Factory energy safety dynamic monitoring method based on Internet of Things and multi-mode perception

Abstract

The invention provides a factory energy safety dynamic monitoring method based on the Internet of Things and multi-mode perception, and belongs to the technical field of industrial Internet of Things. Comprising the following steps: acquiring electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of equipment in real time; carrying out aggregation and protocol conversion on the electrical parameters, the temperature data, the combustible gas leakage concentration and the vibration signals through a multi-protocol intelligent gateway, and uploading the electrical parameters, the temperature data, the combustible gas leakage concentration and the vibration signals to a locally deployed edge computing node in a preset period; preprocessing the received data at the edge computing node, and generating a quantitative risk index based on a dynamic risk assessment model fusing the real-time state of the equipment, the historical aging trend and the environmental parameters; and carrying out risk grade judgment according to the quantitative risk index, and executing a corresponding grading response. Through fusion of multi-modal sensing data and edge intelligence, comprehensive sensing, real-time quantitative evaluation and hierarchical intelligent response of factory energy safety risks are realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of industrial Internet of Things, and particularly relates to a factory energy safety dynamic monitoring method based on Internet of Things and multi-modal perception. BACKGROUND

[0002] In modern industrial production, the safe and stable operation of the factory energy system (including electricity, gas, steam, etc.) is the key to ensuring production continuity and personnel safety.

[0003] However, the traditional energy monitoring method relies on a single type of sensor (such as only monitoring current or temperature), and the perception dimension is limited, which is difficult to fully capture the potential composite risks of equipment. At the same time, the monitoring data of each subsystem often exists in isolation, forming an "information island", and lacking effective fusion analysis means. In addition, the existing system generally adopts a "threshold alarm" mechanism, and the response strategy is simple and extensive, which cannot be finely and intelligently disposed according to the risk level, resulting in a high false alarm rate and insufficient early warning capability for early hidden dangers. SUMMARY

[0004] In order to solve at least one aspect of the technical problems in the background art, the application provides a factory energy safety dynamic monitoring method based on Internet of Things and multi-modal perception, which realizes comprehensive perception, real-time quantitative evaluation and hierarchical intelligent response of factory energy safety risks by fusing multi-modal perception data and edge intelligence, and balances the rapidity of safety protection and the refinement of energy efficiency management.

[0005] The technical scheme adopted by the application is: The first aspect of the application provides a factory energy safety dynamic monitoring method based on Internet of Things and multi-modal perception, comprising: real-time collection of electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of equipment; aggregation and protocol conversion of the electrical parameters, the temperature data, the combustible gas leakage concentration and the vibration signals by a multi-protocol intelligent gateway, and uploading to a locally deployed edge computing node at a preset period; preprocessing of the received data at the edge computing node, and generating a quantitative risk index based on a dynamic risk assessment model that fuses real-time state, historical aging trend and environmental parameters of equipment; risk level determination according to the quantitative risk index, and performing corresponding hierarchical response: when determining as high risk, triggering an emergency control instruction to cut off energy supply; when determining as medium risk, generating a structured maintenance work order and pushing it to the management layer; when determining as low risk, feeding back energy efficiency optimization instructions to the operation layer executor.

[0006] According to one embodiment of the present application, the real-time acquisition device collects electrical parameters, temperature data, combustible gas leakage concentration and vibration signals, specifically: A current transformer, an infrared thermal imager, a combustible gas sensor and a vibration acceleration sensor are respectively arranged at the busbar of the power distribution cabinet, the valve of the gas pipeline and the bearing seat of the rotating motor to construct a multi-modal sensor network. The current transformer is used to collect three-phase current and voltage values. The infrared thermal imager is used to obtain a two-dimensional temperature distribution image of the surface of the device. The combustible gas sensor is used to monitor the concentration of methane, hydrogen or carbon monoxide in real time. The vibration acceleration sensor is used to collect vibration acceleration signals of the device in X, Y and Z directions.

[0007] According to one embodiment of the present application, the multi-protocol intelligent gateway aggregates and converts the electrical parameters, the temperature data, the combustible gas leakage concentration and the vibration signals, and uploads them to the locally deployed edge computing node at a preset period, specifically: The multi-protocol intelligent gateway accesses the current transformer using the Modbus-RTU protocol through the RS485 interface, accesses the vibration acceleration sensor through the CAN bus, accesses the infrared thermal imager through the ZigBee wireless network, and accesses the combustible gas sensor through the LoRa wireless network. The intelligent gateway timestamps and unifies the data format of heterogeneous data from different interfaces and protocols, and encapsulates them as messages in JSON structure. The encapsulated messages are uploaded to the locally deployed edge computing node in batches through the MQTT protocol at a period of 100 milliseconds.

[0008] According to one embodiment of the present application, the edge computing node preprocesses the received data, and generates a quantitative risk index based on a dynamic risk assessment model that fuses the real-time state of the device, historical aging trends and environmental parameters, specifically: The effective value, harmonic distortion rate and unbalance degree are calculated for the preprocessed three-phase current and voltage values. The two-dimensional temperature distribution image is regionally divided, and the average temperature of the hot spot region, the temperature rise rate and the difference from the ambient temperature are calculated. The combustible gas concentration value is trend analyzed, and the concentration change rate per unit time is calculated. The amplitude of 1x, 2x and bearing characteristic frequencies is extracted from the vibration spectrum, and the total vibration intensity is calculated. The calculated electrical characteristics, thermal characteristics, gas characteristics and vibration characteristics are input into a dynamic risk assessment model together with historical aging trend data of the equipment and current environmental parameters.

[0009] According to an embodiment of the present application, when it is determined to be high risk, an emergency control instruction is triggered to cut off the power supply, specifically: The edge computing node sends a digital quantity control signal to the intelligent circuit breaker or electric valve controller corresponding to the target equipment through industrial Ethernet; After receiving the control signal, the intelligent circuit breaker or electric valve controller drives the internal relay to act, cutting off the power supply of the equipment or closing the gas-liquid pipeline valve; After the actuator completes the cutting operation, it feeds back a "cut off" state signal to the edge computing node, forming a control closed loop, and the edge computing node uploads the event record to the management layer.

