Digital twin operation and maintenance method for combustible gas detection internet of things device

Abstract

This invention relates to the field of IoT device operation and maintenance technology, specifically to a digital twin operation and maintenance method for combustible gas detection IoT devices, comprising the following steps: establishing a global sensing network for combustible gas detection IoT devices, and collecting multi-dimensional real-time parameter sets of each detection device in real time; constructing a digital twin benchmark model of all device elements to form a virtual benchmark body for real-time benchmarking simulation; performing frame-by-frame time-series benchmarking between the real-time parameter sequence and the simulation benchmark parameter sequence, dynamically calculating the real-time residuals of multi-dimensional parameters, and constructing a dynamic residual sequence for the device; performing feature identification and anomaly tracing on the dynamic residual sequence, distinguishing scenarios, generating differentiated and standardized operation and maintenance work orders, and distributing them to the operation and maintenance terminal; and identifying hidden degradation of the device through time-series residual quantification, thereby achieving quantitative, precise, and closed-loop intelligent operation and maintenance of combustible gas detection IoT devices, thus meeting the needs of high-precision, refined, and predictive full life-cycle operation and maintenance of the device.

Description

Technical Field

[0001] This invention relates to the field of IoT device operation and maintenance technology, and in particular to a digital twin operation and maintenance method for combustible gas detection IoT devices. Background Technology

[0002] Combustible gas detection IoT devices are core sensing equipment for safe production in high-risk scenarios such as chemical industrial parks, gas stations, oil and gas storage facilities, and mining enterprises. They primarily rely on IoT sensing and wireless transmission technologies to achieve real-time monitoring and early warning of combustible gas concentrations, effectively preventing major safety accidents such as explosions and poisoning caused by gas leaks. They are an indispensable basic sensing terminal in industrial safety risk prevention and control systems. With the rapid popularization of industrial IoT and intelligent operation and maintenance technologies, combustible gas detection equipment has now basically achieved full-area networking, automatic data collection, and remote online monitoring. The large number of deployed devices, wide distribution, and long operating cycles place higher demands on equipment operational stability, detection accuracy and reliability, timely fault identification, and precise operation and maintenance.

[0003] Existing combustible gas detection equipment operation and maintenance monitoring technologies mostly adopt qualitative analysis methods based on fixed threshold comparisons and manual experience judgments, generally lacking dynamic residual quantitative analysis mechanisms. This has become a core technical bottleneck restricting the improvement of the accuracy of intelligent equipment operation and maintenance. Traditional technologies simply compare the deviation between real-time monitoring data and fixed factory thresholds without constructing a standard operating benchmark time series for the equipment. This makes it impossible to achieve frame-by-frame time-series benchmarking between measured data and ideal standard data, making it difficult to generate continuous and dynamic parameter residual sequences. Consequently, it is impossible to accurately capture the implicit and gradual performance degradation characteristics of the equipment from a quantitative perspective. Summary of the Invention

[0004] The purpose of this invention is to provide a digital twin operation and maintenance method for IoT devices for combustible gas detection. It aims to solve the technical problems in the prior art, which simply compare the deviation between real-time monitoring data and fixed factory thresholds, without constructing a standard operating benchmark time sequence for the device, making it impossible to achieve frame-by-frame time-series benchmarking between measured data and ideal standard data, making it difficult to generate continuous and dynamic parameter residual sequences, and failing to accurately capture the implicit and gradual performance degradation characteristics of the device from a quantitative perspective.

[0005] To achieve the above objectives, the present invention employs a digital twin operation and maintenance method for combustible gas detection IoT devices, comprising the following steps: A comprehensive sensing network of IoT devices for combustible gas detection is established to collect multi-dimensional real-time parameter sets from each detection device in real time. These multi-dimensional real-time parameter sets include three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters. Based on the inherent attributes, installation scenarios, and standard operating parameters of combustible gas detection equipment, a digital twin benchmark model of all equipment elements is constructed to replicate the standard operating state and response logic of the physical equipment, forming a virtual benchmark body for real-time benchmarking and simulation. The digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. Input the multi-dimensional real-time parameter set into the digital twin benchmark model to obtain the corresponding time-series simulation benchmark parameter sequence. Then, perform frame-by-frame time-series comparison between the real-time parameter sequence and the simulation benchmark parameter sequence, dynamically calculate the real-time residuals of the multi-dimensional parameters, construct the dynamic residual sequence of the equipment, and quantify the degree of equipment anomaly and deterioration status through the fluctuation amplitude, offset trend, and continuous characteristics of the residual sequence. The system performs feature identification and anomaly tracing on dynamic residual sequences, distinguishing between four scenarios: sensor drift, hardware failure, environmental interference, and actual leakage. Based on the anomaly type, residual offset level, and equipment location information, it generates differentiated and standardized operation and maintenance work orders and sends them to the operation and maintenance terminal.

[0006] Among the steps involved in building a comprehensive sensing network for combustible gas detection IoT devices and collecting multi-dimensional real-time parameter sets from each detection device, including three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters: Deploy wireless sensing nodes with edge computing capabilities, and network various combustible gas detectors, temperature and humidity sensors, barometers and equipment power monitoring devices through a low-power wide area network to form a full-area perception topology covering the target area. Set the acquisition cycle, and synchronously read three types of time-series data in each acquisition cycle: equipment operating parameters, gas detection parameters, and on-site environmental parameters to obtain raw time-series data. Among them, the equipment operating parameters include operating voltage, operating current, signal strength, and continuous operating time; the gas detection parameters include instantaneous value of combustible gas concentration, concentration change rate, and original electrical signal amplitude of the sensor; and the on-site environmental parameters include ambient temperature, ambient humidity, atmospheric pressure, and dust concentration. The original time-series data is timestamped and missing frames due to network latency and packet loss are removed. A sliding window filtering method is used to smooth and denoise the original signal, retaining the key change features of the signal and outputting a multi-dimensional real-time parameter set.

