Gas equipment remote control method and system based on internet of things data collection

Abstract

This application relates to the field of gas equipment monitoring and control technology, and provides a method and system for remote control of gas equipment based on Internet of Things (IoT) data acquisition. The method includes: monitoring the gas equipment to obtain equipment monitoring sequences, environmental monitoring sequences, and operational monitoring sequences; performing anomaly attention association based on the gas equipment monitoring sequences to establish anomaly vectors; predicting gas safety accidents based on the anomaly vectors using a safety accident learning architecture to obtain a first path; optimizing the first path using environmental coupling based on the environmental monitoring sequences to obtain a second path; optimizing the second path using operational interference based on the operational monitoring sequences to obtain a third path; and performing remote intervention control based on the third path. This addresses the problem of gas equipment lacking comprehensive analysis and prediction of multi-source monitoring data in complex operating environments, leading to difficulties in timely detection and effective intervention of abnormal states, thereby improving the safety and intelligence level of gas equipment operation.

Description

Technical Field

[0001] This application relates to the field of gas equipment monitoring and control technology, specifically to a remote control method and system for gas equipment based on Internet of Things (IoT) data acquisition. Background Technology

[0002] In the application of gas equipment, with the continuous expansion of urban gas pipeline networks and the increasing complexity of gas usage scenarios, the operational safety of gas equipment has become increasingly prominent. Traditional gas safety management relies heavily on regular manual inspections, preset static threshold alarms, and isolated equipment status monitoring. These methods have significant limitations when facing complex and ever-changing actual operating conditions. On the one hand, traditional methods struggle to achieve deep integration and correlation analysis of multi-dimensional data such as equipment operating status, surrounding environmental factors, and personnel operating behaviors, resulting in fragmented early warning information and an inability to accurately depict the evolution path of safety accidents. On the other hand, existing monitoring systems often lack forward-looking predictive capabilities, typically only responding passively after an anomaly has reached the alarm threshold, missing the optimal intervention opportunity and failing to truly prevent problems before they occur. Thus, once a gas leak or equipment malfunction occurs, it can easily lead to fires, explosions, and other accidents, posing a serious threat to life and property safety. Summary of the Invention

[0003] This application provides a remote control method and system for gas equipment based on Internet of Things (IoT) data acquisition. It aims to solve the technical problem that gas equipment lacks comprehensive analysis and prediction of multi-source monitoring data in complex operating environments, making it difficult to detect and effectively intervene in abnormal states in a timely manner. The system achieves the technical effect of identifying abnormalities in gas equipment, predicting and optimizing multiple paths of safety accidents based on IoT data, and remotely intervening and controlling the equipment, thereby improving the safety and intelligence level of gas equipment operation.

[0004] The first aspect disclosed in this application provides a method for remote control of gas equipment based on Internet of Things (IoT) data acquisition. The method includes: performing IoT monitoring on the gas equipment to obtain a gas equipment monitoring sequence, an equipment environment monitoring sequence, and an equipment operation monitoring sequence; performing anomaly attention association based on the gas equipment monitoring sequence to establish a gas equipment anomaly vector; predicting gas safety accidents based on the gas equipment anomaly vector using a safety accident learning architecture to obtain a first gas safety accident path; optimizing the first gas safety accident path using environmental factor coupling based on the equipment environment monitoring sequence to obtain a second gas safety accident path; optimizing the second gas safety accident path using operation interference based on the equipment operation monitoring sequence to obtain a third gas safety accident path; and performing remote intervention control on the gas equipment based on the third gas safety accident path.

[0005] Another aspect of this application discloses a remote control system for gas equipment based on Internet of Things (IoT) data acquisition. The system includes: a data monitoring module for IoT monitoring of the gas equipment to obtain a gas equipment monitoring sequence, an equipment environment monitoring sequence, and an equipment operation monitoring sequence; an anomaly association module for anomaly attention association based on the gas equipment monitoring sequences to establish a gas equipment anomaly vector; an accident prediction module for predicting gas safety accidents based on the gas equipment anomaly vector using a safety accident learning architecture to obtain a first path for a gas safety accident; an environmental coupling module for optimizing environmental factor coupling of the first path for a gas safety accident based on the equipment environment monitoring sequence to obtain a second path for a gas safety accident; an interference optimization module for optimizing operational interference of the second path for a gas safety accident based on the equipment operation monitoring sequence to obtain a third path for a gas safety accident; and a remote control module for remotely intervening and controlling the gas equipment based on the third path for a gas safety accident.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The aforementioned remote control method for gas equipment based on IoT data acquisition comprehensively monitors the gas equipment via the IoT, acquiring data on equipment operating status, surrounding environmental conditions, and operational behavior. Subsequently, anomaly vectors are generated by correlating abnormal features in the equipment monitoring data. A safety accident learning architecture is then used to predict potential accidents, resulting in the first accident path. Next, environmental monitoring data is used to couple and correct the prediction results, forming a second accident path. Finally, operational monitoring data is used to optimize the intervention, resulting in a third accident path. Based on the optimized accident paths, remote intervention and control of the gas equipment are implemented, achieving intelligent early warning and safety protection for the gas equipment.

[0008] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0010] Figure 1This is a flowchart illustrating a remote control method for gas equipment based on Internet of Things (IoT) data acquisition in one embodiment.

[0011] Figure 2 This is an architecture diagram of a remote control system for gas equipment based on Internet of Things (IoT) data acquisition in one embodiment.