[0010] According to an embodiment of the present application, when it is determined to be medium risk, a structured maintenance work order is generated and pushed to the management layer, specifically: The edge computing node matches the corresponding diagnostic conclusion and treatment suggestion from the preset maintenance knowledge base according to the quantitative risk index and the type of the equipment, and generates a structured maintenance work order; The structured maintenance work order includes a unique identifier of the equipment, a description of the location of the equipment, a risk type, a quantitative risk index, a current parameter snapshot, a diagnostic conclusion, a recommended measure, and a recommended maintenance level.

[0011] According to an embodiment of the present application, when it is determined to be low risk, an energy efficiency optimization instruction is fed back to the actuator at the operation layer, specifically: Based on the edge computing node, the deviation between the current equipment operating parameters and the preset optimal energy efficiency curve or benchmark model is analyzed; An energy efficiency optimization instruction containing target operating parameters is generated, the target operating parameters at least including target rotating speed, target power valve opening degree; The energy efficiency optimization instruction is sent to the corresponding frequency converter, servo driver or intelligent regulating valve through the industrial communication protocol; The actuator adjusts the equipment operating state according to the received instruction, and feeds back the adjusted actual operating parameters to the edge computing node, forming a closed loop control.

[0012] The second aspect embodiment of the present application provides a factory energy safety dynamic monitoring device based on Internet of Things and multi-modal sensing, comprising: The acquisition module is adapted to acquire electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment in real time; A conversion module is adapted to aggregate and convert protocols of the electrical parameters, the temperature data, the combustible gas leakage concentration and the vibration signal through the multi-protocol intelligent gateway, and upload to a locally deployed edge computing node at a preset period; A risk index generation module is adapted to preprocess the received data at the edge computing node, and generate a quantitative risk index based on a dynamic risk assessment model of fusion equipment real-time state, historical aging trend and environmental parameters; A response control module is adapted to determine a risk level according to the quantitative risk index, and execute corresponding graded responses: when determining as high risk, triggering an emergency control instruction to cut off energy supply; when determining as medium risk, generating a structured maintenance work order and pushing to a management layer; when determining as low risk, feeding back an energy efficiency optimization instruction to an operation layer executor.

[0013] The third aspect of the present application provides an electronic device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the program to realize the method for dynamic monitoring of factory energy safety based on Internet of Things and multi-modal perception in any embodiment of the first aspect.

[0014] The embodiments of the present application also provide a non-volatile computer storage medium having computer executable instructions stored thereon, and the computer executable instructions are executed by a processor to realize the method for dynamic monitoring of factory energy safety based on Internet of Things and multi-modal perception in any embodiment of the first aspect.

[0015] Due to the adoption of the above technical solutions, the present application has the following beneficial effects: According to the method, the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment are collected in real time, multi-dimensional and all-around perception of the running state of the energy equipment is realized, the limitation of traditional single sensor monitoring is broken through, early signs of multiple potential risks such as electrical overload, local overheating, gas leakage and mechanical deterioration can be captured at the same time, and the comprehensiveness and accuracy of hidden danger identification are significantly improved. Secondly, the heterogeneous data from different sensors and using different communication protocols are aggregated and protocol-converted by the multi-protocol intelligent gateway, and uploaded to the local edge computing node at a preset period, which not only solves the communication problem of "seven countries and eight systems" of industrial field equipment, realizes efficient data aggregation, but also provides a solid data flow foundation for subsequent real-time analysis and rapid response through high-frequency data uploading, ensuring the timeliness of the monitoring system. Then, the received data is preprocessed by the locally deployed edge computing node, and a quantitative risk index is generated based on a dynamic risk assessment model that fuses the real-time state of the equipment, historical aging trends and environmental parameters. It integrates multi-source data analysis and converts complex equipment state into an intuitive and comparable quantitative value, realizing the leap from "qualitative judgment" to "quantitative analysis". The model not only considers the current working condition, but also combines historical data and environmental factors, making the evaluation result more scientific and accurate, effectively reducing the false alarm and missed alarm rate. Finally, the risk level is determined according to the generated quantitative risk index, and the corresponding grading response strategy is executed: when it is determined as high risk, the emergency control instruction is triggered to cut off the energy supply, realizing millisecond-level fast and safe protection and ensuring personal and equipment safety; when it is determined as medium risk, a structured maintenance work order is generated and pushed to the management layer, which standardizes and streamlines the hidden danger information, facilitating maintenance personnel to quickly locate and handle, realizing the change from "passive repair" to "active prevention"; when it is determined as low risk, the energy efficiency optimization instruction is fed back to the actuator at the operation layer, which fully utilizes the value of monitoring data and optimizes the equipment operation parameters under the premise of ensuring safety to achieve the purpose of energy saving and consumption reduction. In summary, through the synergistic effect of the above four steps, the method not only greatly improves the intelligent, fine and real-time level of factory energy safety management, but also builds a complete closed loop of "perception-analysis-decision-execution". BRIEF DESCRIPTION OF DRAWINGS

[0016] The drawings described herein are used to provide further understanding of the present application, and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute improper limitations on the present application. In the drawings: Figure 1 A flowchart of the method of factory energy safety dynamic monitoring based on Internet of Things and multi-modal perception provided by the embodiments of the present application is shown. Figure 2 A structural schematic diagram of a factory energy safety dynamic monitoring device based on Internet of Things and multi-modal perception is provided for an embodiment of the present application. Figure 3 A structural schematic diagram of an electronic device is provided for an embodiment of the present application.

[0017] Reference signs: 110, acquisition module; 120, conversion module; 130, risk index generation module; 140, response control module; 810, processor; 820, communication interface; 830, memory; 840, communication bus. DETAILED DESCRIPTION

[0018] In order to more clearly illustrate the overall concept of the present application, the following will be described in detail with reference to the accompanying drawings.

