Hydrogen storage station safety monitoring method and system based on wireless passive three-signal sensor

Abstract

The invention relates to a hydrogen storage station safety monitoring method and system based on a wireless passive three-signal sensor, and belongs to the technical field of safety monitoring. Wireless passive three-signal sensor nodes are deployed in a hydrogen storage station to form a three-dimensional monitoring network; the temperature, strain and hydrogen detection units of the sensor are simultaneously excited through the multi-frequency-band signal transceiver, so that parallel acquisition of three-parameter signals is realized; performing signal decoupling, temperature compensation and primary risk assessment by using an edge computing gateway; deep data mining and risk assessment are carried out through a cloud early warning platform in combination with a digital twinborn model; wherein multi-parameter signal decoupling is realized by adopting a frequency division multiplexing and wavelet denoising algorithm, and safety early warning is carried out in combination with a three-level early warning decision tree. The hydrogen storage station electric spark risk can be eliminated, high-precision synchronous monitoring of temperature, strain and hydrogen concentration is achieved, and the hydrogen leakage early warning time is shortened.

Description

Technical Field

[0001] The present invention belongs to the field of safety monitoring technology, and relates to wireless passive sensor network integration, multi-parameter collaborative monitoring and intelligent early warning. Specifically, it provides a hydrogen storage station safety monitoring system architecture, deployment method and application based on surface acoustic wave (SAW) wireless passive three-signal sensor. Background Art

[0002] With the rapid development of the hydrogen energy industry, hydrogen storage stations, as a key link in the hydrogen energy supply chain, are crucial for the safe and stable operation of the entire industry. However, existing hydrogen storage station monitoring technology has many problems and shortcomings, making it difficult to meet the safety and efficiency requirements of modern hydrogen storage stations.

[0003] In terms of hydrogen storage station monitoring technology, traditional solutions often rely on wired sensors, such as thermocouples and pressure transmitters, in conjunction with independent hydrogen detectors for monitoring. This monitoring method not only requires an external power supply and a large number of signal cables, which increases the complexity and cost of the system, but also has a high risk of electric sparks in the dangerous environment of hydrogen leaks, which may cause serious safety accidents. Some wireless monitoring solutions use battery-powered sensors, but the battery life of such sensors is limited, usually only lasting 3-6 months. In the high-risk environment of hydrogen storage stations, frequent manual battery replacement is not only costly, but may also pose huge safety risks due to human error.

[0004] Existing multi-parameter monitoring systems typically deploy sensors for temperature, pressure, and gas concentration independently, then integrate the data through a PLC or host computer. This monitoring approach results in extremely complex wiring, with the wiring cost of a single hydrogen storage tank accounting for over 30% of the total system cost. Furthermore, because the sampling interval for each sensor is typically greater than or equal to 100ms, signal synchronization between different parameters is poor. In situations such as sudden temperature changes and abnormal pressure, the correlation analysis delay exceeds 200ms, making it difficult to capture early signs of hydrogen embrittlement and resulting in significant monitoring lag. Furthermore, traditional systems also face severe electromagnetic interference issues, further impacting the accuracy and reliability of monitoring data.

[0005] In terms of wireless passive sensing applications, although some single-parameter wireless passive sensors, such as SAW temperature sensors, are currently being used in industrial scenarios, these single-parameter sensors lack the ability to monitor multiple parameters collaboratively and cannot meet the complex requirements of hydrogen storage stations for three-dimensional safety assessments of "temperature, pressure, and leakage." The safe operation of hydrogen storage stations requires real-time monitoring and comprehensive analysis of multiple key parameters to achieve comprehensive safety assessments and risk warnings.

[0006] The shortcomings and disadvantages of the existing technology are mainly reflected in the following aspects:

[0007] High safety risks: Wired-powered and battery-powered devices are susceptible to ignition accidents in environments with hydrogen explosion limits (4%-75%). For example, at 1:20 PM on March 20, 2025, an explosion occurred during construction at the Lotte SK Energy hydrogen fuel cell power plant construction site in Nam-gu, Ulsan, South Korea, resulting in two workers suffering second-degree burns to their faces and legs. An investigation into the accident revealed that the density of hydrogen sensors at the construction site was only 60% of the industry standard, failing to meet safety monitoring requirements.