[0012] Explanation of reference numerals in the attached diagram: Data monitoring module 11, Anomaly correlation module 12, Accident prediction module 13, Environmental coupling module 14, Interference optimization module 15, Remote control module 16. Detailed Implementation

[0013] This application provides a remote control method and system for gas equipment based on Internet of Things (IoT) data acquisition. This addresses the technical problem that gas equipment lacks comprehensive analysis and prediction of multi-source monitoring data in complex operating environments, making it difficult to detect and effectively intervene in abnormal states in a timely manner. The system achieves the technical effect of identifying abnormalities in gas equipment, predicting and optimizing multiple paths of safety accidents based on IoT data, and remotely intervening and controlling the equipment, thereby improving the safety and intelligence level of gas equipment operation.

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0015] It should be noted that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such process, method, product, or device.

[0016] Example 1, as Figure 1 As shown, this application provides a remote control method for gas appliances based on Internet of Things (IoT) data acquisition, the method comprising:

[0017] By conducting IoT monitoring of gas equipment, we can obtain gas equipment monitoring sequences, equipment environment monitoring sequences, and equipment operation monitoring sequences.

[0018] In this embodiment, the gas equipment and its surrounding environment are first monitored via the Internet of Things (IoT) by deploying various types of sensors and data acquisition ports. This data includes equipment operating parameters, environmental conditions, and operational behaviors. Equipment operating parameters can be collected using pressure sensors, temperature sensors, flow meters, valve status detectors, etc. Environmental monitoring data can be collected using temperature and humidity sensors, gas concentration sensors, etc. Operational monitoring data can be collected through human-machine interfaces, operation recording modules, and intelligent control terminals. By classifying and storing this monitoring data, gas equipment monitoring sequences, environmental monitoring sequences, and operational monitoring sequences are established, providing sufficient and reliable data support for subsequent anomaly detection, accident prediction, and remote control.

[0019] Furthermore, this application provides IoT monitoring of gas equipment to obtain gas equipment monitoring sequences, equipment environmental monitoring sequences, and equipment operation monitoring sequences, including:

[0020] The gas equipment is monitored by the Internet of Things (IoT) to obtain IoT data collection results; the data is cleaned based on the IoT data collection results to obtain a device monitoring dataset; the device monitoring dataset is classified by features to obtain the gas equipment monitoring sequence, the device environment monitoring sequence, and the device operation monitoring sequence.

[0021] Preferably, firstly, various types of IoT sensors and data acquisition devices are deployed in and around the gas equipment, such as pressure sensors, temperature sensors, flow sensors, combustible gas concentration sensors, humidity sensors, noise and vibration sensors, as well as human-machine interface recording interfaces connected to the operating terminal. Then, the collected operating parameters, environmental information, and operating data are transmitted to the system via wireless or wired communication networks, forming the original IoT data acquisition results. Subsequently, the obtained IoT data acquisition results undergo data cleaning, including anomaly removal, missing value imputation, redundant data merging, time-series alignment, and unified format conversion, to eliminate noise data and anomalies generated during monitoring, thereby obtaining a high-quality, continuous equipment monitoring dataset. Anomaly removal can be performed using the σ principle and box plot detection; missing value imputation can be performed using the mean and forward imputation; redundant data merging can be performed by comparing timestamps and data sources; time-series alignment can be performed using interpolation, resampling, and clock synchronization protocols; and unified format conversion can be performed using ETL. Subsequently, based on the categories corresponding to each parameter recorded in the preset data feature classification rules, the equipment monitoring dataset is classified by feature. Data such as pressure, temperature, flow rate, and valve status, which are directly related to the operating status of gas equipment, are classified into gas equipment monitoring sequences. Data such as humidity, gas concentration, noise, and vibration, which are related to the external environment of the equipment, are classified into equipment environment monitoring sequences. Data such as control commands, operation logs, and historical operation records, which are related to operational behavior, are classified into equipment operation monitoring sequences. This provides a clear data foundation for subsequent anomaly detection, accident prediction, and remote control.

[0022] Anomaly vectors for gas equipment are established by performing anomaly attention association based on the gas equipment monitoring sequence.

[0023] In one embodiment, gas appliances of the same model or type are first interconnected and clustered. By comparing the operating characteristics of each appliance with those of normal appliances within the cluster, the degree of deviation between the current operating state and the normal state of the gas appliance is obtained. Then, an anomaly attention mechanism is used to weight the obtained deviations, assigning higher attention weights to features with larger deviations, thereby highlighting key anomalies that may cause safety hazards. Finally, the weighted features are concatenated to construct a gas appliance anomaly vector. This gas appliance anomaly vector can accurately represent the abnormal operating state of the gas appliance in multiple dimensions, providing a precise input basis for subsequent safety accident prediction and path deduction.

[0024] Furthermore, this application provides a method for establishing a gas equipment anomaly vector by performing anomaly attention association based on the gas equipment monitoring sequence, including:

[0025] Interconnect the gas equipment with the same model to obtain a gas equipment interconnection cluster; perform multi-degree-of-freedom standard state fitting based on the gas equipment interconnection cluster to construct a standard fitting sequence for the equipment; perform residual analysis based on the standard fitting sequence for the equipment and the gas equipment monitoring sequence to obtain a residual feature sequence for the equipment; perform anomaly attention allocation on the gas equipment monitoring sequence based on the residual feature sequence for the equipment to generate an anomaly vector for the gas equipment.

