A gas monitoring method and system for hydrogen production equipment

Abstract

The invention relates to the field of gas monitoring technology, and specifically discloses a gas monitoring method for hydrogen production equipment, and the specific steps include: real-time collection of gas parameters and flow field parameters of a hydrogen pipeline area, construction of a dynamic flow field model, and identification of a low-speed flow dead zone; discrimination of turbulent boundary areas based on a double threshold method of pulsation intensity and Reynolds number; determination of gas energy concentration areas by clustering a three-dimensional turbulent kinetic energy field; screening of positive concentration areas that grow in the same direction in combination with time series trend analysis; and quantitative evaluation of the risk probability of gas energy concentration areas by fusing gas parameters, turbulent kinetic energy, and historical fault data using a random forest model; a gas monitoring system for hydrogen production equipment, including a dead zone identification module, a turbulence discrimination module, a concentration analysis module, a regional classification module, and a risk prediction module; the invention is conducive to solving the problems of blind areas and response delays in monitoring hydrogen production equipment, and significantly improves the safety of hydrogen production equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of gas monitoring, and in particular to a gas monitoring method and system for hydrogen production equipment. Background Art

[0002] In today's energy field, hydrogen production is a highly promising clean energy production method. Gas monitoring technology plays a key role in the hydrogen production process. However, gas monitoring of existing hydrogen production equipment still has many shortcomings.

[0003] In the prior art, when analyzing the gas energy concentration of hydrogen production equipment, the main focus is on the concentration distribution of the gas, while ignoring important factors such as turbulent kinetic energy. Turbulent kinetic energy reflects the degree of turbulence of the fluid and is of great significance for evaluating the mixing and diffusion of the gas; the lack of analysis of turbulent kinetic energy makes it impossible to accurately identify the location and range of the gas energy concentration area.

[0004] When analyzing the changing patterns of gas energy concentration areas, most existing technologies only consider static information at a single moment and do not conduct dynamic analysis from the perspective of time series; it is difficult to predict the development trend of gas energy concentration areas and it is impossible to discover potential risks in a timely manner. Summary of the invention

[0005] The object of the present invention is to provide a gas monitoring method and system for hydrogen production equipment to solve the above-mentioned background problems.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A gas monitoring method for hydrogen production equipment comprises the following steps:

[0008] Collect gas fluid parameters in the hydrogen pipeline area of ​​the hydrogen production equipment, build a dynamic flow field model of the hydrogen production equipment, and identify the flow dead zone of the hydrogen production equipment;

[0009] Perform turbulence analysis on the gas fluid in the dead zone of flow, and identify the gas turbulence corresponding to the boundary area between the dead zone of flow and the conventional flow channel area;

[0010] Energy concentration analysis is performed on the gas turbulence in the boundary area to obtain the turbulent kinetic energy in the boundary area, and cluster analysis is performed on the turbulent kinetic energy in the boundary area to determine the gas energy concentration area;

[0011] Obtain the positive concentration area of ​​the gas energy concentration area, and determine whether the positive area ratio and turbulent kinetic energy of the positive concentration area grow in the same direction. If they grow in the same direction, classify the gas energy concentration areas that meet the aggregation conditions to obtain the same type of concentration areas;

[0012] A risk prediction model is constructed to extract the gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentrated areas, construct a characteristic evaluation vector, input the characteristic evaluation vector into the risk prediction model, and determine the gas risk prediction value of each gas energy concentration area in the same concentrated area.

[0013] As a further technical solution of the present invention: the identification method of the flow dead zone is:

[0014] Real-time synchronous collection of gas and fluid parameters in the hydrogen pipeline area, and construction of a dynamic flow field model of the hydrogen production equipment based on gas parameters and flow field parameters;

[0015] Based on the dynamic flow field model of hydrogen production equipment, the flow dead zone in the hydrogen pipeline area of ​​the hydrogen production equipment is identified;

[0016] Different from the flow dead zone in the hydrogen production equipment area, other areas of the hydrogen pipeline area are marked as regular flow channel areas.

[0017] As a further technical solution of the present invention: the construction method of the dynamic flow field model of the hydrogen production equipment is:

[0018] Preprocessing and spatial and temporal alignment of gas parameters and flow field parameters;

[0019] Extract gas flow field characteristics;

[0020] Dynamic flow modeling.

[0021] As a further technical solution of the present invention: the gas turbulence is identified as follows:

[0022] Obtaining gas turbulence characteristic parameters in the boundary area between the flow dead zone and the conventional flow channel area during the monitoring period;

[0023] The gas fluid characteristic parameters include: pulsation intensity and Reynolds number within the monitoring period;

[0024] Based on the gas fluid characteristic parameters composed of pulsation intensity and Reynolds number within the monitoring period, the gas turbulence in the boundary area between the flow dead zone and the conventional flow channel area is identified by a dual-parameter threshold method.

[0025] As a further technical solution of the present invention: the pulsation intensity within the monitoring period is obtained in the following manner:

[0026] Obtain the average gas flow rate and instantaneous gas flow rate during the monitoring period;

[0027] Calculate the pulsating velocity component of the average gas flow velocity and the instantaneous gas flow velocity by using the difference formula;

[0028] The root mean square of the pulsation velocity component is calculated, and the root mean square of the pulsation velocity is ratioed to the average gas flow rate to obtain the pulsation intensity within the monitoring period.

[0029] As a further technical solution of the present invention: the gas energy concentration area is determined by:

[0030] By formula: Get the turbulent kinetic energy k in the boundary region, are the root mean square of the pulsating velocity components in the three dimensions of X, Y, and Z. ;

[0031] Based on the turbulent kinetic energy k in the boundary area, the turbulent kinetic energy field is constructed;

[0032] The turbulent kinetic energy field is gridded, the neighborhood radius and the minimum number of points of the clustering algorithm are set, and the density-based clustering algorithm is used to identify the gas energy concentration area.

[0033] As a further technical solution of the present invention: the turbulent kinetic energy field is constructed in the following manner:

[0034] The turbulent kinetic energy corresponding to the boundary area is mapped in space, and the turbulent kinetic energy field is constructed through the spatial interpolation algorithm.