[0008] The monitoring lag is obvious: the data of independent sensors are transmitted through different protocols, resulting in a delay of more than 200ms in the correlation analysis between temperature mutation and pressure anomaly. This makes it impossible to capture the early signs of hydrogen embrittlement in time, missing the best opportunity for early warning and treatment.

[0009] High operation and maintenance costs: Hydrogen storage stations have high annual sensor maintenance costs, and in high-risk environments, the frequency of manual inspections is limited, increasing the difficulty and cost of operation and maintenance.

[0010] Weak data correlation: Traditional monitoring systems lack a coupled model for temperature, strain, and hydrogen concentration, making them incapable of collaborative analysis across multiple parameters. For example, when a crack in a hydrogen storage tank propagates, a single hydrogen concentration alarm will lag behind the actual leak time, failing to accurately and promptly reflect potential safety risks.

[0011] Therefore, the existing hydrogen storage station monitoring technology has significant deficiencies in terms of safety, real-time performance, cost-effectiveness, and data correlation, and is unable to meet the needs of hydrogen storage stations for high-safety, high-precision, and low-cost monitoring. Summary of the Invention

[0012] The present invention aims to build a wireless passive multi-parameter safety monitoring system for hydrogen storage stations, utilizing the power-free characteristics of wireless passive three-signal sensors to eliminate the risk of electric sparks in hydrogen storage stations; through a distributed network architecture and a multi-parameter collaborative algorithm, high-precision synchronous monitoring of temperature, strain, and hydrogen concentration can be achieved; a safety assessment model based on machine learning is established to greatly shorten the hydrogen leak warning time; and the system's full life cycle cost is reduced, thereby reducing the annual operation and maintenance costs of hydrogen storage stations.

[0013] In order to achieve the above objectives, the technical solution of the present invention provides a hydrogen storage station safety monitoring method based on wireless passive three-signal sensors, in which wireless passive three-signal sensor nodes are deployed at the hydrogen storage station to form a three-dimensional monitoring network;

[0014] The multi-band signal transceiver simultaneously stimulates the temperature, strain, and hydrogen detection units of the sensor to achieve parallel acquisition of three-parameter signals;

[0015] Use edge computing gateways for signal decoupling, temperature compensation, and primary risk assessment;

[0016] Conduct in-depth data mining and risk assessment through the cloud-based early warning platform combined with the digital twin model;

[0017] Among them, frequency division multiplexing and wavelet denoising algorithm are used to achieve multi-parameter signal decoupling, and a three-level warning decision tree is combined to perform safety warning.

[0018] Preferably, the sensor signal decoupling process includes:

[0019] The sensor reflects mixed signals;

[0020] FFT spectrum analysis;

[0021] Identify the 200 / 250 / 300MHz peak value. If not, return to the previous step. If yes, proceed to the next step.

[0022] Frequency-parameter preliminary solution;

[0023] Temperature compensation calculation;

[0024] wavelet decomposition;

[0025] Thresholding detail components;

[0026] Signal reconstruction;

[0027] Parameter calibration output;

[0028] Edge gateway local storage / cloud transmission.

[0029] Preferably, the early warning decision tree process is as follows:

[0030] Real-time sensor data;

[0031] The first level of threshold judgment determines whether the threshold warning is triggered. If not, proceed to the next step;

[0032] The second level is trend analysis, which determines whether an abnormal trend is found. If so, it proceeds to the next step.

[0033] The third layer is association rule reasoning, which determines whether there is a parameter coupling risk. If so, the system returns to normal monitoring.

[0034] If the judgment is yes at any level, the corresponding level of warning will be output in sequence, the warning information will be integrated, and decision recommendations will be generated.