[0026] Preferably, firstly, based on the current gas equipment model, gas equipment of the same model is selected from the equipment registration database, and these gas equipment are interconnected to construct a gas equipment interconnection cluster. This gas equipment interconnection cluster includes equipment with the same design parameters. Subsequently, based on the management logs of the IoT platform, normal data samples of the gas equipment interconnection cluster are obtained, and these normal samples are fitted with a multi-degree-of-freedom standard state. After cleaning according to confidence levels, a standard fitting sequence of the gas equipment under normal conditions is obtained, which is used to characterize the baseline state of the gas equipment under ideal or healthy operating conditions. Then, the actual gas equipment monitoring sequence is compared point by point with the standard fitting sequence, and the difference is calculated to obtain the residual analysis result, thus forming the equipment residual feature sequence. This equipment residual feature sequence can reflect the degree and direction of the gas equipment deviating from the standard state in each dimension of operating parameters. Then, for the equipment residual feature sequence, an anomaly attention allocation mechanism is used to allocate weights, that is, the residual of each dimension in the equipment residual feature sequence is divided by the total residual to obtain the attention allocation weight of each dimension feature. This method assigns higher attention weights to feature data with large residual amplitudes and abnormal fluctuation frequencies, thereby highlighting abnormal factors that significantly affect the safety status of equipment. Finally, the obtained attention weights are multiplied by the corresponding features in the gas equipment monitoring sequence to obtain a weighted gas equipment monitoring sequence. This weighted sequence is then concatenated using vectorized encoding to form a gas equipment anomaly vector. This vector accurately represents the current abnormal state of the equipment and provides reliable input for subsequent accident prediction and intervention control, helping to achieve real-time and precise remote control of gas equipment.

[0027] Furthermore, this application provides a method for performing multi-degree-of-freedom standard state fitting based on the gas equipment interconnection cluster to construct a standard fitting sequence for the equipment, including:

[0028] A normal state sample retrieval is performed based on the gas equipment interconnection cluster to obtain a normal gas equipment sample library; multi-degree-of-freedom classification is performed based on the normal gas equipment sample library to obtain multiple normal equipment sample sequences; confidence level identification is performed on each normal equipment sample sequence to obtain multiple sample confidence level sequences; confidence level cleaning is performed on the multiple normal equipment sample sequences based on a predetermined confidence level to obtain multiple confidence normal sample sequences; central tendency analysis is performed on the multiple confidence normal sample sequences to generate the standard fitting sequence for the equipment.

[0029] Optionally, firstly, based on the unique ID of each gas device in the gas device interconnection cluster, a normal state sample retrieval is performed on the management logs of the IoT platform. This involves identifying data segments in the management logs that are operating stably, without faults, and whose fluctuations are within a reasonable range. These data are then stored uniformly to form a normal gas device sample library. This library contains data for multiple monitoring indicators, such as operating pressure, temperature, flow rate, and valve opening. Subsequently, based on this normal gas device sample library, multi-degree-of-freedom classification is performed according to different dimensions of the monitoring indicators, resulting in multiple normal sample sequences. Each normal sample sequence consists of multiple normal samples under the same monitoring indicator, such as pressure sequences, temperature sequences, and flow rate sequences, thus ensuring the independence and integrity of data across each dimension. Next, confidence level identification is performed on each normal sample sequence. This involves statistically analyzing the frequency of each normal sample in the sequence and calculating the corresponding sample confidence level, thus forming multiple sample confidence level sequences. Each sample confidence level sequence contains several confidence level values; higher values ​​indicate a greater probability of the sample occurring under normal conditions and higher reliability. Then, based on a pre-set confidence threshold, multiple normal sample sequences from various devices are cleaned to remove all samples with confidence levels below the threshold, retaining only the more reliable normal samples. This results in multiple confidence-based normal sample sequences. These sequences eliminate low-confidence samples that may contain abnormal fluctuations or errors, ensuring data reliability. Finally, central tendency analysis is performed on these multiple confidence-based normal sample sequences. Statistical methods, such as mean, variance, and cluster centroid extraction, are used to analyze the sequences and extract the central tendency characteristics of each monitoring indicator under normal conditions. By statistically analyzing the central tendency across various dimensions, a standard fitting sequence for the equipment is generated. This standard fitting sequence accurately characterizes the multidimensional benchmarks of gas equipment under normal operating conditions, providing a standard reference for residual analysis and anomaly detection.

[0030] Based on the safety accident learning architecture, the gas equipment is used to predict safety accidents based on the gas equipment anomaly vector, and the first path of gas safety accidents is obtained.

[0031] In one embodiment, after obtaining the gas equipment anomaly vector, the vector is input into the gas safety accident prediction node in the safety accident learning architecture. This node is obtained by using a distillation method to fuse features and minimize losses from multiple prediction models. It can perform feature analysis and risk mapping on the real-time generated gas equipment anomaly vector, outputting the possible accident evolution path of the gas equipment in its current state, thus obtaining the first gas safety accident path. This first path clearly describes the potential trajectory and key influencing factors of the accident, providing accurate initial prediction basis for subsequent environmental coupling optimization and operational intervention optimization, ensuring the safety of gas equipment operation.

[0032] Furthermore, this application provides a safety accident learning architecture that predicts safety accidents of the gas equipment based on the gas equipment anomaly vector to obtain a first path of gas safety accidents, including:

[0033] Gas safety accident record collection is performed based on gas equipment interconnection clusters to obtain a gas safety accident record library; the safety accident learning architecture is activated, which includes multiple edge nodes and cloud nodes; supervised training is performed on the gas safety accident record library based on the multiple edge nodes to obtain multiple gas safety accident prediction models; adversarial loss distillation learning is performed on the multiple gas safety accident prediction models based on the cloud nodes to generate gas safety accident prediction nodes; the gas equipment anomaly vector is input into the gas safety accident prediction nodes to obtain the first path of the gas safety accident.