[0007] The process involves aligning the original time-series data with timestamps, removing missing frames due to network latency and packet loss, smoothing and denoising the original signal using a sliding window filtering method, preserving key signal variation features, and outputting a multi-dimensional real-time parameter set. The multi-dimensional real-time parameter set is lightweight compressed at the edge and uploaded in a unified data encapsulation format.

[0008] The process involves constructing a digital twin benchmark model of all elements of the combustible gas detection equipment based on its inherent properties, installation scenarios, and standard operating parameters. This model replicates the standard operating state and response logic of the physical equipment, forming a virtual benchmark for real-time simulation. The digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. Establish physical performance benchmarks and construct the physical response characteristics of the equipment under the initial healthy state by combining aging and degradation experience curves; among which, physical performance benchmarks include sensor sensitivity nominal value, response time constant, zero-point output benchmark value and range; Define the operating behavior benchmark, which is the ideal response curve of the equipment to a standard concentration of combustible gas under standard environmental conditions; the operating behavior benchmark includes the dynamic response behavior of the entire process from gas contact sensor to signal output, the power consumption behavior and communication behavior of the equipment in different operating modes; Establish scenario adaptation benchmarks. Based on the spatial volume, ventilation conditions, distance to potential leakage sources, and distribution of obstructions at the actual installation locations of the equipment, construct environmental correction factors for different application scenarios to obtain a digital twin benchmark model.

[0009] Before establishing physical performance benchmarks and constructing the physical response characteristics of the equipment under initial healthy conditions using aging degradation experience curves, whereby the physical performance benchmarks include the nominal value of sensor sensitivity, response time constant, zero-point output benchmark value, and measurement range: Obtain geometric reference parameters from the equipment manufacturer's technical manual to construct the mapping relationship between the three-dimensional spatial layout and internal structure of the physical equipment; among which, the geometric reference parameters include the internal air chamber volume of the detector, the sensor probe installation position, and the structural features of the explosion-proof shell.

[0010] Among them, the steps of inputting a multi-dimensional real-time parameter set into a digital twin benchmark model to obtain the corresponding time-series simulation benchmark parameter sequence, performing frame-by-frame time-series comparison between the real-time parameter sequence and the simulation benchmark parameter sequence, dynamically calculating the real-time residuals of multi-dimensional parameters, constructing a dynamic residual sequence of the equipment, and quantifying the degree of equipment anomaly and degradation state through the fluctuation amplitude, offset trend, and continuous characteristics of the residual sequence are as follows: The multi-dimensional real-time parameter set obtained from the IoT platform at the same timestamp is input into the digital twin benchmark model. Based on the currently input real-time environmental parameters and equipment operating parameters, the simulation benchmark value of the equipment under the environmental conditions is calculated and output, forming a simulation benchmark parameter sequence corresponding to the real-time data time sequence. The actual gas concentration value and the measured parameters of the sensor signal amplitude collected in real time are subtracted from the corresponding items in the simulation benchmark parameter sequence frame by frame to obtain the initial deviation value of each dimension at each sampling time. The deviation values ​​are then arranged in chronological order to construct the dynamic residual sequence of the equipment.

[0011] The process involves subtracting the real-time collected actual gas concentration values ​​and sensor signal amplitude measured parameters from the corresponding items in the simulation baseline parameter sequence frame by frame to obtain the initial deviation values ​​for each dimension at each sampling time, and then arranging the deviation values ​​in chronological order to construct the equipment dynamic residual sequence. Sliding window statistical analysis is performed on the dynamic residual sequence to calculate the average fluctuation amplitude of the residual within the window, the slope of the monotonically increasing / decreasing trend of the residual sequence, and the length of time the residual continues to exceed the preset noise band. The stability of the device output is quantified based on the fluctuation amplitude, the speed of sensor sensitivity decay / zero drift is quantified based on the offset trend, and the duration and stability of abnormal device states are quantified based on the continuous characteristics.

[0012] The process includes the steps of quantifying the stability of the device output based on the fluctuation amplitude, quantifying the rate of sensor sensitivity decay / zero drift based on the offset trend, and quantifying the duration and stability of the device's abnormal state based on the continuous characteristics: The three quantitative indicators of residual sequence fluctuation amplitude, offset trend and persistence characteristics are integrated for evaluation to output the abnormality level of the device, and the degradation state is divided into three levels: slight drift, moderate aging and severe failure according to the slope of the offset trend.

[0013] Among them, the steps of performing feature identification and anomaly tracing on dynamic residual sequences, distinguishing between four scenarios—sensor drift, hardware failure, environmental interference, and actual leakage—and generating differentiated and standardized operation and maintenance work orders based on anomaly type, residual offset level, and equipment location information, and then issuing them to the operation and maintenance terminal are as follows: Feature extraction is performed on the dynamic residual sequence, and the abnormal state is classified into four scenarios based on the combination of features: sensor drift, hardware failure, environmental interference, and real leakage. The features include the sign of the residual, the rate of change of the residual, the correlation between the residual and environmental parameters, and the consistency between the residual and the rate of change of gas concentration. Based on the identified anomaly type and residual offset level, the priority and processing time requirements of maintenance work orders are determined. Combining equipment location information, equipment number, and geographical location tag, the work order content is packaged into a standard format and then sent to the mobile terminals of maintenance personnel in the corresponding area through the Internet of Things platform. The execution status and feedback results of the work orders are tracked to form a closed-loop maintenance management system.

