintenance workshop of a petrochemical enterprise was selected for a field comparative test.

[0070] The test area was approximately 60 m long, 8 m wide, and 10 m high, with multiple flanges, valves, and supports arranged longitudinally. Four locations were selected to house controllable trace combustible gas leak sources and dummy arc welding loads to simulate hazardous conditions such as combustible gas accumulation, short-duration arc discharge, and high-temperature molten slag splashing during the pre-combustion stage. During the test, six workers performed typical hot work operations such as welding, cutting, and grinding within the pipe gallery. Two workers wore the helmet-based 3D risk monitoring terminal provided by this invention, two wore comparative helmets with only temperature sensors and basic audible and visual alarms, and the remaining two were equipped only with traditional portable gas detectors and walkie-talkies.

[0071] The ceiling and side walls of the utility tunnel are equipped with several fixed combustible gas detectors and infrared flame detectors in accordance with current specifications, and the original manual inspection record method is retained as a reference for existing technology.

[0072] Before the test, all types of terminals were calibrated and their clocks were synchronized. The basic sampling period of the collaborative sensing network was set to 1 Hz, and automatically increased to 10 Hz when the risk index of any node exceeded the preset warning threshold.

[0073] The helmet-based monitoring terminal integrates a multimodal environmental sensing unit, capable of simultaneously collecting data on combustible gas concentration, temperature, dust concentration, oxygen content, and light and heat radiation. It then calculates local risk parameters locally in real time via an edge computing unit and transmits the data back to the edge server via 4G / Wi-Fi. Comparative devices are installed according to the manufacturer's recommended configuration without any modifications to their internal algorithms to ensure the objectivity of the comparison.

[0074] The experiment consisted of 20 scenarios, covering typical hazardous modes such as normal ventilation and open flame, limited local ventilation, trace gas leakage on the upwind side, gas accumulation on the downwind side, dust and welding slag dispersion, and smoldering in cable trenches. In each scenario, the flow rate and duration of the leak source were controlled by a mass flow meter, the ambient temperature rise and smoke changes were recorded by an infrared thermometer and a smoke meter, and personnel behavior and the development process of the fire source were recorded simultaneously by a high-definition camera.

[0075] During the experiment, the time difference between the initial warning signal issued by various systems and the "starting time of the dangerous state" determined by on-site experts based on video recordings was recorded. Key data such as the accuracy of pre-ignition event identification, the occurrence of false alarms or missed alarms, the coverage area of ​​different systems in the dangerous zone, and the completion status of personnel evacuation were also recorded. To evaluate communication and computing power overhead, the average data upload traffic of the helmet-based monitoring system in multiple scenarios was statistically analyzed to verify the network resource consumption of this invention while ensuring safety performance.

[0076] Test Data Table

[0077] The experimental data above demonstrates that the helmet-based hot work environment monitoring system significantly outperforms existing technologies in several key safety indicators. Firstly, regarding the core indicator of "average lead time for high-risk warnings," the system of this invention achieves 18.7 seconds, while fixed gas alarms and portable gas detectors only achieve 4.3 seconds and 6.1 seconds respectively. Single-temperature monitoring helmets and area infrared flame detectors have even lower lead times, at 3.8 seconds and 5.0 seconds respectively. Manual inspection methods are essentially ineffective in providing an effective lead time, with a statistical result of 0 seconds. This indicates that the present invention, through the fusion of light and heat radiation sensing combined with multimodal analysis of local gas, dust, and oxygen content changes, can provide stable warnings during the pre-ignition stage. Compared to traditional equipment that relies on images of excessive gas concentrations or open flames, this significantly improves the foresight of the warnings, allowing more sufficient response time for personnel evacuation and process interlocking.

[0078] From the perspective of the "accuracy rate of pre-ignition event identification" indicator, the average identification rate of the system of this invention is 92.4%, which is significantly higher than the 61.3% to 68.7% range of the comparative schemes, especially the single temperature monitoring safety helmet of Comparison 3, which only achieved 54.9%. This shows that it is difficult to distinguish between "normal high-temperature operation" and "abnormal energy accumulation" by relying on a single physical quantity. However, this invention, by introducing an edge computing model at the terminal side of the safety helmet to comprehensively interpret the time series characteristics, considers both the temperature rise rate and the coordinated changes in combustible gas and oxygen content, and can more sensitively capture potential combustion chain initiation signs, demonstrating the creative advantages of multi-source fusion algorithms.