[0034] Preferably, the first step is to collect safety incident records from the management logs of the IoT platform according to the unique ID of each gas device in the gas device interconnection cluster. Specifically, this involves identifying data fragments before and after the incident in the management logs and storing this data uniformly to form a gas safety incident record library. This library contains multi-dimensional information before and after the incident, such as equipment operating status parameters, environmental conditions, operational behaviors, and the evolution of the incident, providing a complete data foundation for learning and extracting incident patterns. Subsequently, the safety incident learning architecture is activated. This architecture consists of multiple distributed edge nodes and a centralized cloud node. The edge nodes possess rapid response and localized computing capabilities for parallel processing of incident data and model training; the cloud node undertakes global optimization and model fusion tasks. During the training phase, multiple edge nodes undergo supervised training on the gas safety accident record database. Specifically, equipment operating status parameters and accident evolution processes from the database are used as training samples. Deep neural networks, support vector machines, decision trees, and Markov chains are employed for feature learning and pattern recognition. Cross-entropy loss functions are used to optimize model parameters, and grid search is used to determine the optimal hyperparameter combination. This results in a corresponding gas safety accident prediction model for each edge node. Each model captures the evolutionary patterns of accidents, achieving sensitive prediction of specific accident types or characteristics. Subsequently, the cloud node performs adversarial loss distillation learning on the gas safety accident prediction models at each edge node. Through model testing, attention enhancement, and adversarial reinforcement learning, the distillation loss is minimized during training, generating a unified and more accurate gas safety accident prediction node. Then, during prediction, the real-time obtained gas equipment anomaly vector is input into the gas safety accident prediction node. This gas safety accident prediction node performs multi-dimensional feature analysis and accident path mapping on the input gas equipment anomaly vector, and outputs the first path of gas safety accident, that is, the accident evolution trajectory and potential risk points that the gas equipment may have under the current abnormal conditions. This first path of gas safety accident provides an accurate initial prediction basis for subsequent environmental coupling optimization and operational intervention optimization, thereby improving the timeliness and effectiveness of gas accident prevention and control.

[0035] Furthermore, this application provides a method for generating gas safety accident prediction nodes by performing adversarial loss distillation learning on the multiple gas safety accident prediction models based on the cloud nodes, including:

[0036] The gas safety accident record database is perturbed and injected into the cloud node to obtain a gas safety accident adversarial sample database. Multiple gas safety accident prediction models are tested using the adversarial sample database to obtain multiple adversarial sample loss sets. Attention enhancement is applied to the adversarial sample database using the multiple adversarial sample loss sets to obtain multiple gas safety accident adversarial enhancement databases. Adversarial enhancement learning is performed on the multiple gas safety accident prediction models using the multiple adversarial enhancement databases to obtain multiple gas safety accident predictors. Distillation loss minimization training is performed on the multiple gas safety accident predictors based on the cloud node to obtain the gas safety accident prediction node.

[0037] Optionally, firstly, based on the gas safety accident record database at the cloud node, a certain degree of noise or simulated abnormal factors are injected by changing the values ​​of some operating parameters, randomly adding environmental disturbance factors, or simulating operational errors to generate a safety accident adversarial sample database. This database is used to simulate data conditions under unpredictable or extreme real-world operating conditions to test the robustness and adaptability of the model. Subsequently, based on this database, multiple trained gas safety accident prediction models are tested, calculating the prediction bias of the models when facing adversarial samples, resulting in multiple adversarial sample loss sets. Each set contains the error of the corresponding prediction model under adversarial sample conditions, reflecting the model's stability and reliability in complex environments. Then, based on the multiple adversarial sample loss sets, the cloud node performs attention enhancement on the adversarial sample database. Specifically, it increases the weight of high-loss samples through an attention mechanism, enabling the model to pay more attention to adversarial samples that show significant deviations in prediction. This generates multiple enhanced adversarial sample databases that emphasize the role of critical and sensitive samples, ensuring that the model prioritizes correcting features with a greater impact on risk during training. Then, adversarial reinforcement learning is performed on multiple gas safety accident prediction models using multiple safety accident adversarial enhancement libraries. This involves introducing adversarial enhancement samples to optimize parameters based on the original training, enabling the models to better adapt to abnormal disturbances and improving their robustness and generalization ability. Through these steps, multiple prediction models are enhanced into multiple gas safety accident predictors. Finally, cloud nodes perform distillation loss minimization training on these multiple gas safety accident predictors. This involves unifying and fusing the knowledge of multiple predictors through model compression and knowledge transfer, maintaining the sensitivity of each predictor to key features while extracting their common accident prediction capabilities. By continuously iterating and optimizing to minimize distillation loss, a high-precision, highly robust gas safety accident prediction node is ultimately generated. This node can accurately analyze the anomaly vectors of gas equipment in complex environments and output reliable accident prediction paths, providing a foundation for subsequent environmental coupling optimization and remote control.

[0038] Based on the equipment environmental monitoring sequence, the first path of the gas safety accident is optimized by coupling environmental factors to obtain the second path of the gas safety accident.

[0039] In one embodiment, after obtaining the first path of a gas safety accident, this first path is mapped onto a 3D model of the gas equipment to visually reflect the propagation trajectory and evolution of the potential accident within and around the equipment. Subsequently, real-time data from the equipment's environmental monitoring sequence, such as ambient temperature, humidity, combustible gas concentration, air velocity, and vibration interference, are used to simulate the environmental effects on the 3D model. The simulation results are then coupled with the propagation process of the accident path to quantitatively describe the degree and direction of the coupling influence of environmental conditions on the accident evolution process. Afterward, the first path of the gas safety accident is dynamically corrected based on the coupling results, adjusting parameters such as the accident propagation speed and range to generate a second path. This second path not only considers the evolutionary characteristics of the equipment anomaly itself but also integrates the coupling effect of environmental conditions on accident development, enabling a more realistic and accurate reflection of the accident's evolution trend in actual operating scenarios and providing a reliable basis for subsequent operational intervention optimization and remote control.