[0035] Preferably, the safety warning mechanism sets three levels of warning thresholds, namely warning, alarm and emergency;

[0036] Warning: If any of the following conditions is met: hydrogen concentration reaches 10% LEL, strain exceeds 50 με, or temperature exceeds 60°C, the abnormality will be recorded and continuously monitored;

[0037] Alarm: If two of the following conditions are met: hydrogen concentration reaches 25% LEL, strain exceeds 100 με, or temperature exceeds 80°C, the linkage protection measures will be activated;

[0038] Emergency: If the hydrogen concentration is ≥40%LEL, the strain rate is greater than 10με / s and the temperature exceeds 100°C, or if a single parameter changes suddenly, the machine will be shut down immediately and the fire protection system will be activated.

[0039] Optimized trend analysis mechanism process:

[0040] parameter time series;

[0041] Exponential smoothing;

[0042] Calculate the rate of change Δy / Δt;

[0043] Predict the value y in the next ten minutes pred , prediction model: S t =α×y t +(1-α)×S t-1 ;

[0044] Determine whether Δy / Δt ≥ threshold?

[0045] If so, judge y pre ≥ 90% of the warning threshold? If yes, trigger the warning in advance; otherwise, continue monitoring and update the forecast;

[0046] If otherwise, the abnormality index H=|(y t -μ) / α|≥3? If yes, identify potential gradual risk; otherwise, continue monitoring and update the forecast.

[0047] Preferably, the association rule reasoning mechanism process is as follows:

[0048] Measured values ​​of three parameters;

[0049] Calculate the temperature-strain theoretical value and determine whether the Δε measured-theoretical value is ≥15%. If so, it is determined that there is mechanical damage or leakage;

[0050] Evaluate the hydrogen-temperature risk product and determine whether C H2 ×(T-25)≥1000, if so, a warning of hydride decomposition risk is issued;

[0051] Analyze the strain-hydrogen coupling risk and determine whether ε≥80με and C H2 ≥15%LEL, if so, it will warn of hydrogen embrittlement crack expansion;

[0052] Produce multi-parameter correlation early warning reports.

[0053] The technical solution of the present invention also provides a hydrogen storage station safety monitoring system based on a wireless passive three-signal sensor, comprising:

[0054] Distributed sensor network, deploying 16-24 wireless passive three-signal sensor nodes in hydrogen storage tanks, pipelines, and compressor rooms;

[0055] Multi-band signal transceiver, using a 100-300MHz broadband RF module, simultaneously stimulates the sensor's temperature, strain, and hydrogen detection units;

[0056] Edge computing gateway, integrating spectrum analysis chip and processor, completes signal decoupling, temperature compensation and primary risk assessment;

[0057] The cloud-based early warning platform transmits data via 5G / fiber optics to build an AI analysis system that includes a digital twin model of the hydrogen storage tank, supporting historical data backtracking, trend prediction, and multi-level early warning.

[0058] Preferably, the deployment method of the distributed sensor network includes:

[0059] For hydrogen storage tanks, three sensors are evenly distributed around the tank at 1 / 3 of the height from the bottom of the tank;

[0060] For piping systems, deploy sensors 500 mm upstream and downstream of elbows, reducers, tees, crosses, and valves.

[0061] In the compressor room, monitoring nodes are set at the top of the hydrogen buffer tank and the pipeline interface.

[0062] Preferably, the multi-band signal transceiver uses frequency division multiplexing 200 / 250 / 300MHz and space division multiplexing. The sensor spacing is ≥1.5m, and the signal crosstalk is suppressed to below -40dB.

[0063] In summary, the present invention has the following beneficial technical effects:

[0064] This invention achieves "three-in-one" monitoring of temperature, strain, and hydrogen concentration, effectively establishing a three-dimensional assessment system for the safety status of hydrogen storage equipment, with a key parameter under-detection rate of less than 0.01%. The sensor has a wide operating temperature range of -40°C to 125°C and a vibration acceleration resistance of up to 100G, making it perfectly adapted for low-temperature storage and transportation of hydrogen storage stations and high-frequency vibration scenarios of compressors. At the same time, the LSTM model trained based on historical data can accurately predict the corrosion rate of hydrogen storage tanks 72 hours in advance, with a prediction error of less than 8%, providing a scientific basis for planned maintenance of hydrogen storage stations.