[0079] Regarding the false alarm rate, the statistical results of the system of this invention are 3.1%, significantly lower than the level of 9.6% to 15.2% for the comparative equipment. Fixed gas detectors and area infrared flame detectors are easily affected by instantaneous smoke, sparks, or local gas disturbances during normal hot work activities such as welding and arcing, with false alarm rates reaching 11.8% and 10.7%, respectively. This invention, by comprehensively analyzing the evolution trajectory of the risk field in spatial and temporal dimensions, can effectively filter short-term interference and localized sporadic anomalies, reduce the number of false triggers, and help improve the operator's trust in alarm information, avoiding "alarm fatigue".

[0080] In terms of spatial risk description capabilities, the "high-risk area spatial coverage" index reflects the significant advantage of this invention in using safety helmets as mobile sensing nodes for 3D environmental modeling. In multiple tests, the system of this invention achieved an average coverage rate of 93.5%, while fixed gas alarms and portable gas detectors achieved 42.7% and 55.9%, respectively. Single-temperature monitoring safety helmets and area infrared flame detectors achieved 38.4% and 47.1%, respectively, and manual inspection methods only achieved 29.6%. Traditional systems mostly provide point-based monitoring or localized sector-shaped monitoring, making it difficult to comprehensively cover blind spots, upper platforms, and lower trenches in complex spaces. This invention, however, achieves high-density sampling of risk areas by superimposing the natural movement trajectories of personnel in the work area with a multi-helmet collaborative mechanism, significantly improving spatial coverage and blind spot identification capabilities.

[0081] The mean square error of the 3D risk field reconstruction reflects the accuracy of spatial modeling. The mean square error of the system in this invention is 6.8%, significantly better than the 18.3%–27.4% levels of the comparative schemes. Comparative techniques typically only provide local numerical values ​​or two-dimensional planar trends, making it difficult to reconstruct the actual 3D risk distribution. This invention, by interpolating and temporally smoothing the local risk parameters of each node under a unified coordinate system, can construct a 3D risk field that more closely resembles the actual hazard diffusion state, providing reliable basic data for subsequent risk prediction and evacuation route planning. This feature demonstrates clear novelty and inventiveness.

[0082] The "personnel evacuation completion rate" shows that the system of this invention significantly outperforms existing technologies in protecting personnel safety. The experimental result for this invention is 98.2%, while comparisons 1-4 range from 83.5% to 89.1%, with the manual inspection scheme having the lowest rate at only 78.9%. This difference stems from two aspects: firstly, the invention provides more sufficient early warning time; secondly, the tiered early warning strategy based on a three-dimensional risk field generates differentiated audio-visual and vibration prompts and evacuation guidance information for different risk areas, enabling workers to choose evacuation paths that deviate from the direction of risk field diffusion in the shortest possible time, thereby effectively reducing the probability of being trapped or accidentally entering high-risk areas.

[0083] It should be noted that, regarding the "average data upload traffic per terminal," the system of this invention achieves a value of 146.3 kB / min, significantly higher than the 38.5–61.2 kB / min of some comparative devices. This indicates that while achieving high-precision 3D modeling and real-time early warning, this invention places higher demands on communication and computing resources. However, experiments show that this data volume is perfectly acceptable under the existing industrial wireless network conditions of enterprises, and through dynamic sampling frequency adjustment and hierarchical upload mechanisms, bandwidth consumption can be significantly reduced in low-risk phases. Therefore, from the perspective of overall system performance gains, the moderate increase in data volume is reasonable and engineeringly feasible in exchange for the substantial security improvement of quantifying the 3D risk field and predictive early warning capabilities.

[0084] In summary, the experimental data fully demonstrate that, compared with traditional hot work safety technologies that rely on single-point or single-parameter monitoring, this invention, through its overall technical approach of "multimodal perception + three-dimensional risk field modeling + spatiotemporal evolution prediction + hierarchical early warning and resource scheduling," achieves more forward-looking, precise, and spatialized safety risk management, exhibiting significant beneficial effects, outstanding substantive features, and remarkable progress.