[0040] Furthermore, this application provides a method for optimizing the first path of a gas safety accident based on the equipment environmental monitoring sequence to obtain a second path of a gas safety accident, including:

[0041] A three-dimensional reconstruction is performed on the gas equipment to obtain a gas equipment model; the first path of the gas safety accident is rendered onto the gas equipment model to obtain a gas accident model; environmental effects simulation is performed on the gas accident model based on the equipment environmental monitoring sequence to obtain accident environmental effect simulation data; environmental factor coupling influence analysis is performed on the first path of the gas safety accident based on the accident environmental effect simulation data to obtain accident environmental coupling analysis results; the first path of the gas safety accident is dynamically corrected based on the accident environmental coupling analysis results to generate the second path of the gas safety accident.

[0042] Preferably, the physical information of the gas equipment, such as its size, structure, installation location, and material properties, is first obtained from the design data of the gas equipment. This information is then imported into modeling software for 3D reconstruction to construct a gas equipment model. This model accurately reflects the equipment's geometry, operating components, and internal channels. Subsequently, the first path of a gas safety accident, predicted through anomaly vectors, is loaded and rendered into the gas equipment model to obtain a gas accident model. This model visually displays the propagation direction, diffusion range, and potential hazards of the accident within the equipment. Next, real-time data from the equipment's environmental monitoring sequence is used to simulate the environmental effects of the gas accident model. During the simulation, environmental parameters are used as dynamic input variables and coupled with the accident evolution path to simulate the impact of environmental conditions on the accident. For example, in the case of a gas leak accident, the diffusion process of the leaked gas under specific meteorological conditions is simulated, thus obtaining simulation data on the environmental effects of the accident. Then, environmental factor coupling analysis is performed on the simulation data of the accident environment. In this process, the evolution results of the first path of the gas safety accident under the baseline environmental conditions are used as a reference to extract its key indicators, such as propagation speed, diffusion range, and diffusion direction. Keeping other environmental factors constant, a certain environmental factor is perturbed to obtain new accident simulation results. After obtaining the new accident simulation results, the change magnitude of each key indicator is calculated and divided by the change in the environmental factor to obtain the sensitivity coefficient of each indicator. These sensitivity coefficients are then weighted to obtain the sensitivity result of that environmental factor. Similar operations are performed for other environmental factors to obtain the sensitivity result of each environmental factor. Finally, by dividing each sensitivity result by the largest sensitivity result and then weighting the normalized sensitivity results, the comprehensive environmental impact value is obtained, and this comprehensive environmental impact value is recorded as the accident environment coupling analysis result. Finally, based on the results of this accident environment coupling analysis, the first path of the gas safety accident is dynamically corrected. That is, the accident environment coupling analysis results are multiplied by preset scaling factors of each parameter in the path. These preset scaling factors are used to control the sensitivity of the comprehensive environmental impact to the path correction. The calculated product is then added to 1, and the sum is multiplied by the value of each parameter in the path. This corrects the propagation speed, diffusion range, propagation direction, etc. in the first path of the gas safety accident, forming a second path of the gas safety accident that is more consistent with the actual operating environment. This second path of the gas safety accident can more realistically reflect the evolution trend of the accident in the actual environment and provide a reliable basis for subsequent operational intervention optimization.

[0043] Based on the equipment operation monitoring sequence, the second path of the gas safety accident is optimized by operation interference to obtain the third path of the gas safety accident.

[0044] In one embodiment, after obtaining the second path of a gas safety accident, this second path is mapped onto a 3D model of the gas equipment. Real-time data from the equipment operation monitoring sequence, such as manual operation commands (e.g., valve opening or closing, emergency shutdown operations) and automatic control strategies (e.g., automatic shut-off after alarm triggering, fan start-up), is then used to simulate the operational effects on the 3D model. The simulation results are then interfering with the propagation process of the accident path to analyze the inhibitory, delaying, or deflective effects of operational behaviors on the accident propagation path. Subsequently, the second path of the gas safety accident is dynamically corrected based on the interference results, adjusting parameters such as the accident propagation speed and range to generate a third path. This third path not only considers the effects of equipment malfunctions and environmental factors but also integrates the interference effects of human and automatic operations, making it more closely aligned with the actual accident handling process and providing a reliable decision-making basis for remote intervention and control.

[0045] Furthermore, this application provides a method for optimizing the operational interference of the second path of a gas safety accident based on the equipment operation monitoring sequence to obtain a third path of a gas safety accident, including:

[0046] The second path of the gas safety accident is rendered onto the gas equipment model to obtain a gas accident update model; the operation effect simulation of the gas accident update model is performed based on the equipment operation monitoring sequence to obtain accident operation effect simulation data; the operation factor interference analysis of the second path of the gas safety accident is performed based on the accident operation effect simulation data to obtain accident operation interference analysis results; the second path of the gas safety accident is dynamically optimized based on the accident operation interference analysis results to generate the third path of the gas safety accident.