[0065] The present invention has strong scene adaptation capabilities and is suitable for various hydrogen storage scenarios such as high-pressure gaseous hydrogen storage (35MPa / 70MPa), liquid hydrogen storage (-253°C), and organic liquid hydrogen storage. Its hydrogen leak detection response time is less than 60 seconds, which meets the requirements of GB50516-2021 "Technical Standards for Hydrogen Refueling Stations", and the early warning time is 30-45 seconds earlier than the traditional system. In addition, the present invention reduces the technical threshold and cost of safety monitoring of hydrogen energy infrastructure, provides technical support for the large-scale construction of hydrogen refueling stations, and is expected to reduce the safety accident rate of hydrogen storage stations by more than 70%.

[0066] This invention achieves a technological breakthrough in hydrogen storage station safety monitoring through three levels of technological innovation. It uses a wireless passive three-signal sensor with a power-free design, combined with explosion-proof packaging and broadband radio frequency communication technology, to eliminate the risk of electric sparks in hydrogen environments and achieve continuous operation for 1,000 hours in a 99% LEL hydrogen environment. Frequency division multiplexing and wavelet denoising algorithms achieve efficient decoupling of multi-parameter signals, and a precise temperature compensation model is used to control synchronous sampling delays to less than 50 milliseconds. A three-level early warning decision tree and distributed network architecture are constructed, integrating a digital twin model to achieve leak source location accuracy of ≤2 meters. The early warning response speed is 40 times faster than traditional systems, fully meeting the high-safety, high-precision, and low-cost monitoring needs of hydrogen storage stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 The topology diagram of the hydrogen storage station monitoring system in the hydrogen storage station safety monitoring method and system based on the wireless passive three-signal sensor of the present invention;

[0068] Figure 2 Schematic diagram of the sensor signal decoupling process in the hydrogen storage station safety monitoring method based on wireless passive three-signal sensors and the system of the present invention;

[0069] Figure 3 Schematic diagram of the early warning decision tree model in the hydrogen storage station safety monitoring method and system based on wireless passive three-signal sensors of the present invention;

[0070] Figure 4 This is a flow chart of a hydrogen storage station safety monitoring method based on a wireless passive three-signal sensor and a threshold judgment mechanism in the system according to the present invention;

[0071] Figure 5 This is a flow chart of a hydrogen storage station safety monitoring method based on wireless passive three-signal sensors and a trend analysis mechanism in the system according to the present invention;

[0072] Figure 6 The figure is a flow chart of the hydrogen storage station safety monitoring method based on wireless passive three-signal sensors and the association rule reasoning mechanism in the system of the present invention. DETAILED DESCRIPTION

[0073] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0074] The hydrogen energy industry is currently experiencing rapid development. As a key link in the hydrogen energy supply chain, hydrogen storage stations urgently require real-time monitoring of multiple parameters for their safe operation. The monitoring system provided by this invention can accurately adapt to key areas of hydrogen storage stations, such as hydrogen storage tanks, pipelines, and compressor rooms, meeting safety monitoring needs in complex environments and possessing significant industrial application value. The system can be flexibly deployed according to the scale of the hydrogen storage station. The sensors and monitoring nodes have stable performance, can be repeatedly prepared and installed, and have low operation and maintenance costs. It provides reliable data support for hydrogen storage station safety management and promotes the safe and efficient development of the hydrogen energy industry.

[0075] The present invention discloses a hydrogen storage station safety monitoring method and system based on a wireless passive three-signal sensor. The system aims to address safety hazards, data asynchrony, and delayed warnings, among other issues, that plague existing hydrogen storage station monitoring technologies. This system achieves a technological breakthrough in hydrogen storage station safety monitoring through three levels of technological innovation: a power-free design for the wireless passive three-signal sensor, multi-parameter signal decoupling using frequency division multiplexing and a wavelet denoising algorithm, and the integration of a three-level warning decision tree with a distributed network architecture. This system fully meets the monitoring needs of hydrogen storage stations.

[0076] The system's core architecture comprises a distributed sensor network, a multi-band signal transceiver, an edge computing gateway, and a cloud-based early warning platform. The distributed sensor network collects data, the multi-band signal transceiver generates and transmits signals, the edge computing gateway processes and performs preliminary data analysis, and the cloud-based early warning platform leverages powerful computing and storage resources for in-depth data mining and risk assessment. These components work closely together to ensure the safe operation of the hydrogen storage station.