[0035] As a further technical solution of the present invention: the neighborhood radius is determined as follows:

[0036] The neighborhood radius is characterized by the Taylor microscale within the boundary region, which is calculated by the dynamic viscosity and pulsating velocity components.

[0037] As a further technical solution of the present invention: the forward concentrated area is obtained by:

[0038] Obtain the turbulent kinetic energy in the gas energy concentration area and calculate the first-order difference of the turbulent kinetic energy at the initial and end times of the monitoring period;

[0039] If the first-order difference is positive, the gas energy concentration area is marked as a positive concentration area.

[0040] As a further technical solution of the present invention: the method for judging whether the positive area ratio and the turbulent kinetic energy of the positive concentration zone increase in the same direction is:

[0041] Perform ratio analysis on the area of ​​the positive concentrated area to obtain the positive area ratio, and construct the positive area sequence based on the positive area ratio;

[0042] Obtain the turbulent kinetic energy of the positive concentration area in multiple monitoring periods and construct a kinetic energy sequence;

[0043] The positive area ratio and turbulent kinetic energy series of the positive concentration zone were analyzed in the same direction, and the slope of the fitting equation of the positive area series and turbulent kinetic energy was obtained;

[0044] If the slopes of the fitting equations of the forward area series and the turbulent kinetic energy are both greater than 0, it is considered that the forward area ratio and the turbulent kinetic energy in the forward concentration zone increase in the same direction.

[0045] As a further technical solution of the present invention: the slope of the fitting equation of the forward area series and turbulent kinetic energy is obtained as follows:

[0046] By performing linear regression analysis on the forward area series and turbulent kinetic energy series, the slopes of the fitting equations of the forward area series and turbulent kinetic energy were obtained.

[0047] As a further technical solution of the present invention: the method for obtaining the same type of concentrated areas is:

[0048] The gas energy concentration areas in all boundary areas are screened, and the gas energy concentration areas that meet the aggregation conditions are classified to obtain similar concentration areas.

[0049] As a further technical solution of the present invention: the polymerization conditions are:

[0050] Polymerization conditions 1. The gas energy concentration zone is a positive concentration zone;

[0051] Aggregation condition 2: The positive area ratio and turbulent kinetic energy of the positive concentration zone increase in the same direction.

[0052] As a further technical solution of the present invention: the gas risk prediction value is obtained in the following manner:

[0053] Obtain gas fluid parameters, turbulent kinetic energy and fault distribution rate in similar concentrated areas and construct characteristic evaluation vectors;

[0054] A risk prediction model is constructed through the random forest algorithm, and the characteristic evaluation vectors of similar concentrated areas are input into the risk prediction model;

[0055] The gas risk prediction value of each similar concentration area is obtained through the risk prediction model.

[0056] A gas monitoring system for hydrogen production equipment: comprising the following modules:

[0057] Dead zone identification module: used to collect gas fluid parameters in the hydrogen pipeline area of ​​the hydrogen production equipment, build a dynamic flow field model of the hydrogen production equipment, and identify the flow dead zone of the hydrogen production equipment;

[0058] Turbulence identification module: Based on the dead zone of hydrogen production equipment, the turbulence analysis of the gas fluid in the dead zone is performed to identify the gas turbulence corresponding to the boundary area between the dead zone and the conventional flow channel area;

[0059] Concentration analysis module: conducts energy concentration analysis on the gas turbulence in the boundary area to obtain the turbulent kinetic energy in the boundary area, performs cluster analysis on the turbulent kinetic energy in the boundary area, and determines the gas energy concentration area;

[0060] Regional classification module: obtain the positive concentration area of ​​the gas energy concentration area, determine whether the positive area ratio and turbulent kinetic energy of the positive concentration area grow in the same direction, and if they grow in the same direction, classify the gas energy concentration areas that meet the aggregation conditions to obtain the same type of concentration areas;

[0061] Risk prediction module: extract gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentrated areas, construct characteristic evaluation vectors, build a risk prediction model based on the characteristic evaluation vectors, and determine the gas risk prediction value of each gas energy concentration area in the same concentrated area.

[0062] Beneficial effects of the present invention:

[0063] (1) By collecting gas and fluid parameter data in real time, data preprocessing and time-space alignment are beneficial to improving data accuracy. The improved k-ωSST turbulence model is adopted and the adaptive time step is set based on the CFL condition to construct a dynamic flow field model of the hydrogen production equipment, which is beneficial to identifying the flow dead zone and tracking the flow dead zone in real time under various working conditions of the hydrogen production equipment, thus making up for the monitoring blind spots of traditional monitoring methods and providing a reliable basis for subsequent analysis.

[0064] (2) Based on the flow dead zone, the gas turbulence at the boundary between the flow dead zone and the conventional flow channel area can be identified by obtaining the two key characteristic parameters, pulsation intensity and Reynolds number. The dual-parameter threshold method is more reliable than the single-parameter judgment, and the dynamic viscosity is calculated by the Sutherland formula, which provides important preliminary support for the analysis of the gas energy concentration area.

[0065] (3) The turbulent kinetic energy is calculated by decomposing the pulsating velocity components, and the turbulent kinetic energy field is constructed by combining the spatial interpolation method. The gas energy concentration area is determined by the clustering algorithm, and the neighborhood radius and the minimum number of points are set to improve the clustering accuracy, which is conducive to the division of the gas energy concentration area. Through the differential calculation of the turbulent kinetic energy of the gas energy concentration area, the positive area ratio and the same direction trend analysis of the turbulent kinetic energy can be used to screen out the positive concentration area and classify it into the same type of concentration area, and the change law of the gas energy concentration area can be deeply analyzed from the two dimensions of time and space.

[0066] (4) The gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentrated areas are extracted to construct a characteristic evaluation vector, and a risk prediction model is constructed using the random forest algorithm. This method can integrate multiple aspects of information and determine the gas risk prediction value more scientifically than traditional methods, providing a quantitative basis for the risk assessment of hydrogen production equipment, assisting operators to promptly discover potential risks, take measures in advance, and reduce the probability of accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The present invention will be further described below in conjunction with the accompanying drawings.