[0047] Preferably, the second path of the gas safety accident is first mapped and rendered into a 3D digital model of the gas equipment to obtain a gas accident update model. This gas accident update model can not only intuitively show the propagation trajectory of the accident within the equipment structure and its surrounding area, but also provide a dynamic interactive carrier for subsequent superimposed operational factors. Subsequently, operational effect simulation is performed on the gas accident update model based on the equipment operation monitoring sequence. By combining these operational behaviors with the gas accident update model, the evolution trend of the accident under operational intervention is simulated using control logic and physical effects, resulting in accident operational effect simulation data. Next, operational factor interference analysis is performed on the accident operational effect simulation data. In this process, key indicators are extracted from the accident operational effect simulation data, and these key indicators are compared with the corresponding indicators of the second path of the gas safety accident before operational intervention, obtaining the difference before and after the intervention. The calculated difference is then divided by the indicator before intervention to obtain the interference influence coefficient of each operation on the key parameters of the accident. By weighting the interference influence coefficient of each operation, the interference result corresponding to each operation is obtained. These interference results are then weighted to obtain the accident operational interference analysis result. Finally, the results of the accident operation interference analysis were applied to the second path of the gas safety accident for dynamic optimization. Following similar steps, parameters such as propagation speed, diffusion radius, and direction angle in the path were corrected, and the accident evolution trajectory was recalculated to form the third path of the gas safety accident. This third path of the gas safety accident integrates the influence of equipment abnormality, environmental factors, and operation interference, and is closer to the actual development of the accident under intervention conditions, providing a scientific basis for remote intervention control and emergency decision-making.

[0048] Furthermore, this application provides a third path to obtaining information about gas safety incidents, including:

[0049] Based on the third path of the gas safety accident, a gas equipment warning signal is generated.

[0050] Preferably, after obtaining the third path of a gas safety accident, the third path is analyzed to extract key parameters such as the accident propagation speed, diffusion range, diffusion direction, and potential hazardous node locations. These key parameters are then compared with preset safety thresholds to determine whether the accident propagation speed exceeds permissible limits, whether the diffusion range reaches densely populated areas or critical facilities, and whether the accident direction may affect flammable or high-pressure equipment. Following this, the corresponding risk level is matched based on the judgment results, and a gas equipment warning signal is automatically generated based on this risk level. This warning signal includes alarm level, trigger time, key parameters, and other information, and can be output in various ways, such as on-site alerts via audible and visual alarms, transmission to the monitoring center via an IoT communication module, or push notifications to operators' mobile terminals, thereby improving the safety and emergency response capabilities of gas equipment operation.

[0051] The gas equipment is remotely intervened and controlled according to the third path of the gas safety accident.

[0052] In one embodiment, a gas equipment warning signal generated based on a third path of a gas safety accident will invoke a corresponding intervention control command from a pre-set safety strategy library. This intervention control command can be various operational methods such as emergency valve closure, pipeline segment isolation, gas supply cut-off, ventilation and exhaust equipment activation, ignition source isolation, and alarm system linkage. Subsequently, the matched intervention control command is remotely transmitted to the execution terminal of the gas equipment via IoT communication, such as an electric valve controller, relay switch, fan driver, or automatic flameout device. Upon receiving the remote command, the execution terminal will immediately execute the corresponding action to block, weaken, or deflect the possible propagation path of the accident, thereby reducing the probability of the accident or mitigating its impact range, ensuring that the gas equipment is quickly and effectively handled safely in a remote state, and achieving proactive accident prevention and dynamic control.

[0053] In summary, the embodiments of this application have at least the following technical effects:

[0054] This application first performs IoT monitoring on the gas equipment to obtain a gas equipment monitoring sequence, an equipment environment monitoring sequence, and an equipment operation monitoring sequence. Then, it establishes a gas equipment anomaly vector by performing anomaly attention association based on the gas equipment monitoring sequence. Next, based on a safety accident learning architecture, it predicts safety accidents for the gas equipment based on the gas equipment anomaly vector, obtaining a first path for a gas safety accident. Further, it optimizes the first path for gas safety accidents by coupling environmental factors based on the equipment environment monitoring sequence, obtaining a second path for gas safety accidents. Then, it optimizes the second path for gas safety accidents by performing operational interference based on the equipment operation monitoring sequence, obtaining a third path for gas safety accidents. Finally, it performs remote intervention control on the gas equipment based on the third path for gas safety accidents. These technical effects collectively solve the technical problem of gas equipment lacking comprehensive analysis and prediction of multi-source monitoring data in complex operating environments, leading to difficulties in timely detection and effective intervention of abnormal states. This achieves the technical effect of realizing anomaly identification, multi-path prediction and optimization of safety accidents, and remote intervention control of gas equipment based on IoT data, thereby improving the safety and intelligence level of gas equipment operation.

[0055] Example 2 is based on the same inventive concept as the remote control method for gas equipment based on IoT data acquisition in the previous examples, such as... Figure 2 As shown, this application provides a remote control system for gas equipment based on Internet of Things (IoT) data acquisition. The system includes: a data monitoring module 11: performing IoT monitoring on the gas equipment to obtain a gas equipment monitoring sequence, an equipment environment monitoring sequence, and an equipment operation monitoring sequence; an anomaly association module 12: performing anomaly attention association based on the gas equipment monitoring sequence to establish a gas equipment anomaly vector; an accident prediction module 13: based on a safety accident learning architecture, performing safety accident prediction on the gas equipment according to the gas equipment anomaly vector to obtain a first path for a gas safety accident; an environment coupling module 14: performing environmental factor coupling optimization on the first path for a gas safety accident based on the equipment environment monitoring sequence to obtain a second path for a gas safety accident; an interference optimization module 15: performing operation interference optimization on the second path for a gas safety accident based on the equipment operation monitoring sequence to obtain a third path for a gas safety accident; and a remote control module 16: performing remote intervention control on the gas equipment according to the third path for a gas safety accident.

[0056] Furthermore, the data monitoring module 11 is also used to perform the following methods:

[0057] The gas equipment is monitored by the Internet of Things (IoT) to obtain IoT data collection results; the data is cleaned based on the IoT data collection results to obtain a device monitoring dataset; the device monitoring dataset is classified by features to obtain the gas equipment monitoring sequence, the device environment monitoring sequence, and the device operation monitoring sequence.