[0019] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, however, the present application can also be implemented in other ways different from those described herein, therefore, the protection scope of the present application is not limited by the specific embodiments disclosed below. It should be noted that the embodiments of the present application and the features in each embodiment can be combined with each other without conflict.

[0020] In the present application, unless otherwise explicitly specified and limited, the first feature is "on" or "under" the second feature, which can be direct contact between the first and second features, or indirect contact between the first and second features through an intermediate medium. In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0021] As Figure 1 shown, the first aspect embodiment of the present application provides a factory energy safety dynamic monitoring method based on Internet of Things and multi-modal perception, comprising: Step 100, real-time acquisition of electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the device.

[0022] Step 200, aggregation and protocol conversion of the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals by the multi-protocol intelligent gateway, and uploading to the locally deployed edge computing node at a preset period.

[0023] Step 300, pre-processing the received data at the edge computing node, and generating a quantitative risk index based on the dynamic risk assessment model of the fusion device real-time state, historical aging trend and environmental parameters.

[0024] Step 400, risk level determination according to the quantitative risk index, and execution of corresponding hierarchical response: when determining as high risk, triggering emergency control instruction to cut off energy supply. When determining as medium risk, generating structured maintenance work order and pushing to management layer. When determining as low risk, feedback energy efficiency optimization instruction to operation layer executor.

[0025] In step 100, by deploying current transformers, infrared thermal imagers, combustible gas sensors and vibration acceleration sensors at key nodes of factory energy equipment (such as power distribution cabinet busbar, gas pipeline valve, rotating motor bearing seat), a multi-modal sensor network is constructed to collect real-time electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment. This step breaks through the limitations of traditional monitoring relying on single sensor, realizes the synchronous perception of multi-dimensional states such as electrical, thermal, gas and vibration of the equipment, can fully capture early fault features caused by different reasons such as overcurrent, overheating, leakage, looseness or wear, significantly improves the breadth and depth of risk perception, provides a rich and comprehensive data basis for subsequent accurate analysis, and effectively reduces the risk of missed reports due to perception blind area.

[0026] In step 200, using a multi-protocol intelligent gateway to access sensor devices with different communication protocols through RS485, CAN, ZigBee, LoRa and other interfaces, the collected heterogeneous data is time-stamped, formatted and protocol-converted, and packaged as JSON messages, then uploaded to the local edge computing node in batches with a period of 100 milliseconds through the MQTT protocol. This step solves the "information island" problem of industrial field device communication protocol diversity and data standard inconsistency, realizes efficient aggregation and standardized transmission of multi-source data. High-frequency and low-latency data upload mechanism ensures the real-time performance of the monitoring system, providing stable and reliable data flow support for rapid decision-making on the edge side. At the same time, the intelligent gateway as a data aggregation center reduces the communication burden of the edge node, improves the data processing efficiency and stability of the whole system.

[0027] In step 300, the edge computing node pre-processes the received data for noise reduction, alignment, etc., and calculates electrical characteristics (effective value, harmonic distortion rate), thermal characteristics (hot spot temperature, temperature rise rate), gas characteristics (concentration change rate), and vibration characteristics (characteristic frequency amplitude, total vibration intensity) based on a dynamic risk assessment model that fuses device real-time state, historical aging trend, and environmental parameters, and finally generates a quantitative risk index of 0-100. This step is the core intelligent hub of the method, and through multi-modal data fusion and deep analysis, it converts complex device status into intuitive and quantifiable risk scores. This model takes into account the real-time working condition of the device, the historical degradation trajectory, and the external environmental influence, making the risk assessment more scientific and accurate, achieving a leap from "experience judgment" to "data-driven", and significantly improving the accuracy of early hidden danger identification, providing a reliable decision basis for graded response.

[0028] In step 400, risk level determination is performed according to the generated quantitative risk index, and corresponding graded response strategies are executed: when the index ≥ 80, it is determined as high risk, and emergency control instructions are triggered to cut off energy supply, achieving millisecond-level safety protection; when the index ≥ 50 and < 80, it is determined as medium risk, and a structured maintenance work order containing device ID, risk type, diagnosis conclusion, etc. is generated and pushed to the management layer, achieving preventive maintenance; when the index < 50, it is determined as low risk, and energy efficiency optimization instructions are fed back to the frequency converter and other actuators to optimize operating parameters. This step builds a "high, medium, low" three-level intelligent response closed loop, balancing safety and economy. It not only can quickly cut off the risk source in extreme cases to protect personnel and equipment safety, but also can convert medium risk into standardized maintenance tasks to improve operation and maintenance efficiency, while tapping energy saving potential in a safe state, achieving a comprehensive upgrade from "passive alarm" to "active governance" and "intelligent optimization", significantly improving the intelligent level of factory energy management.

[0029] According to the method for monitoring factory energy safety based on the Internet of Things and multi-modal sensing provided in the first aspect of the present application, firstly, by collecting the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment in real time, multi-dimensional and all-around sensing of the running state of the energy equipment is realized, breaking through the limitations of traditional single sensor monitoring, and early signs of multiple potential risks such as electrical overload, local overheating, gas leakage and mechanical deterioration can be captured at the same time, significantly improving the comprehensiveness and accuracy of hidden danger identification. Secondly, the multi-protocol intelligent gateway is used to aggregate and protocol convert the heterogeneous data from different sensors using different communication protocols, and upload them to the local edge computing node at a preset period, which not only solves the communication problem of "seven countries and eight systems" of industrial field equipment, realizes efficient data aggregation, but also provides a solid data flow foundation for subsequent real-time analysis and rapid response through high-frequency data upload, ensuring the timeliness of the monitoring system. Then, the edge computing node deployed locally pre-processes the received data, and generates a quantitative risk index based on a dynamic risk assessment model that fuses real-time state, historical aging trend and environmental parameters of the equipment, which integrates multi-source data analysis and converts complex equipment state into an intuitive and comparable quantitative value, realizing the leap from "qualitative judgment" to "quantitative analysis". The model not only considers the current working condition, but also combines historical data and environmental factors, making the evaluation result more scientific and accurate, effectively reducing the false alarm and missed alarm rate. Finally, the risk level is determined according to the generated quantitative risk index, and the corresponding grading response strategy is executed: when it is determined as high risk, the emergency control instruction is triggered to cut off the energy supply, realizing millisecond-level fast and safe protection, and ensuring personal and equipment safety; when it is determined as medium risk, a structured maintenance work order is generated and pushed to the management layer, which standardizes and streamlines the hidden danger information, facilitating maintenance personnel to quickly locate and handle, realizing the change from "passive repair" to "active prevention"; when it is determined as low risk, the energy efficiency optimization instruction is fed back to the actuator at the operation layer, which fully utilizes the value of monitoring data, optimizes the equipment operation parameters under the premise of safety, and achieves the purpose of energy saving and consumption reduction. In summary, through the synergistic effect of the above four steps, the method not only greatly improves the intelligent, refined and real-time level of factory energy safety management, but also builds a complete closed loop of "sensing-analysis-decision-execution".