[0068] Figure 1It is a flow chart of a gas monitoring method for hydrogen production equipment of the present invention;

[0069] Figure 2 It is a flow chart of the construction method of the dynamic flow field model of the hydrogen production equipment in the present invention;

[0070] Figure 3 It is a module diagram of a gas monitoring system for hydrogen production equipment of the present invention. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0072] Under dynamic conditions such as start-up and shutdown, load fluctuations, etc., the flow field parameters (pressure, temperature, flow rate) of hydrogen production equipment change dramatically, resulting in dynamic evolution of the range and position of the flow dead zone. Traditional monitoring methods cannot track the changes in the dead zone in real time, and there are monitoring blind spots and response delays. It is necessary to establish a closed-loop monitoring system of dynamic flow field perception-intelligent compensation-coordinated control.

[0073] Embodiment 1

[0074] See also Figure 1 As shown, the present invention is a gas monitoring method for hydrogen production equipment, comprising the following steps:

[0075] Step 1: collect gas fluid parameters in the hydrogen pipeline area of ​​the hydrogen production equipment in real time, build a dynamic flow field model of the hydrogen production equipment, and identify the flow dead zone of the hydrogen production equipment;

[0076] In some embodiments, a micro gas monitoring sensor array is deployed in the hydrogen pipeline area of ​​the hydrogen production equipment to synchronously collect gas fluid parameters in real time;

[0077] Among them, the gas fluid parameters include: gas parameters, flow field parameters; the hydrogen pipeline area includes: the pipeline area composed of the pipeline elbows and valve connections corresponding to the electrolyzer electrode area, the reactor area, and the gas separation tower area;

[0078] It should be noted that the gas parameters include: hydrogen concentration, oxygen concentration; the flow field parameters include: gas flow rate, gas pressure, fluid density, temperature;

[0079] The gas in the gas flow rate and gas pressure of the flow field parameters includes hydrogen and oxygen;

[0080] Based on gas parameters and flow field parameters, a dynamic flow field model of hydrogen production equipment is constructed;

[0081] like Figure 2 As shown, the construction method of the dynamic flow field model of the hydrogen production equipment is:

[0082] S1, preprocessing and time-space alignment of gas parameters and flow field parameters;

[0083] Performing data preprocessing on gas parameters and flow field parameters, and constructing a gas flow field data set including the gas parameters and flow field parameters;

[0084] Among them, data processing includes: removing abnormal values ​​of gas parameters and flow field parameters, identifying missing values ​​of gas parameters and flow field parameters and filling the missing data, and aligning gas parameters and flow field parameters in time and space;

[0085] It should be noted that the IQR algorithm is used to identify the outliers of gas parameters and flow field parameters, and the outlier data points are deleted; the Kriging interpolation method is used to restore the missing data caused by sensor failure; by obtaining the timestamps corresponding to the gas parameters and flow field parameters, the gas parameters and flow field parameters are time-aligned, and the spatial coordinates of the gas monitoring sensor are converted into the device coordinate system for spatial alignment;

[0086] S2, extracting gas flow field characteristics;

[0087] Extract gas velocity gradient, hydrogen concentration gradient, and oxygen concentration gradient based on the gas flow field data set;

[0088] The gas velocity gradient, hydrogen concentration gradient and oxygen concentration gradient are used as gas flow field characteristics;

[0089] S3, dynamic flow field modeling;

[0090] The improved k-ωSST turbulence model is adopted, the adaptive time step is set, and the dynamic flow field model of hydrogen production equipment is constructed;

[0091] Among them, the adaptive time step algorithm based on the CFL (Courant-Friedrichs-Lewy) condition is used to set the adaptive time step;

[0092] Combined with the gas flow rate obtained in S1, the CFL condition specifies the time step With space step and the gas flow rate V, namely ;

[0093] It should be further explained that during the calculation process, the time step is adjusted in real time according to the velocity distribution and spatial grid size of the current flow field to ensure that the CFL number always meets the conditions. For example, when the flow field velocity is large, the time step is automatically reduced; when the flow field velocity is small, the time step is appropriately increased;

[0094] It should be noted that the traditional k-ωSST (Shear-Stress Transport) turbulence model has certain limitations when dealing with complex flow fields;

[0095] The improved k-ωSST turbulence model can better simulate the dynamic flow field in the hydrogen production equipment by improving the calculation method of the turbulent viscosity coefficient and introducing adaptive turbulence model parameters.

[0096] The calculation formula of the turbulent viscosity coefficient is adjusted by introducing the gas velocity gradient, hydrogen concentration gradient, oxygen concentration gradient, pressure and temperature obtained in S2. For example, the correction terms related to the pressure gradient and concentration gradient are introduced to make the turbulent viscosity coefficient more accurately reflect the actual situation of the flow field;

[0097] According to the local characteristics of the flow field, such as gas velocity gradient, pressure gradient, etc., the parameters in the turbulence model are automatically adjusted, such as the coefficients of the generation and dissipation terms of turbulent kinetic energy and turbulent dissipation rate; by automatically adjusting the parameters in the turbulence model, the turbulence model can better adapt to different flow field conditions;

[0098] In dynamic flow field modeling, the choice of time step directly affects the accuracy and efficiency of calculation. If the time step is too large, the calculation results will be unstable; if the time step is too small, the calculation amount and calculation time will increase.

[0099] Based on the dynamic flow field model of hydrogen production equipment, the flow dead zone in the hydrogen pipeline area of ​​the hydrogen production equipment is identified;

[0100] Different from the flow dead zone in the hydrogen production equipment area, other areas in the hydrogen pipeline area are marked as regular flow channel areas;

[0101] The dead zone refers to the area with low velocity and weak material exchange in the flow field. By analyzing the velocity distribution of the flow field calculated by the improved k-ωSST turbulence model, the area with a velocity lower than the preset velocity threshold is defined as the dead zone.