[0058] Furthermore, the anomaly association module 12 is also used to perform the following method:

[0059] Interconnect the gas equipment with the same model to obtain a gas equipment interconnection cluster; perform multi-degree-of-freedom standard state fitting based on the gas equipment interconnection cluster to construct a standard fitting sequence for the equipment; perform residual analysis based on the standard fitting sequence for the equipment and the gas equipment monitoring sequence to obtain a residual feature sequence for the equipment; perform anomaly attention allocation on the gas equipment monitoring sequence based on the residual feature sequence for the equipment to generate an anomaly vector for the gas equipment.

[0060] Furthermore, the anomaly association module 12 is also used to perform the following method:

[0061] A normal state sample retrieval is performed based on the gas equipment interconnection cluster to obtain a normal gas equipment sample library; multi-degree-of-freedom classification is performed based on the normal gas equipment sample library to obtain multiple normal equipment sample sequences; confidence level identification is performed on each normal equipment sample sequence to obtain multiple sample confidence level sequences; confidence level cleaning is performed on the multiple normal equipment sample sequences based on a predetermined confidence level to obtain multiple confidence normal sample sequences; central tendency analysis is performed on the multiple confidence normal sample sequences to generate the standard fitting sequence for the equipment.

[0062] Furthermore, the accident prediction module 13 is also used to perform the following method:

[0063] Gas safety accident record collection is performed based on gas equipment interconnection clusters to obtain a gas safety accident record library; the safety accident learning architecture is activated, which includes multiple edge nodes and cloud nodes; supervised training is performed on the gas safety accident record library based on the multiple edge nodes to obtain multiple gas safety accident prediction models; adversarial loss distillation learning is performed on the multiple gas safety accident prediction models based on the cloud nodes to generate gas safety accident prediction nodes; the gas equipment anomaly vector is input into the gas safety accident prediction nodes to obtain the first path of the gas safety accident.

[0064] Furthermore, the accident prediction module 13 is also used to perform the following method:

[0065] The gas safety accident record database is perturbed and injected into the cloud node to obtain a gas safety accident adversarial sample database. Multiple gas safety accident prediction models are tested using the adversarial sample database to obtain multiple adversarial sample loss sets. Attention enhancement is applied to the adversarial sample database using the multiple adversarial sample loss sets to obtain multiple gas safety accident adversarial enhancement databases. Adversarial enhancement learning is performed on the multiple gas safety accident prediction models using the multiple adversarial enhancement databases to obtain multiple gas safety accident predictors. Distillation loss minimization training is performed on the multiple gas safety accident predictors based on the cloud node to obtain the gas safety accident prediction node.

[0066] Furthermore, the environment coupling module 14 is also used to perform the following methods:

[0067] A three-dimensional reconstruction is performed on the gas equipment to obtain a gas equipment model; the first path of the gas safety accident is rendered onto the gas equipment model to obtain a gas accident model; environmental effects simulation is performed on the gas accident model based on the equipment environmental monitoring sequence to obtain accident environmental effect simulation data; environmental factor coupling influence analysis is performed on the first path of the gas safety accident based on the accident environmental effect simulation data to obtain accident environmental coupling analysis results; the first path of the gas safety accident is dynamically corrected based on the accident environmental coupling analysis results to generate the second path of the gas safety accident.

[0068] Furthermore, the interference optimization module 15 is also used to perform the following method:

[0069] The second path of the gas safety accident is rendered onto the gas equipment model to obtain a gas accident update model; the operation effect simulation of the gas accident update model is performed based on the equipment operation monitoring sequence to obtain accident operation effect simulation data; the operation factor interference analysis of the second path of the gas safety accident is performed based on the accident operation effect simulation data to obtain accident operation interference analysis results; the second path of the gas safety accident is dynamically optimized based on the accident operation interference analysis results to generate the third path of the gas safety accident.

[0070] Furthermore, the interference optimization module 15 is also used to perform the following method:

[0071] Based on the third path of the gas safety accident, a gas equipment warning signal is generated.

[0072] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0073] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0074] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A remote control method for gas equipment based on Internet of Things data acquisition, characterized in that, The method includes: Based on IoT monitoring of gas equipment, gas equipment monitoring sequences, equipment environment monitoring sequences, and equipment operation monitoring sequences are obtained; Anomaly vectors for gas equipment are established by performing anomaly attention association based on the gas equipment monitoring sequence. Based on the safety accident learning architecture, the gas equipment is used to predict safety accidents based on the gas equipment anomaly vector to obtain the first path of gas safety accidents. Based on the equipment environmental monitoring sequence, the first path of the gas safety accident is optimized by coupling environmental factors to obtain the second path of the gas safety accident. Based on the equipment operation monitoring sequence, the second path of the gas safety accident is optimized by operation interference to obtain the third path of the gas safety accident. The gas equipment is remotely intervened and controlled according to the third path of the gas safety accident. Based on the equipment environmental monitoring sequence, the first path of the gas safety accident is coupled and optimized with environmental factors to obtain the second path of the gas safety accident, including: A three-dimensional reconstruction of the gas equipment is performed to obtain a gas equipment model. The first path of the gas safety accident is rendered onto the gas equipment model to obtain the gas accident model; Based on the equipment environmental monitoring sequence, environmental impact simulation is performed on the gas accident model to obtain accident environmental impact simulation data; Based on the simulation data of the accident environment, the environmental factors coupling influence analysis of the first path of the gas safety accident is performed to obtain the accident environment coupling analysis results. Based on the accident environment coupling analysis results, the first path of the gas safety accident is dynamically corrected to generate the second path of the gas safety accident. Based on the equipment operation monitoring sequence, the second path of the gas safety accident is optimized by operational intervention to obtain the third path of the gas safety accident, including: The second path of the gas safety accident is rendered onto the gas equipment model to obtain the updated gas accident model. Based on the equipment operation monitoring sequence, the gas accident update model is simulated for operational effects to obtain accident operational effect simulation data. Based on the simulation data of the accident operation effect, the second path of the gas safety accident is analyzed for operational factors to obtain the accident operation interference analysis results. Based on the analysis results of the accident operation interference, the second path of the gas safety accident is dynamically optimized to generate the third path of the gas safety accident.