[0030] In some embodiments of the present application, the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment are collected in real time, specifically: A current transformer, an infrared thermal imager, a combustible gas sensor and a vibration acceleration sensor are respectively deployed at the busbar of the power distribution cabinet, the valve of the gas pipeline and the bearing seat of the rotating motor, to build a multi-modal sensor network; The current transformer is used to collect three-phase current and voltage values. Infrared thermal imager is used to obtain two-dimensional temperature distribution images of the equipment surface. Combustible gas sensor is used to monitor the concentration of methane, hydrogen or carbon monoxide in real time. Vibration acceleration sensor is used to collect vibration acceleration signals of the equipment in X, Y and Z directions.

[0031] Current transformer is installed on the busbar of the power distribution cabinet, used to accurately collect three-phase current and voltage values. These data are crucial for assessing the health of the power system, which can help detect problems such as overload, short circuit or unbalanced load.

[0032] Infrared thermal imager is deployed at key positions such as gas pipeline valves, which can obtain two-dimensional temperature distribution images of the equipment surface. By analyzing these thermal images, it can be found whether there are overheating points in the equipment, so as to early warn the risk of failure caused by overheating.

[0033] Combustible gas sensor is specially designed to monitor the concentration of methane, hydrogen or carbon monoxide in a specific area in real time. They are usually installed in places where gas leakage may occur, such as near gas pipelines, to ensure timely detection of dangerous gas leakage and ensure the safe operation of the plant.

[0034] Vibration acceleration sensor is installed at the bearing seat of rotating motor, used to measure the vibration acceleration signals of the equipment in X, Y and Z directions. By analyzing the vibration pattern, it can be identified whether the mechanical equipment has abnormal vibration, which is often an early sign of equipment wear or failure.

[0035] By integrating the above four types of sensors, a multi-modal sensor network can provide comprehensive monitoring coverage, collecting data from multiple dimensions such as electrical parameters, temperature changes, gas concentrations and mechanical vibrations.

[0036] In some embodiments of the present application, the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals are aggregated and protocol-converted by a multi-protocol intelligent gateway, and uploaded to a locally deployed edge computing node at a preset period, specifically: The multi-protocol intelligent gateway accesses the current transformer using Modbus-RTU protocol through RS485 interface, accesses the vibration acceleration sensor through CAN bus, accesses the infrared thermal imager through ZigBee wireless network, and accesses the combustible gas sensor through LoRa wireless network; The intelligent gateway timestamps and unifies the data format of heterogeneous data from different interfaces and protocols, and encapsulates them into JSON structured messages; The encapsulated messages are uploaded to the locally deployed edge computing node in batches through MQTT protocol at a period of 100 milliseconds.

[0037] The multi-protocol intelligent gateway accesses the current transformer using the Modbus-RTU protocol through the RS485 interface, which is suitable for the common serial communication in industrial sites, and can stably and reliably obtain the three-phase current and voltage data of the power distribution equipment; the vibration acceleration sensor is accessed through the CAN bus, which utilizes its high anti-interference capability and real-time performance to ensure the accurate collection of the vibration signal of the rotating equipment; the infrared thermal imager is accessed through the ZigBee wireless network, which is suitable for short-distance and low-power wireless data transmission, and facilitates the flexible deployment of the thermal imaging equipment in complex industrial environments; the combustible gas sensor is accessed through the LoRa wireless network, which utilizes its long-distance and low-power characteristics to realize reliable monitoring of the gas concentration in scattered or remote areas. This multi-interface and multi-protocol access method effectively solves the communication problem of the “seven countries and eight systems” of industrial site equipment, and realizes the seamless integration of different brands and types of sensors.

[0038] At the data processing level, the intelligent gateway timestamps the heterogeneous data from different physical interfaces and communication protocols, ensuring that all data have a unified time reference, providing a key timing alignment basis for subsequent multi-modal data fusion analysis. At the same time, the gateway formats the data uniformly and encapsulates them as JSON structure messages with clear structure and easy to parse, realizing data standardization and greatly reducing the data analysis and processing complexity of the upper layer system (edge computing node).

[0039] Finally, the encapsulated JSON messages are uploaded in batches to the locally deployed edge computing node at a period of 100 milliseconds through the MQTT protocol, a lightweight and efficient Internet of Things message transmission protocol. The MQTT protocol is based on the publish / subscribe mode, has low bandwidth occupation, high reliability and supports line reconnection, etc., and is very suitable for industrial Internet of Things scenarios. The 100-millisecond upload period ensures the high real-time performance of the data, providing timely data support for rapid risk assessment and response on the edge side. Batch uploading optimizes network resource utilization while ensuring real-time performance, reducing communication overhead. In summary, this step realizes the efficient, reliable and real-time aggregation of multi-source heterogeneous data.