[0102] Among them, the velocity data is calculated based on the feature extraction in S2 and the improved model in S3;

[0103] Step 2: Based on the flow dead zone of the hydrogen production equipment, a turbulence analysis is performed on the gas fluid in the flow dead zone to identify the gas turbulence corresponding to the boundary area between the flow dead zone and the conventional flow channel area;

[0104] The method for identifying the gas turbulence in the flow dead zone is:

[0105] Obtaining gas turbulence characteristic parameters in the boundary area between the flow dead zone and the conventional flow channel area during the monitoring period;

[0106] The gas fluid characteristic parameters include: pulsation intensity and Reynolds number within the monitoring period;

[0107] The pulsation intensity is obtained as follows:

[0108] Get the average gas flow rate during the monitoring period , and the instantaneous gas flow rate ;

[0109] By formula: Get the pulsating velocity component ;

[0110] By formula: Get the RMS value of the pulsating velocity component , where T is the length of the monitoring period;

[0111] The rms of the pulsating velocity The average gas velocity Perform ratio processing to obtain the pulsation intensity within the monitoring period;

[0112] Among them, the Reynolds number within the monitoring period is obtained as follows:

[0113] By formula: Get the Reynolds number Re during the monitoring period, where They represent fluid density, average gas velocity during the monitoring period, pipeline diameter, and dynamic viscosity respectively;

[0114] It should be noted that the Reynolds number is a dimensionless number used to describe the flow state of a fluid. It is calculated based on the density of the fluid, the average flow velocity, the diameter of the pipe, and the dynamic viscosity of the fluid. When the Reynolds number is greater than a certain critical value, the flow state of the fluid usually changes from laminar flow to turbulent flow.

[0115] Among them, dynamic viscosity (μ) is the ability of the fluid to resist shear deformation, and the dynamic viscosity is calculated by Sutherland's formula;

[0116] Based on the gas fluid characteristic parameters composed of pulsation intensity and Reynolds number during the monitoring period, the gas turbulence in the boundary area between the flow dead zone and the conventional flow channel area is identified by a dual-parameter threshold method;

[0117] Specifically, the pulsation intensity within the monitoring period is compared with a preset pulsation intensity threshold, and the Reynolds number within the monitoring period is compared with a preset Reynolds number threshold;

[0118] If the pulsation intensity within the monitoring period is higher than a preset pulsation intensity threshold, and the Reynolds number within the monitoring period is higher than a preset Reynolds number threshold, it is considered that the gas in the boundary area between the flow dead zone and the conventional flow channel area is turbulent.

[0119] It should be noted that the role of identifying gas turbulence is:

[0120] Function 1: Define the boundary area between the dead zone and the conventional flow channel. By monitoring the pulsation intensity (reflecting the degree of velocity fluctuation) and the Reynolds number (the critical value for judging the flow state from laminar to turbulent), the dual-parameter threshold method is used to identify turbulence, which can accurately divide the boundary range between the dead zone and the conventional flow channel. This solves the misjudgment problem caused by traditional single parameters (such as relying only on the Reynolds number), and provides precise spatial positioning for subsequent focus on high-risk boundary areas;

[0121] Function 2: It provides a key premise for the analysis of gas energy concentration areas. Turbulence is an important sign of energy concentration, and turbulence identification in boundary areas is the basis for the subsequent calculation of turbulent kinetic energy. By confirming the boundary areas where turbulence exists and combining the pulsating velocity components decomposed by the three-dimensional sensor, the turbulent kinetic energy field can be constructed and cluster analysis can be performed to determine the location and range of the gas energy concentration area. If turbulence is not accurately identified, it may lead to missed or misjudgment of the energy concentration area.

[0122] Function 3: Improve the analysis accuracy of the dynamic characteristics of the flow field. The existence of turbulence indicates that there is a strong exchange of matter and energy dissipation in the flow field. The dynamic viscosity is calculated by the Sutherland formula and combined with the Reynolds number, which can reflect the flow field changes under different working conditions (such as equipment start-up and shutdown, load fluctuations) in real time. This provides more accurate boundary conditions for the improved k-ωSST turbulence model, enhances the adaptability of flow field modeling to complex working conditions, and makes up for the shortcomings of traditional models in dynamic flow field simulation.

[0123] The technical solution of this embodiment is: to collect gas fluid parameters in the hydrogen pipeline area of ​​the hydrogen production equipment in real time, to construct a dynamic flow field model of the hydrogen production equipment, and to identify the flow dead zone of the hydrogen production equipment; based on the flow dead zone of the hydrogen production equipment, to perform turbulence analysis on the gas fluid in the flow dead zone, to identify the gas turbulence corresponding to the boundary area between the flow dead zone and the conventional flow channel area, and to provide a key prerequisite for the analysis of the gas energy concentration area.

[0124] Embodiment 2

[0125] like Figure 1 As shown, a gas monitoring method for hydrogen production equipment also includes the following steps:

[0126] Step 3: Based on the gas turbulence in the boundary area, an energy concentration analysis is performed on the gas turbulence in the boundary area to obtain the turbulent kinetic energy in the boundary area, and a cluster analysis is performed on the turbulent kinetic energy in the boundary area to determine the gas energy concentration area;

[0127] The method for performing energy concentration analysis on the gas turbulence in the plurality of boundary regions is as follows:

[0128] By formula: Get the turbulent kinetic energy k in the boundary region, are the root mean square of the pulsating velocity components in the three dimensions of X, Y, and Z. ;

[0129] It should be noted that the pulsating velocity components are decomposed by the three-dimensional sensor to obtain the pulsating velocity components in three dimensions: X, Y, and Z;

[0130] Based on the turbulent kinetic energy k of the boundary area, the turbulent kinetic energy corresponding to the boundary area is mapped in space, and the turbulent kinetic energy field is constructed through the spatial interpolation algorithm;

[0131] It should be noted that the pulsating velocity components of each point in the boundary area are decomposed by using a three-dimensional sensor, and the turbulent kinetic energy of each discrete point is calculated according to the formula. The turbulent kinetic energy data of the discrete points are mapped to the spatial coordinate system. At this time, these data are scattered and discontinuous. The Kriging interpolation method is selected. Based on the existing discrete turbulent kinetic energy data points, by analyzing their spatial distribution laws, correlations and trends, the turbulent kinetic energy values ​​of other unmeasured points in the space are estimated and supplemented, thereby constructing a continuous and complete turbulent kinetic energy field that can reflect the distribution of turbulent kinetic energy in the boundary area.

