Method for detecting methane catalytic performance of catalytic fiber

Abstract

The invention discloses a methane catalytic performance detection method for catalytic fibers, relates to the technical field of catalytic performance detection, and aims to solve the problem of methane catalytic performance detection of the catalytic fibers in different regions and under long-term operation conditions. The method comprises the following steps: firstly, putting catalytic fibers into a controllable atmosphere furnace before detection, and carrying out programmed heating and constant-temperature pretreatment according to a preset heating curve; after pretreatment, mechanical cutting, flatness checking and shaping are conducted on the catalytic fibers, so that the catalytic fibers meet the size and installation requirements. Independent control and proportion setting are carried out on gas inlet flow of methane and air in a reaction stage, a stable atmosphere is formed through a mixing section and a homogenizing section, and sampling detection is carried out on reaction gas in different temperature sections. A micro-channel array and an adjustable micro-valve are arranged at a detection outlet, spatio-temporal data are collected through a multi-point sensor, micro-valve adjustment is carried out, finally, gas response data of partitions are obtained, and long-term operation monitoring is completed.

Description

Technical Field

[0001] The present invention relates to the technical field of catalytic performance detection, and more particularly to a method for detecting the methane catalytic performance of a catalytic fiber. Background Art

[0002] In existing methane catalytic detection technologies, only single-point gas sampling and average response analysis are usually performed on the entire catalytic unit, which cannot identify the local differences in the catalytic reaction caused by factors such as thickness deviation, surface roughness and carbon deposition in different areas of the catalytic fiber after long-term operation. Existing technologies mostly use unified proportional regulation for gas supply, lacking refined control of the independent flow rates of methane and air, resulting in fluctuations in the reaction gas components and affecting detection stability. At the same time, when sampling the gas at the outlet of the reaction channel, most of them rely on single-channel analysis of the confluence outlet, ignoring the differences in fluid properties between micro-scale channels, and cannot provide response data reflecting the true state of the local area. The verification of the temperature rise curve and gas mixing effect during the detection process also relies on manual experience, and lacks repeatable closed-loop verification methods. The tracking and analysis of changes in catalytic fiber performance during long-term operation also remains at the macro level, and it is impossible to dynamically adjust the microvalve channel and correct the local response through time series data. Therefore, the existing technology has deficiencies in atmosphere control, independent detection of micro-scale flow channels, long-term performance tracking and multi-channel synchronous regulation; In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0003] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for detecting the methane catalytic performance of a catalytic fiber to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions: In a preferred embodiment, it comprises: The catalytic fiber is pretreated in a controlled atmosphere furnace by monitoring oxygen concentration, controlling the atmosphere, heating according to a heating curve and maintaining the temperature constant; The catalytic fiber is subjected to size collection, CNC cutting, flatness verification, and shape and size adjustment in the finalized tooling; The methane and air flow rates are set by mass flow controllers to complete gas mixing and perform gas chromatography detection according to the temperature rise curve; A microchannel array is processed in the outlet area and an adjustable microvalve is assembled to collect local data and perform adjustment and sampling analysis.

[0005] In a preferred embodiment, the catalytic fiber is pretreated in a controlled environment before testing. After heating and constant temperature treatment according to a set heating curve, its size is collected, CNC cut, flatness and thickness are checked, and its shape and size are adjusted in a fixed tooling to meet installation requirements.

[0006] In a preferred embodiment, the flow rates of methane and air are set and controlled separately, the two gases are introduced into the mixing section and the homogenizing pipe section for mixing, the components are detected at the outlet detection point, the heating starting temperature, target temperature, holding time and heating rate are set and a heating curve file is generated, a simulation run is performed and the temperature curves are compared, a standard gas is selected to establish a chromatographic response relationship, the reaction outlet gas is sampled and detected in each temperature section, and the methane conversion rate is calculated.

[0007] In a preferred embodiment, an interlaced microchannel array is formed in the outlet area, a mounting slot is processed at the starting end of each microchannel and an adjustable microvalve is assembled, multiple pressure and temperature sensing units are arranged to collect data at time intervals, a spatiotemporal parameter matrix is ​​generated, and the microvalve adjustment coefficient is calculated.