[0040] In some embodiments of the present application, the received data is preprocessed at the edge computing node, and a quantitative risk index is generated based on a dynamic risk assessment model of the fusion device real-time state, historical aging trend and environmental parameters, specifically: The effective value, harmonic distortion rate and unbalance degree are calculated for the preprocessed three-phase current and voltage values; The two-dimensional temperature distribution image is divided into regions, and the average temperature of the hot spot region, the temperature rise rate and the difference from the ambient temperature are calculated; The combustible gas concentration value is trend analyzed, and the concentration change rate per unit time is calculated; The amplitude of the 1st, 2nd frequency and bearing characteristic frequency of the vibration spectrum is extracted, and the total vibration intensity is calculated; The above calculated electrical, thermal, gas and vibration characteristics, together with the historical aging trend data of the equipment and the current environmental parameters, are input into the dynamic risk assessment model.

[0041] First, in the data preprocessing and feature extraction stage, the system deeply analyzes various types of sensor data and extracts key features with clear physical meaning and diagnostic value. For electrical parameters, the effective value is calculated to assess the load level, the total harmonic distortion (THD) is calculated to identify power quality problems, and the three-phase imbalance is calculated to discover potential circuit faults, which are the basic indicators for judging the health of the electrical system. For infrared thermal imaging data, by dividing the two-dimensional temperature distribution image into regions, the "hot spot" can be accurately located, and the average temperature, temperature rise rate and difference from the ambient temperature can be calculated, which not only reflects the severity of overheating, but also the temperature rise rate can effectively warn of impending thermal failure. For flammable gas concentration, the concentration change rate per unit time is more valuable than simply looking at the instantaneous concentration, a rapidly rising concentration curve is often a clear signal that the leak is expanding. For vibration signals, the amplitude of the 1st, 2nd frequency and bearing characteristic frequency is extracted by spectral analysis, which can diagnose specific types of mechanical failures such as imbalance, misalignment and bearing wear, and the total vibration intensity is calculated as a comprehensive evaluation of the overall vibration level of the device.

[0042] Secondly, in the risk fusion and quantification stage, the system deeply fuses the above extracted electrical, thermal, gas and vibration features with the historical aging trend data of the equipment (such as cumulative running time, historical maximum temperature, historical maximum vibration value, past maintenance records, etc.) and the current environmental parameters (such as ambient temperature and humidity, production load, etc.). It breaks the limitation of "only looking at the present, not considering the history", making the evaluation model have the ability of "memory" and "context awareness". For example, a medium temperature hot spot may be a serious problem for a new device, but it may be within the expected range for a long-term high-load running old device.

[0043] Finally, all these fused data are input into the dynamic risk assessment model (which can be built based on machine learning or deep learning algorithms), the model weighs all factors comprehensively and outputs a quantitative risk index of 0-100. This single value is intuitive and comparable, simplifying the complex multi-dimensional state evaluation result into an easily understood and decision-making indicator, providing a scientific and unified decision-making basis for subsequent grading response (high, medium, low risk).

[0044] In some embodiments of the present application, when the risk is determined to be high, an emergency control instruction is triggered to cut off the power supply, specifically: The edge computing node sends a digital control signal to the intelligent circuit breaker or electric valve controller corresponding to the target device through industrial Ethernet. After receiving the control signal, the intelligent circuit breaker or electric valve controller drives the internal relay to act, cutting off the power supply of the device or closing the gas-liquid pipeline valve. After the actuator completes the cutting operation, it feeds back a "cut off" status signal to the edge computing node, forming a control loop, and the edge computing node uploads the event record to the management layer.

[0045] First, the edge computing node sends a digital control signal to the intelligent circuit breaker or electric valve controller corresponding to the target device through industrial Ethernet. The key to this step is "edge computing node" and "industrial Ethernet". Placing decision-making and instruction generation in the local edge node avoids the delay of uploading data to the cloud and waiting for the instruction to be issued, achieving millisecond-level fast response and meeting the stringent real-time requirements of emergency cutting. Industrial Ethernet, as a widely used communication standard in industrial sites, has high bandwidth, low delay, strong anti-interference ability and high reliability, which can ensure the stable and accurate transmission of control instructions in complex industrial electromagnetic environments, avoiding misoperation or refusal to act due to communication failure.

[0046] Second, after receiving the control signal, the intelligent circuit breaker or electric valve controller drives the internal relay to act, cutting off the power supply of the device or closing the gas-liquid pipeline valve. This describes the execution process of the instruction. After receiving the signal, the intelligent circuit breaker will trigger its internal split excitation release, instantly disconnecting the main circuit to achieve rapid cutting of the power supply; while the electric valve controller will drive the motor to act, rotating the pipeline valve to the closed position to block the flow of flammable gas or liquid. This "intelligent" actuator can directly analyze the digital instructions from the control system and complete complex mechanical operations, which is the basis for realizing automatic safety interlocking.

[0047] Finally, after completing the shutdown operation, the actuator feeds back a "shutdown" status signal to the edge computing node, forming a control closed loop. This step is crucial as it establishes a complete "command-execution-feedback" closed loop control logic. Only after receiving the "shutdown" confirmation signal, the edge node can confirm that the safety action has been successfully executed. This avoids the uncertainty that may be brought by open-loop control, ensuring the effectiveness of safety protection. If no feedback is received within a set time, the system can determine that the execution has failed and trigger higher-level alarms or backup emergency measures. At the same time, the edge computing node uploads event records to the management layer, enabling management personnel to learn about the occurrence, processing process, and results of emergency events in the first time, facilitating post-traceability, analysis, and improvement, and realizing the information connection between on-site execution and upper management.

[0048] In some embodiments of the present application, when the risk is determined to be medium, a structured maintenance work order is generated and pushed to the management layer, specifically: The edge computing node matches the corresponding diagnostic conclusion and treatment suggestion from the pre-set maintenance knowledge base according to the quantified risk index and device type, and generates a structured maintenance work order. Wherein, the structured maintenance work order includes device unique identifier, device location description, risk type, quantified risk index, current parameter snapshot, diagnostic conclusion, recommended measures and recommended maintenance level.