[0132] The turbulent kinetic energy field is gridded and a density-based clustering algorithm is used to identify gas energy concentration areas;

[0133] It should be noted that the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm is used to identify the gas energy concentration area. The DBSCAN algorithm can divide the data into different clusters according to the density of the data points, and can identify the noise points. Each grid point in the interpolated turbulent kinetic energy field is used as a data point, and cluster analysis is performed according to its turbulent kinetic energy value and spatial coordinates. After clustering, each cluster represents a gas energy concentration area, while the noise points represent isolated areas that do not constitute obvious gas energy concentration.

[0134] Among them, when using the DBSCAN algorithm, two key parameters need to be set: the neighborhood radius ε and the minimum number of points MinPts; the neighborhood radius defines the neighborhood range of a point, and the minimum number of points MinPts indicates the minimum number of points that must be included in the neighborhood of a point to consider the point as a core point;

[0135] Among them, the neighborhood radius ε is characterized by the Taylor microscale in the calculated boundary area, and the Taylor microscale is calculated by the ratio of the dynamic viscosity and the pulsating velocity component;

[0136] Preferably, the minimum number of points MinPts is 4. The selection of the minimum number of points MinPts is related to the dimension D of the data. Since the space is composed of three dimensions, X, Y, and Z, the minimum number of points MinPts is D+1=4.

[0137] It should be noted that the role of determining the gas energy concentration area is:

[0138] Function 1: Locate high-energy turbulence areas, identify potential risk hotspots, calculate turbulent kinetic energy by decomposing pulsating velocity components through three-dimensional sensors, construct turbulent kinetic energy fields by combining spatial interpolation methods, and use the DBSCAN clustering algorithm to identify high-density areas (i.e. gas energy concentration areas). These areas are usually locations with strong turbulence and concentrated energy at the boundaries between flow dead zones and conventional flow channels. The material exchange is intense and the flow field is unstable. They are risk hotspots that are prone to cause failures (such as local corrosion and gas leakage) during the operation of hydrogen production equipment. By accurately defining such areas, it is possible to avoid the omission or misjudgment of "isolated high-energy points" by traditional monitoring methods, and to achieve precise spatial positioning of risk areas;

[0139] Function 2: Provide core target areas for dynamic trend analysis. Gas energy concentration areas are the basis for subsequent positive concentration area screening and time series trend analysis. By calculating the first-order difference of turbulent kinetic energy during the monitoring period, marking the positive concentration area (area with increased turbulent kinetic energy), and combining the positive area ratio (area ratio of the concentration area) to construct a sequence, it is determined whether the two are growing in the same direction. This process relies on the accurate division of gas energy concentration areas, so as to identify dangerous areas where energy continues to accumulate and expand over time, and provide a quantitative basis for the dynamic evaluation of the equipment operation status.

[0140] Step 4: perform differential calculation on the turbulent kinetic energy of the gas energy concentration zone in the boundary area to obtain the positive concentration zone, perform a same-direction trend analysis on the positive area ratio and turbulent kinetic energy of the positive concentration zone to determine whether the positive area ratio and turbulent kinetic energy of the positive concentration zone grow in the same direction. If they grow in the same direction, the gas energy concentration zones that meet the aggregation conditions are classified to obtain the same type of concentration zones;

[0141] Obtain the turbulent kinetic energy in the gas energy concentration area and calculate the first-order difference of the turbulent kinetic energy at the initial and end times of the monitoring period;

[0142] If the first-order difference is positive, the gas energy concentration area is marked as a positive concentration area, otherwise it is marked as a negative concentration area;

[0143] Obtaining the area of ​​the gas energy concentration zone corresponding to the positive concentration zone, calculating the ratio of the area of ​​the gas energy concentration zone corresponding to the positive concentration zone to the area of ​​the boundary zone, and obtaining the positive area ratio;

[0144] The gas energy concentration area corresponding to the positive area ratio is marked as the positive concentration area;

[0145] Obtain the positive area ratio of the positive concentration area in multiple monitoring cycles and construct a positive area sequence;

[0146] Obtain the turbulent kinetic energy of the positive concentration area in multiple monitoring periods and construct a kinetic energy sequence;

[0147] Conduct a trend analysis on the positive area ratio and turbulent kinetic energy in the positive concentration zone to determine whether the positive area ratio and turbulent kinetic energy in the positive concentration zone increase in the same direction;

[0148] The method for determining whether the positive area ratio and the turbulent kinetic energy of the positive concentration zone increase in the same direction is:

[0149] By performing linear regression analysis on the forward area series and turbulent kinetic energy series, the slopes of the fitting equations of the forward area series and turbulent kinetic energy are obtained;

[0150] If the slopes of the fitting equations of the forward area series and turbulent kinetic energy are both greater than 0, it is considered that the forward area ratio and turbulent kinetic energy in the forward concentration zone increase in the same direction;

[0151] Screen the gas energy concentration areas in all boundary areas, classify the gas energy concentration areas that meet the aggregation conditions, and obtain similar concentration areas;

[0152] Among them, the polymerization condition is 1. The gas energy concentration area is a positive concentration area;

[0153] Aggregation condition 2: the positive area ratio and turbulent kinetic energy of the positive concentration zone grow in the same direction;

[0154] It should be noted that the role of judging whether the positive area ratio and turbulent kinetic energy of the positive concentration zone grow in the same direction is:

[0155] Function 1: Identify the dynamic risk evolution trend and warn of safety hazards. The positive area ratio growth indicates that the spatial range of the gas energy concentration area is expanding, which may mean that the turbulent area at the boundary of the flow dead zone and the conventional flow channel is expanding, and the spatial range of material exchange and energy accumulation continues to increase.