[0008] In a preferred embodiment, the short-term flow regulation trend is extracted, the abnormal time range is screened, the short-term disturbance adjustment interval is marked and a microvalve adjustment mask is generated, the local adjustment direction vector is calculated and the microvalve is driven to perform the adjustment, so that the gas is diverted into multiple independent sub-channels.

[0009] In a preferred embodiment, a sampling probe is configured at the outlet of each sub-channel to collect gas flow rate, temperature and concentration data, and a gas response curve of each sub-region is formed within a preset sampling interval.

[0010] In a preferred embodiment, the collected data is normalized and the conversion rate offset of each sub-region is calculated, a multi-dimensional feature vector is generated and clustered to obtain a performance level label, the performance level label is associated with the time series data and the microvalve adjustment coefficient to form a comprehensive data matrix, the data is analyzed to filter out the sub-regions with abnormal performance and the adjustment trajectory is traced back, the expected performance recovery target curve is calculated and the microvalve adjustment correction amount is obtained, the correction amount is transmitted to the corresponding microvalve control unit to execute the adjustment, all sub-region data are integrated to calculate the cluster-level adjustment coefficient and synchronously drive the microvalves in the cluster to adjust.

[0011] In a preferred embodiment, water vapor is introduced through a constant temperature liquid container and the gas is uniformly mixed through a mixing section, and a humidity monitoring probe is used at the outlet to detect the humidity and provide feedback correction.

[0012] In a preferred embodiment, the furnace body is driven to heat according to the temperature rise curve and the reaction gas is sent to the gas chromatograph for sampling and analysis, the mass flow controller is called to adjust the intake flow rate of methane and air, the reaction channel is kept under kinetic control, the reaction data of different temperature sections are repeatedly detected and recorded, the furnace body is switched to a constant temperature operation mode, and automatic sampling is performed at time intervals and the test results are recorded, the conversion rate change curve is drawn, and the activity retention rate and stability index under long-term operation conditions are calculated.

[0013] The technical effects and advantages of the method for detecting the methane catalytic performance of catalytic fibers of the present invention are as follows: The present invention achieves the unification of sample states before detection by pre-treating the catalytic fibers with a controlled atmosphere and finalizing their dimensions before detection. During the reaction detection process, the flow rates of methane and air are set and independently controlled, a stable detection atmosphere is formed through the mixing section and the homogenization section, and the reaction data of each temperature section are made comparative through gas chromatography response calibration. By arranging a microchannel array and an adjustable microvalve in the outlet area, it is possible to independently sample the local gas channel, and to perform closed-loop adjustment of the microvalve opening according to the threshold in each sampling cycle, thereby forming a dynamic association between the local fluid state and the response data. Further combining time series sampling and cluster-level integration, it is possible to correct areas of performance differences, achieve continuous monitoring and dynamic adjustment under long-term operating conditions, and possess higher detection accuracy and zoning management capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a timing diagram of a method for detecting the methane catalytic performance of a catalytic fiber according to the present invention.

[0015] Figure 2 This is a flow chart of an implementation method for detecting the methane catalytic performance of a catalytic fiber according to the present invention. DETAILED DESCRIPTION

[0016] 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.

[0017] Example The present invention discloses a method for detecting the methane catalytic performance of catalytic fibers. Figure 1 As shown, including: The catalytic fiber is pretreated in a controlled atmosphere furnace by monitoring oxygen concentration, controlling the atmosphere, heating according to a heating curve and maintaining the temperature constant; The catalytic fiber is subjected to size collection, CNC cutting, flatness verification, and shape and size adjustment in the finalized tooling; The methane and air flow rates are set by mass flow controllers to complete gas mixing and perform gas chromatography detection according to the temperature rise curve; A microchannel array is processed in the outlet area and an adjustable microvalve is assembled to collect local data and perform adjustment and sampling analysis.

[0018] like Figure 2 As shown, before the catalytic fiber enters the test, the catalytic fiber to be tested is first placed in a controllable environment of air atmosphere, wherein the controlled environment furnace is pre-installed with an air circulation system and an oxygen concentration online monitoring probe. The oxygen concentration value is collected by the probe in real time and fed back to the furnace control program to ensure the stability of the atmosphere composition.