[0049] First, the edge computing node matches the corresponding diagnostic conclusion and treatment suggestion from the pre-set maintenance knowledge base according to the quantified risk index and device type, and generates a structured maintenance work order. This step reflects the "intelligence" and "automation" of the system. The system is no longer just an alarm, but has a preliminary "diagnosis" capability. By taking the "quantified risk index" (reflecting the severity of the risk) and "device type" (such as motor, transformer, valve, etc., different devices have different failure modes) as inputs, the system can automatically infer the most likely fault cause (diagnostic conclusion) and recommended treatment method (treatment suggestion) by intelligently matching in the pre-constructed "maintenance knowledge base". This rule-based or machine learning model-based knowledge base, which combines expert experience and historical data, makes the diagnostic results more professional and reliable, effectively reducing the dependence on on-site personnel experience.

[0050] Secondly, the generated structured maintenance work order contains device unique identifier, device location description, risk type, quantified risk index, current parameter snapshot, diagnostic conclusion, recommended measures and recommended maintenance level, etc., realizing the standardization and completeness of the work order content.

[0051] The device unique identifier and location description ensure that the maintenance personnel can quickly and accurately locate the target device.

[0052] Risk types and quantified risk indices allow maintenance personnel to have a clear understanding of the nature and severity of the problem before arriving on site.

[0053] The current parameter snapshot provides a multi-dimensional data record at the time of the incident, which is an important basis for analyzing the root cause of the failure.

[0054] The diagnostic conclusions and recommended measures directly provide the possible causes and solutions to the problem, which greatly shortens the troubleshooting time and improves maintenance efficiency and accuracy.

[0055] The suggested maintenance levels (such as emergency, high, medium, and low) provide management with a basis for decision-making, making it easier to allocate maintenance resources and priorities rationally.

[0056] In summary, this mechanism, by automatically generating structured work orders containing rich contextual information, seamlessly integrates potential risks detected by the system into the enterprise's operation and maintenance management system. This drives a shift in the operation and maintenance model from traditional "passive emergency repair" to "proactive prevention" and "precision maintenance," significantly improving equipment availability and the scientific nature of maintenance work.

[0057] In some embodiments of this application, when the risk is determined to be low, an energy efficiency optimization command is fed back to the operator layer actuator, specifically as follows: Based on edge computing nodes, analyze the deviation between the current device operating parameters and the preset optimal energy efficiency curve or benchmark model; Generate energy efficiency optimization instructions containing target operating parameters, which include at least target speed, target power, and valve opening. Energy efficiency optimization commands are sent to the corresponding frequency converters, servo drives, or intelligent control valves via industrial communication protocols. The actuator adjusts the operating state of the equipment according to the received instructions and feeds back the adjusted actual operating parameters to the edge computing node to form a closed-loop control.

[0058] First, the deviation between the current equipment operating parameters and the preset optimal energy efficiency curve or benchmark model is analyzed based on edge computing nodes. This step forms the basis for energy efficiency optimization decisions. The system no longer focuses solely on "safety" but further pursues "optimal performance." Edge computing nodes continuously monitor the actual operating status of the equipment (such as motor speed, load power, valve opening, etc.) and compare it with the preset "optimal energy efficiency curve" or "benchmark model." This "optimal energy efficiency curve" can be the theoretical optimal value provided by the equipment manufacturer or the measured optimal operating condition derived from historical data analysis. By calculating the deviation between the actual value and the target value, the system can accurately identify whether the equipment is operating under overload, inefficiently, or wasting energy.

[0059] Secondly, the energy efficiency optimization instructions containing the target operating parameters are generated and sent to the corresponding frequency converters, servo drives or intelligent regulating valves through industrial communication protocols. When the system identifies the operating deviation, it will automatically generate a clear optimization instruction containing specific "target operating parameters", such as adjusting the motor speed to X revolutions, reducing the water pump power to Y kilowatts, and setting the regulating valve opening to Z%. These instructions are issued through Modbus TCP, Profinet and other industrial communication protocols, directly affecting the actuators in the field. The frequency converter adjusts the motor power frequency according to the target speed, the servo drive accurately controls the position and force of the actuator, and the intelligent regulating valve adjusts the fluid flow according to the instructions. This process realizes active intervention and fine control of the device operating state.

[0060] Finally, the actuators adjust the device operating state according to the received instructions and feed back the adjusted actual operating parameters to the edge computing node, forming a closed-loop control. This is the key to ensuring optimization effect. The system does not "fire and forget", but establishes a complete "measurement-analysis-decision-implementation-re-measurement" closed-loop control loop. After the actuator completes the adjustment, it will immediately feed back the new operating parameters to the edge node. The edge node again assesses whether the current state has approached or reached the optimal energy efficiency point. If there is still deviation, it will continue to issue new optimization instructions; if it has reached the standard, it will maintain the current state. This dynamic and continuous closed-loop adjustment ensures that the device is always running in the high-efficiency interval, effectively reducing unnecessary energy consumption and achieving real "dynamic" energy efficiency management.

[0061] As shown in Figure 2 The second aspect of the present application provides a factory energy safety dynamic monitoring device based on Internet of Things and multi-modal sensing, comprising: The acquisition module 110 is adapted to acquire the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the device in real time; The conversion module 120 is adapted to aggregate and convert the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals through a multi-protocol intelligent gateway, and upload them to the locally deployed edge computing node at a preset period; The risk index generation module 130 is adapted to preprocess the received data at the edge computing node, and generate a quantitative risk index based on a dynamic risk assessment model that integrates real-time device state, historical aging trend and environmental parameters; The response control module 140 is adapted to determine the risk level according to the quantitative risk index, and execute the corresponding graded response: when it is determined as high risk, trigger the emergency control instruction to cut off the energy supply; when it is determined as medium risk, generate a structured maintenance work order and push it to the management layer; when it is determined as low risk, feed back the energy efficiency optimization instruction to the operation layer actuator.

[0062] The device for dynamically monitoring factory energy safety based on the Internet of Things and multi-modal sensing provided by the second aspect of the present application can implement the method for dynamically monitoring factory energy safety based on the Internet of Things and multi-modal sensing in any of the embodiments of the first aspect, and thus can implement any of the technical effects of the method for dynamically monitoring factory energy safety based on the Internet of Things and multi-modal sensing, which will not be repeated here.