[0014] In the process of extracting features from the dynamic residual sequence and classifying abnormal states into four categories based on feature combinations: sensor drift, hardware failure, environmental interference, and actual leakage, the following steps are involved: Features include the sign of the residual, the rate of change of the residual, the correlation between the residual and environmental parameters, and the consistency between the residual and the rate of change of gas concentration. For sensor drift scenarios, when the residual changes monotonically and slowly and is unrelated to the actual gas concentration fluctuation, it is identified as zero drift and sensitivity decay, and calibration and sensor replacement suggestions are generated, along with an estimated remaining lifespan. For hardware failure scenarios, when the residual shows a large jump, a continuous constant output, or irregular fluctuations, it is identified as a circuit fault or sensor damage, and an emergency shutdown repair work order is generated. For scenarios involving environmental interference, when residual fluctuations are highly correlated with sudden changes in ambient temperature and humidity and increases in dust concentration, and when there is no abnormal upward trend in gas concentration, false alarms caused by environmental factors are identified, and maintenance suggestions for cleaning equipment protective covers and adjusting environmental compensation parameters are generated. For real-world leak scenarios, when the residual shows a rapid positive increase that is consistent with the rising trend of the actual gas concentration, and the simulation baseline output does not change significantly, it is identified as a real leak event, generating the highest priority alarm and triggering the opening and closing command of the exhaust valve.

[0015] This invention discloses a digital twin operation and maintenance method for combustible gas detection IoT devices. It establishes a comprehensive sensing network for the combustible gas detection IoT devices and collects multi-dimensional real-time parameter sets from each detection device in real time. These multi-dimensional real-time parameter sets include three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters. Based on the inherent attributes, installation scenarios, and standard operating parameters of the combustible gas detection devices, a digital twin benchmark model of all device elements is constructed to replicate the standard operating state and response logic of the physical devices, forming a virtual benchmark for real-time benchmarking simulation. The digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. The multi-dimensional real-time parameter sets are input into the digital twin benchmark model to obtain the corresponding time-series simulation. The benchmark parameter sequence is used to compare the real-time parameter sequence with the simulation benchmark parameter sequence frame by frame, dynamically calculate the real-time residuals of multi-dimensional parameters, and construct the dynamic residual sequence of the equipment. The fluctuation amplitude, offset trend and continuous characteristics of the residual sequence are used to quantify the degree of equipment anomaly and deterioration status. Feature identification and anomaly tracing are performed on the dynamic residual sequence to distinguish four scenarios: sensor drift, hardware failure, environmental interference and actual leakage. Based on the anomaly type, residual offset level and equipment location information, differentiated and standardized operation and maintenance work orders are generated and sent to the operation and maintenance terminal. By quantifying the time-series residuals to identify the hidden deterioration of the equipment, quantitative, precise and closed-loop intelligent operation and maintenance of combustible gas detection IoT equipment can be realized, thereby meeting the high-precision, refined and predictive operation and maintenance needs of the entire equipment life cycle. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the steps of the digital twin operation and maintenance method for combustible gas detection IoT devices of the present invention.

[0018] Figure 2 This is a flowchart of steps S100 of the present invention.

[0019] Figure 3 This is a flowchart of steps S200 of the present invention.

[0020] Figure 4 This is a flowchart of steps S300 of the present invention.

[0021] Figure 5 This is a flowchart of steps S400 of the present invention.

[0022] Figure 6 This is a schematic diagram of the digital twin operation and maintenance system for combustible gas detection IoT devices according to the present invention.

[0023] Figure 7 This is a schematic diagram of the electronic device of the present invention. Detailed Implementation

[0024] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0025] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0026] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0027] Please see Figures 1-5 This invention provides a digital twin operation and maintenance method for combustible gas detection IoT devices, comprising the following steps: S100: Establish a comprehensive sensing network for combustible gas detection IoT devices, and collect multi-dimensional real-time parameter sets from each detection device in real time; these multi-dimensional real-time parameter sets include three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters; the specific process is as follows: S101: Deploy wireless sensing nodes with edge computing capabilities to network various combustible gas detectors, temperature and humidity sensors, barometers and equipment power monitoring devices through a low-power wide area network to form a full-area perception topology covering the target area. S102: Set the acquisition cycle, and synchronously read three types of time-series data in each acquisition cycle: equipment operating parameters, gas detection parameters, and on-site environmental parameters to obtain raw time-series data; among them, equipment operating parameters include operating voltage, operating current, signal strength, and continuous operating time; gas detection parameters include instantaneous value of combustible gas concentration, concentration change rate, and original electrical signal amplitude of sensor; on-site environmental parameters include ambient temperature, ambient humidity, atmospheric pressure, and dust concentration. S103: Timestamp alignment is performed on the original time series data, missing frames due to network latency and packet loss are removed, and the original signal is smoothed and denoised using a sliding window filtering method, while retaining the key change features of the signal and outputting a multi-dimensional real-time parameter set. S104: Perform edge-side lightweight compression on the multi-dimensional real-time parameter set and upload the multi-dimensional real-time parameter set in a unified data encapsulation format.

[0028] In the above process, wireless sensing nodes with edge computing capabilities are deployed, and various combustible gas detectors, temperature and humidity sensors, barometers and equipment power monitoring devices are networked through a low-power wide area network to form a full-area perception topology covering the target area, realizing the interconnection and interoperability of multiple types of sensing devices and full-area data collection coverage.