[0156] Turbulent kinetic energy growth: reflects the intensification of turbulence intensity in the area and the increase in the concentration of kinetic energy of the fluid, which may be accompanied by higher pressure fluctuations, temperature gradients or uneven concentrations, increasing the risk of local corrosion and fatigue failure of equipment;

[0157] Growth in the same direction: The simultaneous growth of the two indicates that energy accumulation and spatial expansion form a positive feedback, which is a significant signal of increased risk. For example, at the pipe elbow of hydrogen production equipment, if the growth in the same direction continues, it may indicate that the area is about to enter an unstable state and requires immediate attention;

[0158] Function 2: Screen key risk areas and exclude invalid fluctuations: Through dual-condition screening (positive concentration area and same-direction growth), short-term noise or isolated energy fluctuations can be filtered out; locate core risk areas: only areas that meet both "energy increase" and "scale expansion" are classified as similar concentration areas, ensuring that subsequent risk assessments focus on high-risk areas with sustained growth potential, thereby improving monitoring efficiency;

[0159] Function 3: Provide a dynamic basis for risk quantification in the time dimension and construct time series characteristics: Through linear regression analysis of the slope, the same-direction growth is converted into a quantifiable trend indicator, avoiding the limitations of traditional methods that only rely on data at a single moment.

[0160] Step 5: construct a risk prediction model, extract gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentrated areas, construct a characteristic evaluation vector, input the characteristic evaluation vector into the risk prediction model, and determine the gas risk prediction value of each gas energy concentration area in the same concentrated area;

[0161] The risk prediction model is constructed as follows:

[0162] Obtain gas fluid parameters, turbulent kinetic energy and fault distribution rate in similar concentrated areas and construct characteristic evaluation vectors;

[0163] Exemplarily, the feature evaluation vector is ;

[0164] Among them, P i represents the gas fluid parameters of the same type of concentrated area, n is the number of parameters, k is the turbulent kinetic energy, and f is the fault distribution rate;

[0165] It should be noted that the fault distribution rate is obtained based on the equipment historical operation data of the hydrogen production equipment;

[0166] For example, P 1 Indicates the hydrogen concentration in the gas parameters, P 2 Indicates oxygen concentration;

[0167] A risk prediction model is constructed through the random forest algorithm, and the characteristic evaluation vectors of similar concentrated areas are input into the risk prediction model;

[0168] By formula: Build a risk prediction model to obtain the gas risk prediction value RiskScore for each similar concentration area;

[0169] Where K is the number of trees in the random forest algorithm, It represents the risk prediction value of the j-th pair of feature vectors P, where j is the number of the tree in the random forest algorithm.

[0170] The technical solution of this embodiment is: based on the gas turbulence in the boundary area, energy concentration analysis is performed on the gas turbulence in the boundary area to obtain the turbulent kinetic energy of the boundary area, and the turbulent kinetic energy of the boundary area is clustered to determine the gas energy concentration area; the turbulent kinetic energy of the gas energy concentration area in the boundary area is differentially calculated to obtain the positive concentration area, and the positive area ratio and turbulent kinetic energy of the positive concentration area are analyzed in the same direction to determine whether the positive area ratio and turbulent kinetic energy of the positive concentration area increase in the same direction. If they increase in the same direction, the gas energy concentration areas that meet the aggregation conditions are classified to obtain similar concentration areas; the gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentration areas are extracted to construct a characteristic evaluation vector, and a risk prediction model is constructed based on the characteristic evaluation vector to determine the gas risk prediction value of each gas energy concentration area in the similar concentration area; it can integrate multiple aspects of information and determine the gas risk prediction value more scientifically than traditional methods, provide a quantitative basis for risk assessment of hydrogen production equipment, and assist operators to promptly discover potential risks.

[0171] Embodiment 3

[0172] like Figure 3 As shown, a gas monitoring system for hydrogen production equipment includes the following modules:

[0173] Dead zone identification module: used to collect gas fluid parameters in the hydrogen pipeline area of ​​the hydrogen production equipment in real time, build a dynamic flow field model of the hydrogen production equipment, and identify the flow dead zone of the hydrogen production equipment;

[0174] By deploying a micro gas monitoring sensor array in the hydrogen pipeline area of ​​the hydrogen production equipment, it is used to synchronously collect gas fluid parameters in real time;

[0175] Based on gas parameters and flow field parameters, a dynamic flow field model of hydrogen production equipment is constructed;

[0176] The construction method of the dynamic flow field model of the hydrogen production equipment is as follows:

[0177] S1, parameter preprocessing and spatiotemporal alignment;

[0178] S2, gas flow field feature extraction;

[0179] S3, dynamic flow field modeling;

[0180] The improved k-ωSST turbulence model is adopted, the adaptive time step is set, and the dynamic flow field model of hydrogen production equipment is constructed;

[0181] Based on the dynamic flow field model of hydrogen production equipment, the flow dead zone in the hydrogen pipeline area of ​​the hydrogen production equipment is identified;

[0182] Turbulence identification module: Based on the dead zone of hydrogen production equipment, it is used to perform turbulence analysis on the gas fluid in the dead zone and identify the gas turbulence corresponding to the boundary area between the dead zone and the conventional flow channel area;

[0183] The method for identifying the gas turbulence in the flow dead zone is:

[0184] Obtaining gas turbulence characteristic parameters in the boundary area between the flow dead zone and the conventional flow channel area during the monitoring period;

[0185] The gas fluid characteristic parameters include: pulsation intensity and Reynolds number within the monitoring period;

[0186] The pulsation intensity is obtained as follows:

[0187] Get the average gas flow rate during the monitoring period , and the instantaneous gas flow rate ;

[0188] By formula: Get the pulsating velocity component ;

[0189] By formula: Get the RMS value of the pulsating velocity component , where T is the length of the monitoring period;

[0190] The rms of the pulsating velocity The average gas velocity Perform ratio processing to obtain the pulsation intensity within the monitoring period;

[0191] Among them, the Reynolds number within the monitoring period is obtained as follows:

[0192] By formula: Get the Reynolds number Re during the monitoring period, where They represent fluid density, average gas velocity during the monitoring period, pipeline diameter, and dynamic viscosity respectively;

[0193] Among them, dynamic viscosity (μ) is the ability of the fluid to resist shear deformation, and the dynamic viscosity is calculated by Sutherland's formula;

[0194] Based on the gas fluid characteristic parameters composed of pulsation intensity and Reynolds number during the monitoring period, the gas turbulence in the boundary area between the flow dead zone and the conventional flow channel area is identified by a dual-parameter threshold method;

[0195] Specifically, the pulsation intensity within the monitoring period is compared with a preset pulsation intensity threshold, and the Reynolds number within the monitoring period is compared with a preset Reynolds number threshold;