[0063] The third aspect of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for dynamically monitoring factory energy safety based on the Internet of Things and multi-modal sensing in any of the embodiments of the first aspect when executing the program.

[0064] Figure 3 An example of a schematic diagram of the physical structure of an electronic device is shown in Figure 3 The electronic device can include a processor 810, a communications interface 820, a memory 830, and a communications bus 840, wherein the processor 810, the communications interface 820, and the memory 830 can communicate with each other through the communications bus 840. The processor 810 can invoke the logical instructions in the memory 830 to execute the method for dynamically monitoring factory energy safety based on the Internet of Things and multi-modal sensing in any of the embodiments of the first aspect, which comprises: Step 100, collecting electrical parameters, temperature data, combustible gas leakage concentration, and vibration signals of the device in real time.

[0065] Step 200, aggregating and converting the electrical parameters, temperature data, combustible gas leakage concentration, and vibration signals through a multi-protocol intelligent gateway, and uploading them to a locally deployed edge computing node at a preset period.

[0066] Step 300, preprocessing the received data at the edge computing node, and generating a quantitative risk index based on a dynamic risk assessment model that fuses real-time state, historical aging trend, and environmental parameters of the device.

[0067] Step 400, determining the risk level according to the quantitative risk index, and executing corresponding hierarchical responses: when determining as high risk, triggering an emergency control instruction to cut off the energy supply; when determining as medium risk, generating a structured maintenance work order and pushing it to the management layer; and when determining as low risk, feeding back energy efficiency optimization instructions to the operation layer executor.

[0068] Moreover, the logic instructions in the memory 830 described above can be implemented in the form of software functional units and sold or used as independent products, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or parts of the present application that make essential contributions or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0069] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program can be executed by a processor to enable a computer to execute the method for dynamic monitoring of factory energy safety based on Internet of Things and multi-modal perception provided by the above-mentioned methods, the method comprising: Step 100, collecting electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment in real time.

[0070] Step 200, aggregating and protocol converting the electrical parameters, temperature data, combustible gas leakage concentration and vibration signals through a multi-protocol intelligent gateway, and uploading to a locally deployed edge computing node at a preset period.

[0071] Step 300, pre-processing the received data at the edge computing node, and generating a quantitative risk index based on a dynamic risk assessment model that fuses real-time state, historical aging trend and environmental parameters of the equipment.

[0072] Step 400, determining the risk level according to the quantitative risk index, and executing corresponding graded responses: when determining as high risk, triggering an emergency control instruction to cut off the energy supply; when determining as medium risk, generating a structured maintenance work order and pushing it to the management layer; when determining as low risk, feeding back energy efficiency optimization instructions to the operation layer executor.

[0073] In yet another aspect, the present application also provides a non-transitory computer readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the method for dynamic monitoring of factory energy safety based on Internet of Things and multi-modal perception provided by the above-mentioned methods, the method comprising: Step 100, real-time collection of electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment.

[0074] Step 200, aggregation and protocol conversion of electrical parameters, temperature data, combustible gas leakage concentration and vibration signals by a multi-protocol intelligent gateway, and uploading to a locally deployed edge computing node at a preset period.

[0075] Step 300, preprocessing of the received data at the edge computing node, and generating a quantitative risk index based on a dynamic risk assessment model of the fusion equipment real-time state, historical aging trend and environmental parameters.

[0076] Step 400, risk level determination according to the quantitative risk index, and executing corresponding hierarchical responses: when determining as high risk, triggering an emergency control instruction to cut off the energy supply; when determining as medium risk, generating a structured maintenance work order and pushing to the management layer; and when determining as low risk, feeding back an energy efficiency optimization instruction to the operation layer executor.

[0077] Finally, the application also provides a non-volatile computer storage medium having computer executable instructions stored thereon, which are executed by a processor to implement the factory energy safety dynamic monitoring method based on Internet of Things and multi-modal perception provided by the above methods, and the method comprises: Step 100, real-time collection of electrical parameters, temperature data, combustible gas leakage concentration and vibration signals of the equipment.

[0078] Step 200, aggregation and protocol conversion of electrical parameters, temperature data, combustible gas leakage concentration and vibration signals by a multi-protocol intelligent gateway, and uploading to a locally deployed edge computing node at a preset period.

[0079] Step 300, preprocessing of the received data at the edge computing node, and generating a quantitative risk index based on a dynamic risk assessment model of the fusion equipment real-time state, historical aging trend and environmental parameters.

[0080] Step 400, risk level determination according to the quantitative risk index, and executing corresponding hierarchical responses: when determining as high risk, triggering an emergency control instruction to cut off the energy supply; when determining as medium risk, generating a structured maintenance work order and pushing to the management layer; and when determining as low risk, feeding back an energy efficiency optimization instruction to the operation layer executor.

[0081] The places not mentioned in the application can be realized by using or referring to the existing technology.

[0082] The various embodiments in the specification are described in progressive manner, and the same or similar parts between the various embodiments can be mutually referred to, and each embodiment focuses on the difference from other embodiments.

[0083] The above merely provides an example of the present application, and is not intended to limit the present application. The present application can have various modifications and changes for those skilled in the art. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for dynamic monitoring of factory energy safety based on the Internet of Things and multimodal sensing, characterized in that, include: Real-time acquisition of equipment electrical parameters, temperature data, combustible gas leakage concentration, and vibration signals; The electrical parameters, temperature data, combustible gas leakage concentration, and vibration signal are aggregated and converted using a multi-protocol smart gateway, and then uploaded to the locally deployed edge computing node at a preset cycle. The received data is preprocessed at the edge computing node, and a quantitative risk index is generated based on a dynamic risk assessment model of the real-time status of the fusion device, historical aging trends, and environmental parameters. The risk level is determined based on the quantitative risk index, and the corresponding graded response is executed: when the risk level is determined to be high, an emergency control command is triggered to cut off the energy supply. When the risk level is determined to be medium, a structured maintenance work order is generated and pushed to management. When the risk level is determined to be low, an energy efficiency optimization command is sent to the work-level actuator.