[0029] A fixed data acquisition cycle is preset, and three types of time-series raw data are read synchronously in each acquisition cycle. Among them, the acquired equipment operating parameters include operating voltage, operating current, signal strength, and continuous operating time, which are used to characterize the power supply and communication operation status of the equipment; the acquired gas detection parameters include instantaneous values ​​of combustible gas concentration, concentration change rate, and sensor raw electrical signal amplitude, which are used to reflect the gas detection response status; the acquired field environmental parameters include ambient temperature, ambient humidity, atmospheric pressure, and dust concentration, which are used to record the environmental conditions at the monitoring site, and finally obtain a multi-dimensional raw time-series dataset.

[0030] Standardized preprocessing was performed on the collected raw time-series data. First, all data frames were aligned with a unified timestamp to eliminate timing misalignment issues caused by multiple devices and parameters. Second, missing data frames and invalid abnormal frames caused by network latency and data packet loss were accurately identified and removed. Finally, a sliding window filtering algorithm was used to smooth and denoise the raw signal data, effectively filtering out high-frequency interference noise while fully preserving key signal change characteristics such as sudden changes in gas concentration and fluctuations in equipment parameters, ultimately outputting a regular and reliable multi-dimensional real-time parameter set.

[0031] At the edge, the pre-processed multi-dimensional real-time parameter set is lightweighted and compressed to reduce data transmission bandwidth consumption and transmission latency; the encapsulation format and transmission protocol of all parameter data are unified, and the standardized multi-dimensional real-time parameter set is stably uploaded to the IoT platform to provide compliant input data for subsequent model simulation and residual calculation.

[0032] S200: Based on the inherent attributes, installation scenarios, and standard operating parameters of combustible gas detection equipment, a digital twin benchmark model of all equipment elements is constructed to replicate the standard operating state and response logic of the physical equipment, forming a virtual benchmark body for real-time benchmarking simulation. The digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. The specific process is as follows: S201: Obtain geometric reference parameters from the equipment manufacturer's technical manual and construct the mapping relationship between the three-dimensional spatial layout and internal structure of the physical equipment; among which, the geometric reference parameters include the internal air chamber volume of the detector, the sensor probe installation position, and the structural features of the explosion-proof shell; S202: Establish physical performance benchmarks and construct the physical response characteristics of the equipment under the initial healthy state by combining the aging and decay experience curves; wherein, the physical performance benchmarks include the nominal value of sensor sensitivity, response time constant, zero-point output benchmark value and range; S203: Define the operating behavior benchmark, which is the ideal response curve of the equipment to a standard concentration of combustible gas under standard environmental conditions; the operating behavior benchmark includes the dynamic response behavior of the entire process from gas contact sensor to signal output, the power consumption behavior and communication behavior of the equipment in different operating modes; S204: Establish scenario adaptation benchmarks. Based on the spatial volume, ventilation conditions, distance to potential leakage sources, and distribution of obstructions at the actual installation locations of the equipment, construct environmental correction factors for different application scenarios to obtain a digital twin benchmark model.

[0033] In the above process, the model geometric reference generation process is as follows: retrieve the technical manual of the combustible gas detector, extract the precise geometric reference parameters of the equipment, including the internal gas chamber volume of the detector, the sensor probe installation position, the explosion-proof shell structural features and other core parameters; based on 3D modeling technology, accurately construct a 3D spatial model of the external structure and internal component layout of the physical equipment, establish the spatial structural mapping relationship between the physical equipment and the virtual model, and complete the geometric reference construction.

[0034] Model physical performance benchmark generation process: Based on the nominal performance parameters of the equipment at the factory and combined with the long-term aging and decay experience curve of the sensor, a physical response characteristic benchmark of the equipment in the initial healthy state is constructed. This physical performance benchmark fully includes the nominal value of sensor sensitivity, response time constant, zero-point output benchmark value and standard range, and establishes the hardware performance standard of the equipment in the state of no aging and no fault, which serves as the core basis for judging the performance degradation of the equipment.

[0035] Model operation behavior benchmark generation process: Define the standard operation behavior logic of the equipment and construct a standardized operation behavior benchmark. This benchmark is the ideal dynamic response curve of the equipment to different concentrations of standard combustible gas under standard ambient temperature and pressure and interference-free environmental conditions. Completely replicate the dynamic response behavior of the entire process from gas contact sensor, sensor signal conversion to electrical signal output, while covering the power consumption change behavior and data communication transmission behavior under different working modes such as standby, detection, and alarm, to realize the virtual replication of the equipment operation logic.

[0036] Model scenario adaptation benchmark generation and complete model fusion process: For each actual installation point of the equipment, collect scenario feature information such as the spatial volume of the point, ventilation conditions, distance of potential leakage sources, and distribution of obstructions around the equipment; Based on the scenario features, construct environmental correction factors corresponding to different application scenarios to compensate for the impact of environmental conditions on the detection accuracy of the equipment; Deeply integrate the geometric benchmark, physical performance benchmark, operational behavior benchmark and scenario adaptation correction factors to finally obtain a full-element digital twin benchmark model adapted to the real working conditions on site, which can output standard simulation parameters in real time.