[0019] On this basis, the first heating curve file preset in the furnace control program is used to drive the electric heating unit in the furnace to perform programmed heating operations. The temperature value in the furnace is collected in real time by the distributed thermocouple array and transmitted back to the furnace control program. The furnace control program performs closed-loop adjustment based on the temperature collection data, so that the furnace temperature gradually increases according to the heating curve and remains constant after reaching the target temperature, further promoting a uniform and controllable oxidation reaction on the surface of the catalytic fiber, so that the active metal oxide on the fiber surface is uniformly formed and stably solidified, thereby achieving the full generation of active sites on the surface of the catalytic fiber and the structural stabilization of the catalytic layer.

[0020] The pre-treated catalytic fibers are then fed into a mechanical cutting device equipped with a high-precision cutting table. Before cutting begins, a laser rangefinder and displacement encoder are used to collect the fiber's dimensions and initial position. During the cutting process, a CNC-driven cutting tool cuts the fibers to the preset dimensions. A clamping device on the cutting table simultaneously secures the catalytic fibers in place, ensuring zero displacement in the thickness direction throughout the entire cutting process.

[0021] After cutting, the built-in optical flatness detection is used to check the cross-sectional flatness and thickness consistency of the cut sheet structure to ensure that the processing meets the design requirements.

[0022] At the same time, during the shaping process after cutting is completed, the sheet catalytic fiber is placed in a shaping tool with temperature control and pressure regulation. The real-time temperature and pressure values ​​are collected by embedded sensors and fed back to the shaping control program. The shaping control program adjusts the working status of the heating element and the pressing unit according to the parameters, and gradually adjusts the shape and size of the sheet structure while ensuring the integrity of the fiber structure, so that its appearance and channel installation dimensions are strictly matched.

[0023] Furthermore, to ensure a constant gas composition entering the reaction zone, mass flow regulation technology is used to set the flow rates of methane and air separately. Specifically, the methane and air gases are connected to independent gas circuits equipped with mass flow control units. The flow control units use internal flow sensors to collect real-time flow rates of methane and air, and feed the collected data into the flow control program. The control program then presets the methane-to-air ratio and drives the electronically controlled valve to a stable output state, achieving constant flow control of the two gases within the independent gas circuits.

[0024] It should be noted that the preset ratio of methane to air and the stable output state of the electronically controlled valve need to be adjusted according to actual conditions. In this embodiment, the default situation is taken as an example.

[0025] Subsequently, the two independently controlled gases are introduced into the gas mixing section. A fixed turbulence structure is designed inside the mixing section, so that the two incoming gases are staggered and diverted between the turbulence blades, and initially uniformed under the action of turbulent diffusion.

[0026] The mixed gas then enters the homogenization section, which re-mixes the gas components spatially by extending their residence time and leveraging the uniform distribution structure within the section. To verify the mixing effect, a gas composition detection point is set up at the outlet of the mixing section. This point uses an online gas composition sensor to continuously collect gas composition data. When the sensor output signal remains stable and without fluctuations within a fixed period after real-time analysis by the control program, uniform mixing is confirmed.

[0027] After the gas environment stabilizes, the starting temperature for heating, the target temperature, the holding time of each temperature section, and the heating rate are determined based on the characteristics of the catalytic fiber after the aforementioned pretreatment that is suitable for catalytic combustion of methane. The starting temperature for heating, the target temperature, the holding time of each temperature section, and the heating rate are then written into the furnace control to generate a second heating curve file for controlling the heating unit. A simulation run test is then performed without a sample. The temperature data is collected in real time by a thermocouple array arranged in the furnace cavity, and the real-time temperature curve is compared with the second heating curve file. The power output of the heating unit is adjusted to gradually approximate the actual heating process to the second heating curve file. When the temperature deviations of multiple consecutive simulation runs are within the allowable range, the second heating curve file is confirmed to be valid.

[0028] The allowable range of temperature deviation needs to be set according to specific actual scenarios, which will not be specifically explained in this embodiment.

[0029] During the product testing phase, to establish a traceable conversion relationship for chromatographic responses, a standard gas with known composition and concentration and certified traceability by a third party is first selected. This gas is then introduced into the gas chromatograph's sampling inlet via a calibrated mass flow control unit, with the flow rate controlled to align with subsequent actual testing conditions. Once the gas chromatograph enters a stable operating state, multiple injections are performed at fixed time intervals. The chromatographic system records the peak area data for methane and carbon dioxide corresponding to the standard gas. A linear fit is then performed between the peak area and the molar content of the standard gas to establish a corresponding curve between peak area and molar content. The response factors of the chromatographic process for methane and carbon dioxide are calculated and stored, ensuring the traceable conversion of subsequent test data from chromatographic signals to actual concentrations.