[0077] In practice, the distributed sensing network forms a three-dimensional monitoring network by deploying 16-24 wireless passive three-signal sensor nodes at key locations, such as the hydrogen tank's equatorial plane (three-point symmetry), pipeline elbows (stress concentration areas), and valve flanges. These sensor nodes precisely adapt to key areas of the hydrogen storage station, such as the hydrogen tank, pipeline, and compressor room, meeting safety monitoring requirements in complex environments. The multi-band signal transceiver utilizes a 100-300MHz wideband RF module to simultaneously excite the sensor's temperature (250MHz), strain (300MHz), and hydrogen (200MHz) detection units, enabling parallel acquisition of three-parameter signals. The edge computing gateway integrates a spectrum analysis chip (such as the AD9361) and an STM32H7 processor to perform signal decoupling, temperature compensation (based on a polynomial fitting algorithm), and preliminary risk assessment. The cloud-based early warning platform transmits data via 5G / fiber optics, building an AI analysis system that includes a digital twin model of the hydrogen tank, supporting historical data backtracking, trend prediction, and multi-level early warning (warning / alarm / emergency shutdown).

[0078] The system adopts a signal anti-crosstalk design, using a dual strategy of frequency division multiplexing (200 / 250 / 300MHz) and space division multiplexing (sensor spacing ≥1.5m), combined with a wavelet packet denoising algorithm, to suppress signal crosstalk to below -40dB. Through a customized algorithm for hydrogen storage scenarios, a temperature-strain coupling model (ΔT = 0.023 × Δε + 0.15) and a hydrogen concentration diffusion equation (based on Fick's second law) are established to achieve a leak source positioning accuracy of ≤2m. In addition, the low-power network protocol TDMA-CSMA hybrid networking protocol developed has a node sleep current of <10μA and a configurable wake-up period (10s-60min), meeting the long-term maintenance-free requirements of hydrogen storage stations.

[0079] The system deployment process includes monitoring point planning, sensor installation, and network construction. In the monitoring point planning, three sensors are evenly distributed around the hydrogen storage tank at a height of 1 / 3 of the tank bottom to monitor wall temperature and strain. Sensors are deployed on the piping system at elbows, reducers (large and small ends), tees, crosses, and 500mm upstream and downstream of valves. Monitoring nodes are set up on the top of the hydrogen buffer tank and at the pipeline interface in the compressor room. The sensor installation process uses high-temperature curing adhesive (temperature resistance ≥200°C) to adhere the sensor to the metal surface, ensuring direct contact between the sensitive layer and the measured medium. The sensor is waterproofed (IP68 rating) and wrapped with a 316L stainless steel protective mesh with a mechanical shock resistance of ≥50G. The network construction steps involve deploying 4-6 edge computing gateways to form a cellular coverage pattern, with a communication radius of ≥50m for each gateway. A local communication network is established using the LoRaWAN protocol, and the frequency of data upload to the cloud platform can be set (from 1 time / minute to 1 time / hour).

[0080] The system realizes multi-parameter synchronous acquisition. The reader transmits a broadband signal (100-300MHz) and activates the three IDTs of the sensor at the same time. In the hydrogen detection unit, the Pd / Al2O3 / Pd film reacts with H2 to generate PdH x , resulting in a surface acoustic wave frequency shift (Δf = -0.8 MHz / % LEL). The temperature sensing unit utilizes the thermoelastic effect of the AlScN film to linearly decrease the frequency with increasing temperature (α = -20 ppm / °C). The strain sensing unit amplifies mechanical deformation through the PI coating, while the LiNbO3 layer converts strain into frequency variation (GF = 35).