[0085] Example 3 provides a hot work environment monitoring system based on a safety helmet, characterized in that: Collaborative sensing network module: Composed of multiple terminals worn by workers on safety helmets, used to synchronously collect data on gas concentration, temperature, dust concentration, oxygen content, light radiation, heat radiation, and corresponding spatial location information in the hot work area; Local risk generation module: Configured in the edge computing unit of each safety helmet terminal, it is used to perform time series analysis on multi-source sensor data, identify abnormal energy release events or pre-ignition events, and integrate environmental parameters to generate local risk parameters associated with spatial location; 3D Risk Field Modeling Module: Used to aggregate local risk parameters uploaded by various safety helmet terminals under a unified work space coordinate system, construct a 3D environmental risk field covering the hot work area through spatial interpolation algorithm, and calculate the dynamic change vector of the risk field to analyze the spatiotemporal evolution law; Safety Status Assessment and Prediction Module: Used to perform weighted fusion of three-dimensional environmental risk field and dynamic change vector, divide into safe zone, early warning zone and high-risk zone, predict the spread range of high-risk area within future time window, and determine whether there is unacceptable risk by combining personnel location and distribution of key equipment. The graded early warning and resource regulation module is used to generate sound, light and vibration early warning strategies for personnel and control instructions for equipment based on the assessment and prediction results. It dynamically adjusts the sampling frequency, upload cycle and sensor working mode of each safety helmet terminal, and sends early warning information, evacuation guidance and sensing resource configuration instructions through the wireless communication network, while receiving execution status feedback.

[0086] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A hard hat-based hot work environment monitoring method, characterized by, The method comprises the following steps: A cooperative perception network composed of safety helmet terminals worn by workers is constructed, and multi-source sensing data and corresponding spatial position information of a hot work area are synchronously collected; Dynamic change characteristics of the multi-source sensing data are monitored, potential dangerous characteristics related to hot work risk are analyzed and identified, and local risk parameters associated with corresponding spatial position information are generated; The local risk parameters of the safety helmet terminals are converged under a unified work space coordinate system, a three-dimensional environmental risk field covering the hot work area is established, and the spatio-temporal evolution law of environmental risk is analyzed; Based on the three-dimensional environmental risk field and the spatio-temporal evolution law, the safety state of the hot work environment is comprehensively evaluated, and the diffusion trend of the high-risk area is predicted; According to the evaluation results and the prediction trend, a hierarchical early warning strategy for personnel and equipment and an adaptive adjustment scheme of the perception resources of each safety helmet terminal are generated, and dynamic early warning information and control instructions are output.

2. The hard hat-based hot work environment monitoring method of claim 1, wherein, The safety helmet terminal comprises: A multi-modal environmental perception unit for collecting gas concentration, temperature, dust concentration, oxygen content, light radiation data and thermal radiation data; A spatial positioning unit for obtaining three-dimensional position and / or attitude information of the safety helmet terminal in the work space; An edge computing unit configured to preprocess and extract features of the collected multi-source sensing data, and generate local risk parameters based on a preset model; A wireless communication unit for uploading the local risk parameters and multi-source sensing data to the cooperative perception network; A power supply unit for supplying power to each component of the safety helmet terminal.

3. The hard hat-based hot work environment monitoring method of claim 2, wherein, The dynamic change characteristics of the multi-source sensing data are monitored, potential dangerous characteristics related to hot work risk are analyzed and identified, and local risk parameters associated with corresponding spatial position information are generated, which specifically comprises the following steps: Time series analysis is performed on the multi-source sensing data to obtain the change rate, change gradient and fluctuation characteristics of the gas concentration, temperature, dust concentration, oxygen content, light radiation data and thermal radiation data; Based on the change characteristics of the light radiation data and the thermal radiation data, abnormal energy release signs and / or pre-combustion signs are identified; The abnormal energy release signs and / or pre-combustion signs are fused with environmental parameters such as gas concentration, temperature, dust concentration and oxygen content at the corresponding time to obtain local risk parameters representing the environmental risk level of the corresponding spatial position.

4. The hard hat-based hot work environment monitoring method of claim 3, wherein, Identifying abnormal energy release signs and / or pre-combustion signs specifically comprises the following steps: When the light radiation data appears a transient strong light pulse and the thermal radiation data has a temperature rise rate exceeding a threshold value, an abnormal energy release event is identified; When combustible gas concentration rises, oxygen content decreases, and thermal radiation data is abnormal but has no continuous flame characteristics, a pre-combustion event is identified.

5. The hard hat-based hot work environment monitoring method of claim 3, wherein, The local risk parameters of the safety helmet terminals are converged under a unified work space coordinate system, a three-dimensional environmental risk field covering the hot work area is established, and the spatio-temporal evolution law of environmental risk is analyzed, which specifically comprises the following steps: The local risk parameters uploaded by each safety helmet terminal and the corresponding spatial position information are mapped to three-dimensional grid nodes in the unified work space coordinate system; generating a continuous three-dimensional environmental risk field by using a spatial interpolation algorithm based on the risk parameters of the three-dimensional grid nodes; performing time-series alignment and difference calculation on the three-dimensional environmental risk fields of multiple continuous time windows to obtain a dynamic change vector of the risk field, the dynamic change vector including a risk intensity growth rate, a high-risk area migration direction, and a risk diffusion speed; analyzing the evolution law of the environmental risk in the spatial and temporal dimensions based on the dynamic change vector to identify potential risk aggregation areas or rapidly deteriorating areas.