2. The method for remote control of gas equipment based on Internet of Things data collection according to claim 1, characterized in that, Anomaly vectors for gas equipment are established by performing anomaly attention association based on the gas equipment monitoring sequence, including: Based on the gas equipment, interconnect the same type of equipment to obtain a gas equipment interconnection cluster; Based on the gas equipment interconnection cluster, multi-degree-of-freedom standard state fitting is performed to construct a standard fitting sequence for the equipment. Based on the equipment standard fitting sequence and the gas equipment monitoring sequence, residual analysis is performed to obtain the equipment residual feature sequence; Anomaly attention is assigned to the gas equipment monitoring sequence based on the equipment residual feature sequence to generate the gas equipment anomaly vector.

3. The method for remote control of gas equipment based on data collection of Internet of Things according to claim 2, characterized in that, Based on the gas equipment interconnection cluster, multi-degree-of-freedom standard state fitting is performed to construct a standard fitting sequence for the equipment, including: A normal state sample retrieval is performed based on the gas equipment interconnection cluster to obtain a normal gas equipment sample library; Based on the gas equipment normal sample library, perform multi-degree-of-freedom classification to obtain multiple normal sample sequences of equipment; Confidence levels are determined based on the normal sample sequences of each device to obtain multiple sample confidence sequences; Based on a predetermined confidence level, the multiple device normal sample sequences are cleaned according to the multiple sample confidence level sequences to obtain multiple confidence normal sample sequences. Based on the central tendency analysis of the multiple confidence normal sample sequences, the standard fitting sequence of the device is generated.

4. The method for remote control of gas equipment based on Internet of Things data collection according to claim 1, characterized in that, Based on a safety incident learning architecture, safety incident prediction is performed on the gas equipment according to the gas equipment anomaly vector to obtain the first path of a gas safety incident, including: Based on the gas equipment interconnection cluster, safety accident records are collected to obtain a gas safety accident record database; Activate the security incident learning architecture, which includes multiple edge nodes and cloud nodes; Based on the multiple edge nodes, the gas safety accident record database is trained under supervision to obtain multiple gas safety accident prediction models. Based on the cloud nodes, adversarial loss distillation learning is performed on the multiple gas safety accident prediction models to generate gas safety accident prediction nodes. The gas equipment anomaly vector is input into the gas safety accident prediction node to obtain the first path of the gas safety accident.

5. The method for remote control of gas equipment based on Internet of Things data collection according to claim 4, characterized in that, Based on the cloud nodes, adversarial loss distillation learning is performed on the multiple gas safety accident prediction models to generate gas safety accident prediction nodes, including: Based on the cloud node, a perturbation injection is performed on the gas safety accident record database to obtain a safety accident countermeasure sample database. The multiple gas safety accident prediction models were tested based on the aforementioned adversarial sample library to obtain multiple adversarial sample loss sets. Based on the multiple adversarial sample loss sets, attention enhancement is performed on the security incident adversarial sample library to obtain multiple security incident adversarial enhancement libraries; Based on the multiple safety accident adversarial enhancement libraries, adversarial enhancement learning is performed on the multiple gas safety accident prediction models to obtain multiple gas safety accident predictors. The gas safety accident prediction node is obtained by training the multiple gas safety accident predictors to minimize distillation loss based on the cloud node.

6. The gas equipment remote control method based on Internet of Things data collection of claim 1, wherein, Based on IoT monitoring of gas equipment, gas equipment monitoring sequences, equipment environmental monitoring sequences, and equipment operation monitoring sequences are obtained, including: The gas equipment is monitored by the Internet of Things (IoT) to obtain IoT data collection results; Data cleaning is performed based on the IoT data collection results to obtain a device monitoring dataset. The equipment monitoring dataset is subjected to feature classification to obtain the gas equipment monitoring sequence, the equipment environment monitoring sequence, and the equipment operation monitoring sequence.

7. The gas equipment remote control method based on Internet of Things data collection of claim 1, wherein, Obtaining the third path for gas safety incidents includes: Based on the third path of the gas safety accident, a gas equipment warning signal is generated.

8. A remote control system for gas equipment based on Internet of Things (IoT) data acquisition, characterized in that: The system is used to implement the remote control method for gas equipment based on Internet of Things data acquisition as described in any one of claims 1-7, the system comprising: Data monitoring module: Based on IoT monitoring of gas equipment, obtain gas equipment monitoring sequences, equipment environment monitoring sequences, and equipment operation monitoring sequences; Anomaly Association Module: Performs anomaly attention association based on the gas equipment monitoring sequence to establish a gas equipment anomaly vector; Accident prediction module: Based on the safety accident learning architecture, it predicts safety accidents of the gas equipment according to the anomaly vector of the gas equipment and obtains the first path of gas safety accidents; Environmental Coupling Module: Based on the equipment environmental monitoring sequence, the first path of the gas safety accident is optimized by coupling environmental factors to obtain the second path of the gas safety accident; Interference optimization module: Based on the equipment operation monitoring sequence, optimize the operation interference of the second path of the gas safety accident to obtain the third path of the gas safety accident; Remote control module: Performs remote intervention and control of the gas equipment according to the third path of the gas safety accident.

Images