[0081] The data processing process is centered on sensor signal decoupling, and includes three major algorithm modules: frequency division multiplexing signal separation, temperature compensation correction, and wavelet denoising filtering. For the mixed RF signal output by the wireless passive three-signal sensor, high-precision calculation of temperature, strain, and hydrogen concentration parameters is achieved through frequency feature extraction, environmental factor calibration, and noise suppression. The processing cycle is 20ms, which meets the real-time monitoring needs of the hydrogen storage station. Specifically, the sensor signal decoupling process is as follows: sensor reflected mixed signal, FFT spectrum analysis, identification of 200 / 250 / 300MHz peaks (if not, return to the previous step, if yes, proceed to the next step), frequency-parameter preliminary solution, temperature compensation calculation, wavelet decomposition (3 layers), threshold processing detail components, signal reconstruction, parameter calibration output, and finally realizing local storage of edge gateway / cloud transmission.

[0082] The early warning decision adopts a three-layer progressive decision-making mechanism of threshold judgment-trend analysis-association rule reasoning. Based on multi-parameter fusion technology, the three types of monitoring data, temperature, strain, and hydrogen concentration, are correlated and analyzed with the operating status of the hydrogen storage station equipment to achieve a full-process safety assessment from abnormal warning to risk root cause location. The model decision cycle is 100ms, supporting real-time response to sudden safety incidents at the hydrogen storage station. Specifically, the early warning decision tree process is as follows: After the real-time data of the sensor is input, it first enters the first layer of threshold judgment to determine whether the threshold warning is triggered. If not triggered, it enters the second layer of trend analysis to determine whether a trend anomaly is found. If the trend analysis does not find an anomaly, it enters the third layer of association rule reasoning to determine whether there is a parameter coupling risk. If the association rule reasoning does not find a risk, it returns to normal monitoring. If the judgment at any level is yes, the corresponding level of warning, warning information fusion, and decision recommendations (maintenance / shutdown / observation) are output in sequence.

[0083] The safety early warning mechanism sets three levels of warning thresholds: warning (yellow), alarm (orange), and emergency (red). Warning (yellow): If any one of the following conditions is met: hydrogen concentration reaches 10% LEL, strain exceeds 50 με, or temperature exceeds 60°C, the system will record the abnormality and continuously monitor. Alarm (orange): If any two of the following conditions are met: hydrogen concentration reaches 25% LEL, strain exceeds 100 με, or temperature exceeds 80°C, the system will initiate coordinated protective measures. Emergency (red): If all of the following conditions are met: hydrogen concentration ≥ 40% LEL, strain rate > 10 με / s, and temperature exceeds 100°C, or if any single parameter changes suddenly, the system will be shut down and the firefighting system will be activated.

[0084] The trend analysis mechanism process includes exponential smoothing of parameter time series, calculation of change rate Δy / Δt, and prediction of the value y in the next ten minutes. pred (Prediction model: S t =α×y t +(1-α)×S t -1). Then determine whether Δy / Δt is ≥ threshold value. If yes, further determine y pred Is it ≥ 90% of the warning threshold? If so, trigger the warning in advance. If not, continue to monitor and update the forecast. If Δy / Δt < threshold, judge the abnormal index H = |(y t -μ) / α|≥3. If so, identify potential gradual risk (such as early hydrogen embrittlement). Otherwise, continue to monitor and update the prediction.

[0085] The association rule reasoning mechanism process involves processing the measured values ​​of three parameters. First, the temperature-strain theoretical value is calculated and the Δε measured-theoretical value is judged to be ≥15%. If so, it is determined that there is mechanical damage or leakage. Secondly, the hydrogen-temperature risk product is evaluated to determine C H2 ×(T-25) ≥ 1000. If so, warn of hydride decomposition risk; analyze the strain-hydrogen coupling risk again and determine whether ε ≥ 80με and C H2 Is it ≥15% LEL? If so, it will warn of hydrogen embrittlement crack expansion and eventually generate a multi-parameter correlation warning report.