[0037] S300: Input the multi-dimensional real-time parameter set into the digital twin benchmark model to obtain the corresponding time-series simulation benchmark parameter sequence. Perform frame-by-frame time-series comparison between the real-time parameter sequence and the simulation benchmark parameter sequence, dynamically calculate the real-time residuals of the multi-dimensional parameters, construct the equipment dynamic residual sequence, and quantify the degree of equipment anomaly and degradation state through the fluctuation amplitude, offset trend, and continuous characteristics of the residual sequence. The specific process is as follows: S301: Input the multi-dimensional real-time parameter set obtained from the IoT platform at the same timestamp into the digital twin benchmark model. Based on the currently input real-time environmental parameters and equipment operating parameters, calculate and output the simulation benchmark value of the equipment under the environmental conditions, forming a simulation benchmark parameter sequence corresponding to the real-time data time sequence. S302: Subtract the real-time collected actual gas concentration value and the measured parameters of sensor signal amplitude from the corresponding items in the simulation reference parameter sequence frame by frame to obtain the initial deviation value of each dimension at each sampling time, and arrange the deviation values ​​in chronological order to construct the equipment dynamic residual sequence; S303: Perform sliding window statistical analysis on the dynamic residual sequence, calculate the average fluctuation amplitude of the residual within the window, the slope of the monotonically increasing / decreasing trend of the residual sequence, and the length of time the residual continues to exceed the preset noise band; S304: Quantify the stability of the device output based on the fluctuation amplitude, quantify the speed of sensor sensitivity decay / zero drift based on the offset trend, and quantify the duration and stability of the device abnormal state based on the continuous characteristics. S305: It integrates and evaluates three quantitative indicators of residual sequence fluctuation amplitude, offset trend and persistence characteristics, outputs the abnormality level of the device, and classifies the deterioration state into three levels: slight drift, moderate aging and severe failure according to the slope of the offset trend.

[0038] In the above process, the simulation benchmark parameter sequence generation process is as follows: a standardized multi-dimensional real-time parameter set under a unified timestamp is retrieved from the Internet of Things platform, and real-time environmental parameters and equipment operating parameters are synchronously input into the trained digital twin benchmark model; the model combines the current real working conditions, adaptively calculates and outputs the theoretical standard detection values ​​and operating values ​​of the equipment under the working conditions, generates simulation benchmark data corresponding one-to-one with the real-time collected data in a time sequence, and continuously splices them to form a complete time-series simulation benchmark parameter sequence.

[0039] The dynamic residual sequence construction process is as follows: using the timestamp as the sole benchmark, the measured data such as the instantaneous value of gas concentration, the amplitude of sensor electrical signals, and the operating parameters of the equipment collected in real time on site are compared with the simulation benchmark parameters of the same sequence frame by frame to calculate the initial deviation value of each dimension of the parameters at each sampling time. According to the time axis sequence, the multidimensional deviation values ​​of all time nodes are continuously connected in series to construct the dynamic residual sequence of the equipment in a structured manner.

[0040] Residual sequence feature data processing and statistical process: A sliding window statistical method is used to perform refined analysis on the dynamic residual sequence. The sliding window size is fixed and the residual sequence is traversed frame by frame. The average fluctuation amplitude of the residual data in each window is calculated in real time to characterize the overall fluctuation stability of the parameters. The residual change curve within the window is fitted to solve the trend slope of the monotonically increasing or monotonically decreasing residual sequence to characterize the continuous shift rate of the parameters. The cumulative time length of the residual value that continuously exceeds the preset normal noise fluctuation range is statistically analyzed to characterize the continuous characteristics of the abnormal state.

[0041] Equipment status quantification assessment process: Equipment status quantification is completed based on residual statistical characteristics. By using the average fluctuation amplitude of the residual window, the stability of the real-time output data of the equipment is accurately quantified to determine whether the equipment has frequent jitter anomalies. By using the slope and direction of the overall residual offset trend, the sensor sensitivity decay rate and the positive or negative zero drift rate are quantified. By using the duration of abnormal residuals, the persistence and stability of the abnormal state of the equipment are quantified, and the transient interference and persistent faults are distinguished.

[0042] The equipment degradation level fusion judgment process involves weighted fusion evaluation of three types of quantitative indicators: residual fluctuation amplitude, offset trend slope, and abnormal duration, to comprehensively calculate the overall abnormality level of the equipment. At the same time, based on the magnitude of the residual offset trend slope, the equipment performance degradation state is further subdivided into three degradation levels: slight parameter drift, moderate equipment aging, and severe equipment failure, providing a precise classification basis for subsequent differentiated operation and maintenance.

[0043] S400: Performs feature identification and anomaly tracing on dynamic residual sequences, distinguishing between four scenarios: sensor drift, hardware failure, environmental interference, and actual leakage. Based on the anomaly type, residual offset level, and equipment location information, it generates differentiated and standardized maintenance work orders and distributes them to the maintenance terminal. The specific process is as follows: S401: Features are extracted from the dynamic residual sequence, and abnormal states are classified into four scenarios based on feature combinations: sensor drift, hardware failure, environmental interference, and actual leakage. Features include the sign of the residual, the rate of change of the residual, the correlation between the residual and environmental parameters, and the consistency between the residual and the rate of change of gas concentration. For sensor drift scenarios, when the residual changes monotonically and slowly and is unrelated to actual gas concentration fluctuations, it is identified as zero-point drift and sensitivity decay, generating calibration and sensor replacement suggestions, along with an estimated remaining lifespan. For hardware failure scenarios, when the residual shows a large jump or a continuous constant... When there are constant output fluctuations or irregular fluctuations, it is identified as a circuit fault or sensor damage, and an emergency shutdown maintenance work order is generated. For environmental interference scenarios, when the residual fluctuations are highly correlated with sudden changes in ambient temperature and humidity and an increase in dust concentration, and the gas concentration does not show an abnormal upward trend, it is identified as a false alarm caused by environmental factors, and maintenance suggestions for cleaning the equipment protective cover and adjusting environmental compensation parameters are generated. For real leakage scenarios, when the residual shows a rapid positive increase and is consistent with the upward trend of the actual gas concentration, while the simulation benchmark output does not change significantly, it is identified as a real leakage event, and the highest priority alarm and linkage exhaust valve opening and closing commands are generated. S402: Based on the identified anomaly type and residual offset level, determine the priority and processing time requirements of the maintenance work order. Combine the equipment location information, equipment number and geographical location tag, encapsulate the work order content into a standard format, and send it to the mobile terminal of the corresponding maintenance personnel through the Internet of Things platform. Track the execution status and feedback results of the work order to form a closed-loop maintenance management system.