6. The hard hat-based live fire operation environment monitoring method of claim 5, wherein, comprehensively evaluating the safety state of the hot work operation environment and predicting the diffusion trend of the high-risk area based on the three-dimensional environmental risk field and the spatiotemporal evolution law, specifically including the following steps: generating a risk index representing the comprehensive risk level by weighting and fusing the risk parameters of each grid node in the three-dimensional environmental risk field at the current time and the corresponding dynamic change vector; comparing the risk index with multiple levels of preset risk thresholds to divide each grid node into a safe zone, a warning zone, or a high-risk zone; predicting the spatial distribution range of the high-risk area in a future preset time window based on the migration direction and diffusion speed of the grid nodes in the high-risk zone by using a spatiotemporal extrapolation algorithm; determining whether there is an unacceptable risk in the hot work operation environment within the time window in combination with the relative relationship between the predicted high-risk area and the current location of the operating personnel and the distribution of key equipment, and taking the determination result as the core basis for the comprehensive evaluation of the safety state and the generation of subsequent warning strategies.

7. The hard hat-based live fire operation environment monitoring method of claim 6, wherein, generating a hierarchical warning strategy for personnel and equipment and an adaptive adjustment scheme for the sensing resources of each safety helmet terminal based on the evaluation result and the prediction trend, and outputting dynamic warning information and control instructions, specifically including the following steps: determining the personnel warning level and the equipment control level corresponding to different risk levels according to the risk level division result of each grid node and the spatial distribution range of the predicted high-risk area within the preset time window; generating a corresponding warning strategy for operating personnel in the warning zone, the high-risk zone, or predicted to enter the high-risk area based on the personnel warning level, including at least one of triggering the audible and visual alarm of the safety helmet terminal and the vibration reminder; generating an instruction to limit the start and stop or reduce the power operation of the equipment performing the hot work operation and an emergency shutdown or isolation instruction for the adjacent key non-hot work equipment based on the equipment control level; dynamically adjusting the sensing resource configuration of the safety helmet terminal according to the range of the predicted high-risk area, the risk change trend, and the remaining power and communication load state of each safety helmet terminal, including at least one of increasing or decreasing the sampling frequency of multi-source sensing data, adjusting the data upload period, or enabling an enhanced monitoring mode of a specific type of sensor; downloading the generated hierarchical warning information and equipment control instructions to the corresponding safety helmet terminal and / or hot work operation equipment control system through the wireless communication module, and receiving execution state feedback to confirm the implementation of the warning prompt and control action.

8. A safety helmet-based hot work operation environment monitoring system using the method of any one of claims 1-7, characterized by: The synergic perception network module is composed of safety helmet terminals worn by workers, and is used for synchronously collecting gas concentration, temperature, dust concentration, oxygen content, light radiation data, heat radiation data and corresponding spatial position information in a hot work area. The local risk generation module is configured in an edge computing unit of each safety helmet terminal, is used for time series analysis on multi-source sensing data, identifies abnormal energy release events or pre-combustion events, and generates local risk parameters associated with spatial positions by fusing environmental parameters. The three-dimensional risk field modeling module is used for gathering local risk parameters uploaded by each safety helmet terminal in a unified work space coordinate system, constructing a three-dimensional environmental risk field covering the hot work area by a spatial interpolation algorithm, and calculating a dynamic change vector of the risk field to analyze the spatio-temporal evolution law. The safety state evaluation and prediction module is used for weighted fusion on the three-dimensional environmental risk field and the dynamic change vector, divides a safety zone, a warning zone and a high-risk zone, predicts the diffusion range of the high-risk zone in a future time window, and determines whether there is an unacceptable risk by combining a worker position and a key equipment distribution. The hierarchical warning and resource regulation module is used for generating a sound-light-vibration warning strategy for workers and a management and control instruction for equipment according to the evaluation and prediction results, dynamically adjusting a sampling frequency, an uploading period and a sensor working mode of each safety helmet terminal, and issuing warning information, evacuation guidance and sensing resource configuration instructions through a wireless communication network, and simultaneously receiving execution state feedback.

Images