[0196] If the pulsation intensity during the monitoring period is higher than the preset pulsation intensity threshold, and the Reynolds number during the monitoring period is higher than the preset Reynolds number threshold, it is considered that the gas turbulence in the boundary area between the flow dead zone and the conventional flow channel area is;

[0197] Concentration analysis module: Based on the gas turbulence in the boundary area, it is used to perform energy concentration analysis on the gas turbulence in the boundary area, obtain the turbulent kinetic energy in the boundary area, perform cluster analysis on the turbulent kinetic energy in the boundary area, and determine the gas energy concentration area;

[0198] The method for performing energy concentration analysis on the gas turbulence in the plurality of boundary regions is as follows:

[0199] By formula: Get the turbulent kinetic energy k in the boundary region, are the root mean square of the pulsating velocity components in the three dimensions of X, Y, and Z. ;

[0200] Based on the turbulent kinetic energy k of the boundary area, the turbulent kinetic energy corresponding to the boundary area is mapped in space, and the turbulent kinetic energy field is constructed through the spatial interpolation algorithm;

[0201] The turbulent kinetic energy field is gridded and a density-based clustering algorithm is used to identify gas energy concentration areas;

[0202] Regional classification module: used to perform differential calculation on the turbulent kinetic energy of the gas energy concentration zone in the boundary area to obtain the positive concentration zone, perform the same-direction trend analysis on the positive area ratio and turbulent kinetic energy of the positive concentration zone, and determine whether the positive area ratio and turbulent kinetic energy of the positive concentration zone increase in the same direction. If they increase in the same direction, the gas energy concentration zones that meet the aggregation conditions will be classified to obtain the same type of concentration zones;

[0203] Obtain the turbulent kinetic energy in the gas energy concentration area and calculate the first-order difference of the turbulent kinetic energy at the initial and end times of the monitoring period;

[0204] If the first-order difference is positive, the gas energy concentration area is marked as a positive concentration area, otherwise it is marked as a negative concentration area;

[0205] Obtaining the area of ​​the gas energy concentration zone corresponding to the positive concentration zone, calculating the ratio of the area of ​​the gas energy concentration zone corresponding to the positive concentration zone to the area of ​​the boundary zone, and obtaining the positive area ratio;

[0206] The gas energy concentration area corresponding to the positive area ratio is marked as the positive concentration area;

[0207] Obtain the positive area ratio of the positive concentration area in multiple monitoring cycles and construct a positive area sequence;

[0208] Obtain the turbulent kinetic energy of the positive concentration area in multiple monitoring periods and construct a kinetic energy sequence;

[0209] Conduct a trend analysis on the positive area ratio and turbulent kinetic energy in the positive concentration zone to determine whether the positive area ratio and turbulent kinetic energy in the positive concentration zone increase in the same direction;

[0210] By performing linear regression analysis on the forward area series and turbulent kinetic energy, the slope of the fitting equation of the forward area series and turbulent kinetic energy is identified;

[0211] If the slopes of the fitting equations of the forward area series and turbulent kinetic energy are both greater than 0, it is considered that the forward area ratio and turbulent kinetic energy in the forward concentration zone increase in the same direction;

[0212] Screen the gas energy concentration areas in all boundary areas, classify the gas energy concentration areas that meet the aggregation conditions, and obtain similar concentration areas;

[0213] Among them, the polymerization condition is 1. The gas energy concentration area is a positive concentration area;

[0214] Aggregation condition 2: the positive area ratio and turbulent kinetic energy of the positive concentration zone grow in the same direction;

[0215] Risk prediction module: used to build a risk prediction model, extract gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentrated areas, build a characteristic evaluation vector, input the characteristic evaluation vector into the risk prediction model, and determine the gas risk prediction value of each gas energy concentration area in the same concentrated area;

[0216] The risk prediction model is constructed as follows:

[0217] Obtain gas fluid parameters, turbulent kinetic energy and fault distribution rate in similar concentrated areas and construct characteristic evaluation vectors;

[0218] Exemplarily, the feature evaluation vector is ;

[0219] Among them, P i represents the gas fluid parameters of the same type of concentrated area, n is the number of parameters, k is the turbulent kinetic energy, and f is the fault distribution rate;

[0220] For example, P 1 Indicates the hydrogen concentration in the gas parameters, P 2 Indicates oxygen concentration;

[0221] A risk prediction model is constructed through the random forest algorithm, and the characteristic evaluation vectors of similar concentrated areas are input into the risk prediction model;

[0222] By formula: Get the gas risk prediction value RiskScore for each similar concentration area;

[0223] Where K is the number of trees in the random forest algorithm, represents the risk prediction value of the j-th pair of feature vectors P, where j is the number of the tree in the random forest algorithm;

[0224] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A gas monitoring method for hydrogen production equipment, characterized in that: The steps include: Collect gas fluid parameters in the hydrogen pipeline area of ​​the hydrogen production equipment, build a dynamic flow field model of the hydrogen production equipment, and identify the flow dead zone of the hydrogen production equipment; Perform turbulence analysis on the gas fluid in the dead zone of flow, and identify the gas turbulence corresponding to the boundary area between the dead zone of flow and the conventional flow channel area; Energy concentration analysis is performed on the gas turbulence in the boundary area to obtain the turbulent kinetic energy in the boundary area, and cluster analysis is performed on the turbulent kinetic energy in the boundary area to determine the gas energy concentration area; The gas energy concentration area is determined in the following manner: By formula: , obtain the turbulent kinetic energy k of the boundary area, where are the root mean square of the pulsating velocity components in the three dimensions of X, Y, and Z. ; Based on the turbulent kinetic energy k in the boundary area, the turbulent kinetic energy field is constructed; The turbulent kinetic energy field is gridded, the neighborhood radius and minimum number of points of the clustering algorithm are set, and the density-based clustering algorithm is used to identify the gas energy concentration area; Obtain the positive concentration area of ​​the gas energy concentration area, and determine whether the positive area ratio and turbulent kinetic energy of the positive concentration area grow in the same direction. If they grow in the same direction, classify the gas energy concentration areas that meet the aggregation conditions to obtain the same type of concentration areas; A risk prediction model is constructed to extract the gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentrated areas, construct a characteristic evaluation vector, input the characteristic evaluation vector into the risk prediction model, and determine the gas risk prediction value of each gas energy concentration area in the same concentrated area.

2. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The flow dead zone is identified as follows: Real-time synchronous acquisition of gas and fluid parameters in the hydrogen pipeline area, and construction of a dynamic flow field model of the hydrogen production equipment based on gas parameters and flow field parameters; Based on the dynamic flow field model of hydrogen production equipment, the flow dead zone in the hydrogen pipeline area of ​​the hydrogen production equipment is identified; Different from the flow dead zone in the hydrogen production equipment area, other areas of the hydrogen pipeline area are marked as regular flow channel areas.

3. A gas monitoring method for hydrogen production equipment according to claim 2, characterized in that: The construction method of the dynamic flow field model of the hydrogen production equipment is as follows: Preprocessing and spatial and temporal alignment of gas parameters and flow field parameters; Extract gas flow field characteristics; Dynamic flow modeling.

4. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The gas turbulence is identified as follows: Obtaining gas turbulence characteristic parameters in the boundary area between the flow dead zone and the conventional flow channel area during the monitoring period; The gas fluid characteristic parameters include: pulsation intensity and Reynolds number within the monitoring period; Based on the gas fluid characteristic parameters composed of pulsation intensity and Reynolds number within the monitoring period, the gas turbulence in the boundary area between the flow dead zone and the conventional flow channel area is identified by the dual-parameter threshold method.

5. A gas monitoring method for hydrogen production equipment according to claim 4, characterized in that: The pulsation intensity within the monitoring period is obtained in the following manner: Obtain the average gas flow rate and instantaneous gas flow rate during the monitoring period; Calculate the pulsating velocity component of the average gas flow velocity and the instantaneous gas flow velocity by using the difference formula; The root mean square of the pulsation velocity component is calculated, and the root mean square of the pulsation velocity is ratioed to the average gas flow rate to obtain the pulsation intensity within the monitoring period.

6. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The turbulent kinetic energy field is constructed as follows: The turbulent kinetic energy corresponding to the boundary area is mapped in space, and the turbulent kinetic energy field is constructed through the spatial interpolation algorithm.

7. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The neighborhood radius is determined as follows: The neighborhood radius is characterized by the Taylor microscale within the boundary region, which is calculated by the dynamic viscosity and pulsating velocity components.

8. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The forward concentration area is obtained in the following manner: Obtain the turbulent kinetic energy in the gas energy concentration area and calculate the first-order difference of the turbulent kinetic energy at the initial and end times of the monitoring period; If the first-order difference is positive, the gas energy concentration area is marked as a positive concentration area.

9. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The method for judging whether the positive area ratio and turbulent kinetic energy of the positive concentration zone increase in the same direction is as follows: Perform ratio analysis on the area of ​​the positive concentrated area to obtain the positive area ratio, and construct the positive area sequence based on the positive area ratio; Obtain the turbulent kinetic energy of the positive concentration area in multiple monitoring periods and construct a kinetic energy sequence; The positive area ratio and turbulent kinetic energy series of the positive concentration zone were analyzed in the same direction, and the slope of the fitting equation of the positive area series and turbulent kinetic energy was obtained; If the slopes of the fitting equations of the forward area series and the turbulent kinetic energy are both greater than 0, it is considered that the forward area ratio and the turbulent kinetic energy in the forward concentration zone increase in the same direction.

10. A gas monitoring method for hydrogen production equipment according to claim 9, characterized in that: The slope of the fitting equation of the forward area series and turbulent kinetic energy is obtained as follows: By performing linear regression analysis on the forward area series and turbulent kinetic energy series, the slopes of the fitting equations of the forward area series and turbulent kinetic energy were obtained.

11. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The method for obtaining the same type of concentrated area is as follows: The gas energy concentration areas in all boundary areas are screened, and the gas energy concentration areas that meet the aggregation conditions are classified to obtain similar concentration areas.

12. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The polymerization conditions are: Polymerization conditions 1. The gas energy concentration zone is a positive concentration zone; Aggregation condition 2: The positive area ratio and turbulent kinetic energy of the positive concentration zone increase in the same direction.

13. A gas monitoring method for hydrogen production equipment according to claim 1, characterized in that: The gas risk prediction value is obtained in the following manner: Obtain gas fluid parameters, turbulent kinetic energy and fault distribution rate in similar concentrated areas and construct characteristic evaluation vectors; A risk prediction model is constructed through the random forest algorithm, and the characteristic evaluation vectors of similar concentrated areas are input into the risk prediction model; The gas risk prediction value of each similar concentration area is obtained through the risk prediction model.

14. A gas monitoring system for hydrogen production equipment, used to implement a gas monitoring method for hydrogen production equipment according to any one of claims 1 to 13, characterized in that: Includes the following modules: Dead zone identification module: used to collect gas fluid parameters in the hydrogen pipeline area of ​​the hydrogen production equipment, build a dynamic flow field model of the hydrogen production equipment, and identify the flow dead zone of the hydrogen production equipment; Turbulence identification module: used to perform turbulence analysis on the gas fluid in the dead zone of flow, and identify the gas turbulence corresponding to the boundary area between the dead zone of flow and the conventional flow channel area; Concentration analysis module: conducts energy concentration analysis on the gas turbulence in the boundary area to obtain the turbulent kinetic energy in the boundary area, performs cluster analysis on the turbulent kinetic energy in the boundary area, and determines the gas energy concentration area; Regional classification module: obtain the positive concentration area of ​​the gas energy concentration area, determine whether the positive area ratio and turbulent kinetic energy of the positive concentration area grow in the same direction, and if they grow in the same direction, classify the gas energy concentration areas that meet the aggregation conditions to obtain the same type of concentration areas; Risk prediction module: used to build a risk prediction model, extract gas fluid parameters, turbulent kinetic energy and fault distribution rate of similar concentrated areas, build a characteristic evaluation vector, input the characteristic evaluation vector into the risk prediction model, and determine the gas risk prediction value of each gas energy concentration area in the same concentrated area.

Images