[0086] The invention achieves a technological breakthrough in hydrogen storage station safety monitoring through three levels of technological innovation: First, the power-free design of the wireless passive three-signal sensor, combined with explosion-proof packaging (IP68+316L stainless steel protection) and broadband radio frequency communication (100-300MHz), eliminates the risk of electric sparks in hydrogen environments and can operate continuously for 1000 hours without abnormalities in a 99% LEL hydrogen environment; second, frequency division multiplexing (200 / 250 / 300MHz) and wavelet denoising algorithm (DB4 wavelet basis three-layer decomposition) are used to realize multi-parameter signal Decoupling is carried out, and a temperature compensation model (ΔT = 0.023 × Δε + 0.15) is used to improve detection accuracy, and the synchronous sampling delay is controlled within 50ms; finally, a three-level early warning decision tree (threshold judgment-trend analysis-association reasoning) and a distributed network architecture (sensor-edge gateway-cloud platform) are constructed, and combined with the digital twin model to achieve a leakage source positioning accuracy of ≤ 2m. The early warning response speed is 40 times faster than that of the traditional system, greatly shortening the hydrogen leakage accident warning time, and fully meeting the high-safety, high-precision, and low-cost monitoring needs of hydrogen storage stations.

[0087] Compared with existing technologies, this invention addresses safety hazards, data asynchrony, and delayed warnings in traditional monitoring through a multi-parameter collaborative algorithm (temperature-strain coupling model, hydrogen-temperature risk product criterion), a three-level early warning decision tree (threshold judgment-trend analysis-association reasoning), and distributed network design. It provides a full-scenario, high-security, and high-precision monitoring system solution. Actual testing has shown that the system operates stably in high-concentration hydrogen environments, effectively identifying and warning of potential risks. Its early warning response speed is 40 times faster than that of traditional systems, significantly shortening the warning time for hydrogen leak accidents and fully meeting the high-safety, high-precision, and low-cost monitoring requirements of hydrogen storage stations.

[0088] The system architecture combines wireless passive three-signal sensors with edge computing and digital twin technology for the first time to build a dedicated monitoring system for hydrogen storage stations, solving the problems of multi-parameter synchronization and real-time warning, which is not a simple sensor application expansion. The algorithm collaborative design uses the coupled analysis model of temperature-strain-hydrogen concentration (R 2 =0.987), breaking through the limitations of traditional single-parameter alarms. In practical applications at hydrogen storage stations, this invention effectively improves the safe operation of hydrogen storage stations and reduces accident risks through precise monitoring and intelligent early warning, providing strong technical support for the safe and efficient development of the hydrogen energy industry.

[0089] Based on surface acoustic wave technology, the present invention constructs a three-level architecture of "wireless passive three-signal sensor-edge computing gateway-cloud platform", deeply integrates the monitoring data of the three sensing units of temperature, strain and hydrogen concentration with the hydrogen storage station scenario, and adopts a strategy combining frequency division multiplexing (200 / 250 / 300MHz) and spatial partitioning deployment (differentiated layout of hydrogen storage tanks / pipelines / compressor rooms) to realize wireless passive synchronous monitoring and intelligent early warning of key parameters.

[0090] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A hydrogen storage station safety monitoring method based on a wireless passive three-signal sensor is characterized in that: Deploy wireless passive three-signal sensor nodes at hydrogen storage stations to form a three-dimensional monitoring network; The multi-band signal transceiver simultaneously stimulates the temperature, strain, and hydrogen detection units of the sensor to achieve parallel acquisition of three-parameter signals; Use edge computing gateways for signal decoupling, temperature compensation, and primary risk assessment; Conduct in-depth data mining and risk assessment through the cloud-based early warning platform combined with the digital twin model; Among them, frequency division multiplexing and wavelet denoising algorithm are used to achieve multi-parameter signal decoupling, and a three-level warning decision tree is combined to perform safety warning.

2. The hydrogen storage station safety monitoring method based on wireless passive three-signal sensor according to claim 1 is characterized in that: The sensor signal decoupling process includes: The sensor reflects mixed signals; FFT spectrum analysis; Identify the 200 / 250 / 300MHz peak value. If not, return to the previous step. If yes, proceed to the next step. Frequency-parameter preliminary solution; Temperature compensation calculation; wavelet decomposition; Thresholding detail components; Signal reconstruction; Parameter calibration output; Edge gateway local storage / cloud transmission.