2. The factory energy safety dynamic monitoring method based on the Internet of Things and multimodal sensing according to claim 1, characterized in that, The electrical parameters, temperature data, combustible gas leakage concentration, and vibration signals of the real-time acquisition device are specifically as follows: A multimodal sensor network is constructed by deploying current transformers, infrared thermal imagers, combustible gas sensors, and vibration acceleration sensors at the busbar of the power distribution cabinet, the valve of the gas pipeline, and the bearing seat of the rotating motor. The current transformer is used to collect three-phase current and voltage values; The infrared thermal imager is used to acquire a two-dimensional temperature distribution image of the device surface; The combustible gas sensor is used to monitor the concentration of methane, hydrogen, or carbon monoxide in real time. The vibration acceleration sensor is used to collect vibration acceleration signals of the equipment in the X, Y, and Z directions.

3. The factory energy safety dynamic monitoring method based on the Internet of Things and multimodal sensing according to claim 2, characterized in that, The process involves aggregating and converting the electrical parameters, temperature data, combustible gas leakage concentration, and vibration signal via a multi-protocol smart gateway, and then uploading them to a locally deployed edge computing node at a preset cycle. The multi-protocol smart gateway connects to a current transformer using the Modbus-RTU protocol via an RS485 interface, a vibration acceleration sensor via a CAN bus, an infrared thermal imager via a ZigBee wireless network, and a combustible gas sensor via a LoRa wireless network. The smart gateway timestamps and unifies the data format of heterogeneous data from different interfaces and protocols, encapsulating it into a JSON structure message. The encapsulated messages are uploaded in batches to the locally deployed edge computing nodes via the MQTT protocol at 100-millisecond intervals.

4. The factory energy safety dynamic monitoring method based on the Internet of Things and multimodal sensing according to claim 2, characterized in that, The received data is preprocessed at the edge computing node, and a quantitative risk index is generated based on a dynamic risk assessment model of the fusion device's real-time status, historical aging trends, and environmental parameters. Specifically: Calculate the effective values, harmonic distortion rate, and unbalance of the pre-processed three-phase current and voltage values; The two-dimensional temperature distribution image is divided into regions, and the average temperature, temperature rise rate, and difference between the hot spot region and the ambient temperature are calculated. Trend analysis was performed on the concentration values ​​of combustible gases to calculate the rate of concentration change per unit time. Extract the amplitudes of the first harmonic, second harmonic, and bearing characteristic frequency from the vibration spectrum, and calculate the total vibration intensity; The electrical, thermal, gas, and vibration characteristics calculated above, along with the equipment's historical aging trend data and current environmental parameters, are input into the dynamic risk assessment model.

5. The factory energy safety dynamic monitoring method based on the Internet of Things and multimodal sensing according to claim 1, characterized in that, When a high-risk situation is identified, an emergency control command is triggered to cut off the energy supply, specifically as follows: Edge computing nodes send digital control signals to the smart circuit breaker or electric valve controller corresponding to the target device via industrial Ethernet. After receiving the control signal, the intelligent circuit breaker or electric valve controller drives the internal relay to cut off the power supply to the equipment or close the gas and liquid pipeline valve. After the actuator completes the cut-off operation, it sends a "cut-off" status signal to the edge computing node, forming a control closed loop. The edge computing node then uploads the event record to the management layer.

6. The factory energy safety dynamic monitoring method based on the Internet of Things and multimodal sensing according to claim 1, characterized in that, When the risk level is determined to be medium, a structured maintenance work order is generated and pushed to the management level, specifically as follows: Edge computing nodes match corresponding diagnostic conclusions and handling suggestions from a pre-set maintenance knowledge base based on the quantified risk index and equipment type, and generate structured maintenance work orders. The structured maintenance work order includes a unique equipment identifier, equipment location description, risk type, quantified risk index, current parameter snapshot, diagnostic conclusion, recommended measures, and recommended maintenance level.

7. The method for dynamic monitoring of factory energy safety based on the Internet of Things and multimodal sensing according to claim 1, characterized in that, When the risk level is determined to be low, an energy efficiency optimization command is sent to the work-level actuator. Specifically: Based on edge computing nodes, analyze the deviation between the current device operating parameters and the preset optimal energy efficiency curve or benchmark model; Generate an energy efficiency tuning instruction containing target operating parameters, wherein the target operating parameters include at least the target speed and the target power valve opening; The energy efficiency optimization command is sent to the corresponding frequency converter, servo drive or intelligent regulating valve through an industrial communication protocol. The actuator adjusts the operating state of the equipment according to the received instructions and feeds back the adjusted actual operating parameters to the edge computing node to form a closed-loop control.

8. A dynamic monitoring device for factory energy safety based on the Internet of Things and multimodal sensing, characterized in that, include: The data acquisition module is suitable for real-time acquisition of equipment's electrical parameters, temperature data, combustible gas leakage concentration, and vibration signals. The conversion module is adapted to aggregate and convert the electrical parameters, temperature data, combustible gas leakage concentration and vibration signal through a multi-protocol smart gateway, and upload them to the locally deployed edge computing node at a preset cycle. The risk index generation module is suitable for preprocessing the received data at the edge computing node and generating a quantitative risk index based on a dynamic risk assessment model of the real-time status of the fusion device, historical aging trends, and environmental parameters. The response control module is adapted to determine the risk level based on the quantified risk index and execute the corresponding graded response: when the risk level is determined to be high, an emergency control command is triggered to cut off the energy supply. When the risk level is determined to be medium, a structured maintenance work order is generated and pushed to management. When the risk level is determined to be low, an energy efficiency optimization command is sent to the work-level actuator.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the factory energy safety dynamic monitoring method based on the Internet of Things and multimodal sensing as described in any one of claims 1 to 7.

10. A non-volatile computer storage medium storing computer-executable instructions thereon, characterized in that, When the computer-executable instructions are executed by the processor, they implement the dynamic monitoring method for factory energy safety based on the Internet of Things and multimodal sensing as described in any one of claims 1 to 7.

Images