[0044] In the above process, the residual feature extraction and multi-scenario anomaly tracing and classification process involves: extracting multi-dimensional core features of the dynamic residual sequence, including the sign of the residual, the real-time rate of change of the residual, the correlation between the residual data and environmental parameters, and the consistency between residual fluctuations and the rate of change of gas concentration. Accurate classification of anomalies is achieved through multi-feature combination matching: if the residual exhibits a slow, monotonic change and the fluctuation trend is unrelated to the actual gas concentration fluctuations on site, it is determined to be an anomaly of sensor zero-point drift or sensitivity decay. Simultaneously, the estimated remaining lifespan of the sensor is calculated, and maintenance suggestions for equipment calibration and on-demand sensor replacement are generated; if the residual shows a sudden, large jump or a long-term constant output... If the residual fluctuations are irregular and violent, it is determined to be a hardware failure such as equipment circuit failure or sensor component damage, and an emergency shutdown and maintenance work order is immediately generated. If the residual fluctuation pattern is highly correlated with environmental changes such as sudden changes in temperature and humidity or increased dust concentration, and the gas concentration does not show an abnormal upward trend, it is determined to be a false alarm caused by environmental interference, and maintenance suggestions for cleaning the equipment protective cover and fine-tuning the environmental compensation parameters are generated. If the residual shows a positive and rapid increasing trend, and is highly consistent with the upward trend of the gas concentration at the scene, while the simulation output of the digital twin benchmark model shows no abnormal fluctuations, it is determined to be a real combustible gas leak event, and the highest priority alarm command is generated, and safety handling commands such as exhaust ventilation and valve opening and closing are triggered simultaneously.

[0045] Differentiated maintenance work order generation and closed-loop maintenance process: Based on the identified abnormal scenario types and residual offset degradation levels, match the corresponding work order priorities and standardized processing time requirements; bind the unique number, precise geographical location, and installation point information of the abnormal equipment, integrate the cause of the abnormality, degradation level, and handling suggestions to form a complete work order content, and complete the work order packaging according to a unified standard format; through the IoT platform, accurately distribute the maintenance work order to the mobile terminals of maintenance personnel in the corresponding area, track the entire process of work order reception, handling, completion feedback, and combine maintenance results to iteratively optimize model parameters and judgment rules to achieve complete closed-loop maintenance management.

[0046] Corresponding to the aforementioned embodiments of the digital twin operation and maintenance method for combustible gas detection IoT devices, this application also provides embodiments of a digital twin operation and maintenance system for combustible gas detection IoT devices.

[0047] Figure 6 This is a schematic diagram illustrating the structural principle of a digital twin operation and maintenance system for a combustible gas detection Internet of Things (IoT) device, according to an exemplary embodiment. (Refer to...) Figure 6 The system may include: a real-time parameter set acquisition module, a digital twin model construction module, a dynamic residual sequence calculation module, and an operation and maintenance work order generation module; wherein: The real-time parameter set acquisition module is used to build a global sensing network for combustible gas detection IoT devices and collect multi-dimensional real-time parameter sets from each detection device in real time; the multi-dimensional real-time parameter sets include three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters. The digital twin model construction module is used to construct a digital twin benchmark model of all elements of the combustible gas detection equipment based on its inherent attributes, installation scenarios, and standard operating parameters. This model replicates the standard operating state and response logic of the physical equipment, forming a virtual benchmark body for real-time benchmarking and simulation. The digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. The dynamic residual sequence calculation module is used to input a multi-dimensional real-time parameter set into the digital twin benchmark model, obtain the simulation benchmark parameter sequence corresponding to the time series, perform frame-by-frame time series comparison between the real-time parameter sequence and the simulation benchmark parameter sequence, dynamically calculate the real-time residual of the multi-dimensional parameters, construct the dynamic residual sequence of the equipment, and quantify the degree of equipment anomaly and deterioration status through the fluctuation amplitude, offset trend and continuous characteristics of the residual sequence. The maintenance work order generation module is used to perform feature recognition and anomaly tracing on dynamic residual sequences, distinguish between four types of scenarios: sensor drift, hardware failure, environmental interference, and actual leakage, and generate differentiated and standardized maintenance work orders based on anomaly type, residual offset level, and equipment location information, and then send them to the maintenance terminal.

[0048] In this embodiment, the real-time parameter set acquisition module establishes a comprehensive sensing network for combustible gas detection IoT devices, acquiring multi-dimensional real-time parameter sets from each detection device in real time. These multi-dimensional real-time parameter sets include three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters. The digital twin model construction module, based on the inherent attributes, installation scenarios, and standard operating parameters of the combustible gas detection devices, constructs a digital twin benchmark model for all elements of the device, replicating the standard operating state and response logic of the physical device to form a virtual benchmark for real-time benchmarking simulation. This digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. The dynamic residual sequence calculation module inputs the multi-dimensional real-time parameter sets into the digital twin benchmark model to obtain the corresponding time-series data. The simulation baseline parameter sequence is used to perform frame-by-frame time-series comparison between the real-time parameter sequence and the simulation baseline parameter sequence, dynamically calculate the real-time residuals of multi-dimensional parameters, and construct a dynamic residual sequence for the equipment. The fluctuation amplitude, offset trend, and continuous characteristics of the residual sequence are used to quantify the degree of equipment anomaly and degradation status. The maintenance work order generation module performs feature recognition and anomaly tracing on the dynamic residual sequence, distinguishing between four scenarios: sensor drift, hardware failure, environmental interference, and actual leakage. Based on the anomaly type, residual offset level, and equipment location information, differentiated and standardized maintenance work orders are generated and sent to the maintenance terminal. By quantifying the time-series residuals to identify hidden equipment degradation, quantitative, precise, and closed-loop intelligent maintenance of combustible gas detection IoT devices is achieved, thereby meeting the high-precision, refined, and predictive full life-cycle maintenance needs of the equipment.