3. The hydrogen storage station safety monitoring method based on wireless passive three-signal sensor according to claim 2 is characterized in that: The early warning decision tree process is as follows: Real-time sensor data; The first level of threshold judgment determines whether the threshold warning is triggered. If not, proceed to the next step; The second level is trend analysis, which determines whether an abnormal trend is found. If so, it proceeds to the next step. The third layer is association rule reasoning, which determines whether there is a parameter coupling risk. If so, the system returns to normal monitoring. If the judgment is yes at any level, the corresponding level of warning will be output in sequence, the warning information will be integrated, and decision recommendations will be generated.

4. The hydrogen storage station safety monitoring method based on wireless passive three-signal sensor according to claim 3 is characterized in that: The safety early warning mechanism sets three levels of warning thresholds: warning, alarm, and emergency; Warning: If any of the following conditions is met: hydrogen concentration reaches 10% LEL, strain exceeds 50 με, or temperature exceeds 60°C, the abnormality will be recorded and continuously monitored; Alarm: If two of the following conditions are met: hydrogen concentration reaches 25% LEL, strain exceeds 100 με, or temperature exceeds 80°C, the linkage protection measures will be activated; Emergency: If the hydrogen concentration is ≥40%LEL, the strain rate is greater than 10με / s and the temperature exceeds 100°C, or if a single parameter changes suddenly, the machine will be shut down immediately and the fire protection system will be activated.

5. The hydrogen storage station safety monitoring method based on wireless passive three-signal sensor according to claim 4 is characterized in that: Trend analysis mechanism process: parameter time series; Exponential smoothing; Calculate the rate of change Δy / Δt; Predict the value y in the next ten minutes pred , prediction model: S t =α×y t +(1-α)×S t -1; Determine whether Δy / Δt ≥ threshold? If so, judge y pre ≥ 90% of the warning threshold? If yes, trigger the warning in advance; otherwise, continue monitoring and update the forecast; Otherwise, determine whether the abnormal index H = |(yt-μ) / α|≥3? If so, identify the potential gradual risk; otherwise, continue monitoring and update the forecast.

6. The hydrogen storage station safety monitoring method based on wireless passive three-signal sensor according to claim 5 is characterized in that: The association rule reasoning mechanism process is as follows: Measured values ​​of three parameters; Calculate the temperature-strain theoretical value and determine whether the Δε measured-theoretical value is ≥15%. If so, it is determined that there is mechanical damage or leakage; Evaluate the hydrogen-temperature risk product and determine whether C H2 ×(T-25)≥1000, if so, a warning of hydride decomposition risk is issued; Analyze the strain-hydrogen coupling risk and determine whether ε≥80με and C H2 ≥15%LEL, if so, it will warn of hydrogen embrittlement crack expansion; Produce multi-parameter correlation early warning reports.

7. A system using the hydrogen storage station safety monitoring method based on wireless passive three-signal sensor according to claims 1-6, characterized in that: include: Distributed sensor network, deploying 16-24 wireless passive three-signal sensor nodes in hydrogen storage tanks, pipelines, and compressor rooms; Multi-band signal transceiver, using a 100-300MHz broadband RF module, simultaneously stimulates the sensor's temperature, strain, and hydrogen detection units; Edge computing gateway, integrating spectrum analysis chip and processor, completes signal decoupling, temperature compensation and primary risk assessment; The cloud-based early warning platform transmits data via 5G / fiber optics to build an AI analysis system that includes a digital twin model of the hydrogen storage tank, supporting historical data backtracking, trend prediction, and multi-level early warning.

8. The hydrogen storage station safety monitoring system based on wireless passive three-signal sensor according to claim 7 is characterized in that: Distributed sensor network deployment methods include: For hydrogen storage tanks, three sensors are evenly distributed around the tank at 1 / 3 of the height from the bottom of the tank; For piping systems, deploy sensors 500 mm upstream and downstream of elbows, reducers, tees, crosses, and valves. In the compressor room, monitoring nodes are set at the top of the hydrogen buffer tank and the pipeline interface.

9. The hydrogen storage station safety monitoring system based on wireless passive three-signal sensor according to claim 8 is characterized in that: The multi-band signal transceiver uses frequency division multiplexing 200 / 250 / 300MHz and space division multiplexing. The sensor spacing is ≥1.5m, and the signal crosstalk is suppressed to below -40dB.

Images