[0049] Regarding the system in the above embodiments, the specific ways in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0050] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0051] Accordingly, this application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; and when the one or more programs are executed by the one or more processors, causing the one or more processors to implement the digital twin operation and maintenance method for a combustible gas detection IoT device as described above. Figure 7 The diagram shown is a hardware structure diagram of any device with data processing capabilities, used in a digital twin operation and maintenance system for combustible gas detection IoT devices according to an embodiment of the present invention. (Except for...) Figure 7 In addition to the processor, memory, and network interface shown, any data processing device in the embodiment may also include other hardware depending on the actual function of the data processing device, which will not be described in detail here.

[0052] Accordingly, this application also provides a computer-readable storage medium storing computer instructions thereon, which, when executed by a processor, implement the digital twin operation and maintenance method for a combustible gas detection IoT device as described above. The computer-readable storage medium can be an internal storage unit of any data-processing device as described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, smart media card (SMC), SD card, flash card, etc., equipped on the device. Furthermore, the computer-readable storage medium can include both internal storage units of any data-processing device and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the data-processing device, and can also be used to temporarily store data that has been output or will be output.

[0053] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0054] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A digital twin operation and maintenance method for combustible gas detection IoT devices, characterized in that, Includes the following steps: A comprehensive sensing network of IoT devices for combustible gas detection is established to collect multi-dimensional real-time parameter sets from each detection device in real time. These multi-dimensional real-time parameter sets include three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters. Based on the inherent attributes, installation scenarios, and standard operating parameters of combustible gas detection equipment, a digital twin benchmark model of all equipment elements is constructed to replicate the standard operating state and response logic of the physical equipment, forming a virtual benchmark body for real-time benchmarking and simulation. The digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. Input the multi-dimensional real-time parameter set into the digital twin benchmark model to obtain the corresponding time-series simulation benchmark parameter sequence. Then, perform frame-by-frame time-series comparison between the real-time parameter sequence and the simulation benchmark parameter sequence, dynamically calculate the real-time residuals of the multi-dimensional parameters, construct the dynamic residual sequence of the equipment, and quantify the degree of equipment anomaly and deterioration status through the fluctuation amplitude, offset trend, and continuous characteristics of the residual sequence. The system performs feature identification and anomaly tracing on dynamic residual sequences, distinguishing between four scenarios: sensor drift, hardware failure, environmental interference, and actual leakage. Based on the anomaly type, residual offset level, and equipment location information, it generates differentiated and standardized operation and maintenance work orders and sends them to the operation and maintenance terminal.

2. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 1, characterized in that, In the process of building a comprehensive sensing network for combustible gas detection IoT devices, and collecting multi-dimensional real-time parameter sets from each detection device in real time; these multi-dimensional real-time parameter sets include three types of time-series data: device operating parameters, gas detection parameters, and on-site environmental parameters: Deploy wireless sensing nodes with edge computing capabilities, and network various combustible gas detectors, temperature and humidity sensors, barometers and equipment power monitoring devices through a low-power wide area network to form a full-area perception topology covering the target area. Set the acquisition cycle, and synchronously read three types of time-series data in each acquisition cycle: equipment operating parameters, gas detection parameters, and on-site environmental parameters to obtain raw time-series data. Among them, the equipment operating parameters include operating voltage, operating current, signal strength, and continuous operating time; the gas detection parameters include instantaneous value of combustible gas concentration, concentration change rate, and original electrical signal amplitude of the sensor; and the on-site environmental parameters include ambient temperature, ambient humidity, atmospheric pressure, and dust concentration. The original time-series data is timestamped and missing frames due to network latency and packet loss are removed. A sliding window filtering method is used to smooth and denoise the original signal, retaining the key change features of the signal and outputting a multi-dimensional real-time parameter set.

3. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 2, characterized in that, After aligning the original time-series data with timestamps, removing missing frames due to network latency and packet loss, and smoothing and denoising the original signal using a sliding window filtering method while preserving key signal variation features, a multi-dimensional real-time parameter set is output: The multi-dimensional real-time parameter set is lightweight compressed at the edge and uploaded in a unified data encapsulation format.

4. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 1, characterized in that, Based on the inherent attributes, installation scenarios, and standard operating parameters of combustible gas detection equipment, a digital twin benchmark model of all equipment elements is constructed to replicate the standard operating state and response logic of the physical equipment, forming a virtual benchmark body for real-time benchmarking simulation. This digital twin benchmark model includes geometric benchmarks, physical performance benchmarks, operational behavior benchmarks, and scenario adaptation benchmarks. Establish physical performance benchmarks and construct the physical response characteristics of the equipment under the initial healthy state by combining aging and degradation experience curves; among which, physical performance benchmarks include sensor sensitivity nominal value, response time constant, zero-point output benchmark value and range; Define the operating behavior benchmark, which is the ideal response curve of the equipment to a standard concentration of combustible gas under standard environmental conditions; the operating behavior benchmark includes the dynamic response behavior of the entire process from gas contact sensor to signal output, the power consumption behavior and communication behavior of the equipment in different operating modes; Establish scenario adaptation benchmarks. Based on the spatial volume, ventilation conditions, distance to potential leakage sources, and distribution of obstructions at the actual installation locations of the equipment, construct environmental correction factors for different application scenarios to obtain a digital twin benchmark model.

5. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 4, characterized in that, Before establishing physical performance benchmarks and constructing the physical response characteristics of the equipment under initial healthy conditions using aging degradation experience curves; whereby the physical performance benchmarks include the nominal value of sensor sensitivity, response time constant, zero-point output benchmark value, and measurement range: Obtain geometric reference parameters from the equipment manufacturer's technical manual to construct the mapping relationship between the three-dimensional spatial layout and internal structure of the physical equipment; among which, the geometric reference parameters include the internal air chamber volume of the detector, the sensor probe installation position, and the structural features of the explosion-proof shell.

6. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 1, characterized in that, In the steps of inputting a multi-dimensional real-time parameter set into a digital twin benchmark model, obtaining the corresponding time-series simulation benchmark parameter sequence, performing frame-by-frame time-series comparison between the real-time parameter sequence and the simulation benchmark parameter sequence, dynamically calculating the real-time residuals of the multi-dimensional parameters, constructing the equipment dynamic residual sequence, and quantifying the degree of equipment anomaly and degradation state through the fluctuation amplitude, offset trend, and continuous characteristics of the residual sequence: The multi-dimensional real-time parameter set obtained from the IoT platform at the same timestamp is input into the digital twin benchmark model. Based on the currently input real-time environmental parameters and equipment operating parameters, the simulation benchmark value of the equipment under the environmental conditions is calculated and output, forming a simulation benchmark parameter sequence corresponding to the real-time data time sequence. The actual gas concentration value and the measured parameters of the sensor signal amplitude collected in real time are subtracted from the corresponding items in the simulation benchmark parameter sequence frame by frame to obtain the initial deviation value of each dimension at each sampling time. The deviation values ​​are then arranged in chronological order to construct the dynamic residual sequence of the equipment.

7. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 6, characterized in that, After subtracting the real-time collected actual gas concentration values ​​and sensor signal amplitude measured parameters from the corresponding items in the simulation baseline parameter sequence frame by frame to obtain the initial deviation values ​​of each dimension at each sampling time, and arranging the deviation values ​​in chronological order to construct the equipment dynamic residual sequence: Sliding window statistical analysis is performed on the dynamic residual sequence to calculate the average fluctuation amplitude of the residual within the window, the slope of the monotonically increasing / decreasing trend of the residual sequence, and the length of time the residual continues to exceed the preset noise band. The stability of the device output is quantified based on the fluctuation amplitude, the speed of sensor sensitivity decay / zero drift is quantified based on the offset trend, and the duration and stability of abnormal device states are quantified based on the continuous characteristics.

8. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 7, characterized in that, After quantifying the stability of the device output based on the fluctuation amplitude, quantifying the rate of sensor sensitivity decay / zero drift based on the offset trend, and quantifying the duration and stability of the device's abnormal state based on the continuous characteristics: The three quantitative indicators of residual sequence fluctuation amplitude, offset trend and persistence characteristics are integrated for evaluation to output the abnormality level of the device, and the degradation state is divided into three levels: slight drift, moderate aging and severe failure according to the slope of the offset trend.

9. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 1, characterized in that, In the process of performing feature identification and anomaly tracing on dynamic residual sequences, distinguishing between four scenarios—sensor drift, hardware failure, environmental interference, and actual leakage—and generating differentiated and standardized maintenance work orders based on anomaly type, residual offset level, and equipment location information, and then distributing these work orders to the maintenance terminal: Feature extraction is performed on the dynamic residual sequence, and the abnormal state is classified into four scenarios based on the combination of features: sensor drift, hardware failure, environmental interference, and real leakage. The features include the sign of the residual, the rate of change of the residual, the correlation between the residual and environmental parameters, and the consistency between the residual and the rate of change of gas concentration. Based on the identified anomaly type and residual offset level, the priority and processing time requirements of maintenance work orders are determined. Combining equipment location information, equipment number, and geographical location tag, the work order content is packaged into a standard format and then sent to the mobile terminals of maintenance personnel in the corresponding area through the Internet of Things platform. The execution status and feedback results of the work orders are tracked to form a closed-loop maintenance management system.

10. The digital twin operation and maintenance method for combustible gas detection IoT devices as described in claim 9, characterized in that, In the process of extracting features from dynamic residual sequences and classifying abnormal states into four categories based on feature combinations: sensor drift, hardware failure, environmental interference, and actual leakage; the features include the sign of the residual, the rate of change of the residual, the correlation between the residual and environmental parameters, and the consistency between the residual and the rate of change of gas concentration. For sensor drift scenarios, when the residual changes monotonically and slowly and is unrelated to the actual gas concentration fluctuation, it is identified as zero drift and sensitivity decay, and calibration and sensor replacement suggestions are generated, along with an estimated remaining lifespan. For hardware failure scenarios, when the residual shows a large jump, a continuous constant output, or irregular fluctuations, it is identified as a circuit fault or sensor damage, and an emergency shutdown repair work order is generated. For scenarios involving environmental interference, when residual fluctuations are highly correlated with sudden changes in ambient temperature and humidity and increases in dust concentration, and when there is no abnormal upward trend in gas concentration, false alarms caused by environmental factors are identified, and maintenance suggestions for cleaning equipment protective covers and adjusting environmental compensation parameters are generated. For real-world leak scenarios, when the residual shows a rapid positive increase that is consistent with the rising trend of the actual gas concentration, and the simulation baseline output does not change significantly, it is identified as a real leak event, generating the highest priority alarm and triggering the opening and closing command of the exhaust valve.

Images