[0030] After calibration, set temperature ranges and continuously sample the reaction outlet gas within each temperature range using the previously calibrated gas chromatography analysis technology. A sampling probe introduces the outlet gas in real time, and the gas chromatograph collects the peak areas of methane and carbon dioxide. These peak areas are then substituted into the established response factor calculation formula to output the methane conversion rate for that temperature range.

[0031] It should be noted that, considering that within a single catalytic fiber, due to slight differences in local thickness, local processing roughness, or carbon deposits formed after long-term operation, the reaction states of different regions of the same catalytic fiber are different; at the same time, when the gas flows through the catalytic fiber, there are slight deviations in the inlet gas distribution, and the temperature gradient in the local area will also cause inconsistent catalytic effects. When sampling and analyzing the gas at the outlet of a single catalytic fiber, the existing detection step usually only performs a single point detection at the confluence outlet of the entire piece. Although this simplifies the analysis process, it is unable to identify the performance differences of different regions of the same catalytic fiber, which results in the test results only reflecting the average value and failing to present local anomalies. Therefore, in this embodiment, when performing sampling and detection, high-precision microfluidic control processing technology is adopted. Under the drive of a CNC micromachining platform, the tool position, channel depth, and channel width are collected in real time by the laser ranging sensor and displacement sensor embedded in the platform, and the tool position, channel depth, and channel width data are fed back to the processing control program in real time. The processing control program performs closed-loop correction on the tool path based on the three-dimensional layout parameters in the design input file, and finally accurately forms an interlaced microchannel array in the outlet area.

[0032] At the starting end of each microchannel, micro-nano processing technology is further used to process a spiral winding installation slot. The slot geometric dimensions are collected by displacement sensors and visual recognition modules and fed back to the micro-nano processing controller. The micro-nano processing controller makes real-time corrections based on the slot geometric parameters derived from the previous CAD model in the design drawings.

[0033] Subsequently, the adjustable microvalve is fixed to the slot through an embedded locking method. The locking process is performed by a torque-controlled micro-actuator, and the locking depth and position are collected by a displacement detection unit and fed back to the locking control program for dynamic adjustment. This ensures that the spatial index position of the adjustable microvalve body in the channel array corresponds one-to-one with the design parameters, achieving equidistant distribution of the adjustable microvalves in space and enabling the outlet structure to have multiple independently controllable microvalve channels. After the adjustable microvalve is assembled, multiple microfluidic pressure sensing units and microtemperature sensing units are arranged in a ring to obtain the local fluid state of each microchannel. Each sensing unit senses the pressure and temperature of the gas flowing through that location through a built-in sensitive element. Then, the process setting document is inspected, and the sampling time parameter N1 is selected based on the experimental requirements and the equipment response time. All sensor units are synchronously triggered to sample at the time interval N1. The collected raw data are marked in the control program with the time coordinate t automatically generated by the sampling timestamp and the spatial index x written into the control program when the adjustable microvalve is assembled, forming a spatiotemporal fluid parameter matrix.

[0034] Subsequently, a difference operation is performed on the spatiotemporal fluid parameter matrix to calculate the pressure change rate and temperature change rate of each position within the N1 time window, and the time accumulation and summation are performed according to the weight to generate the spatiotemporal composite characteristic curve of each microchannel.

[0035] Among them, the mathematical model of the microvalve adjustment coefficient K(t,x) is as follows: K(t,x)=α×(ΔP(t,x) / Δt)+β×(ΔT(t,x) / Δt); ΔP(t,x) represents the pressure change at the spatial index position within the time window N1, ΔT(t,x) represents the temperature change at the spatial index position within the time window N1, Δt represents the sampling interval, and α and β are derived from the previous experimental data fitting and tuning process. The specific values ​​are recorded in the algorithm configuration file.

[0036] For example, in a certain temperature range T1, the pressure change at the spatial index position in the time window N1 is ΔP = 0.6 kPa, the temperature change is ΔT = 0.8 degrees Celsius, the sampling interval Δt is one second, and the weight coefficients α and β are determined by experimental fitting to be 0.55 and 0.45, respectively. Substituting into the above formula, the microvalve adjustment coefficient K(t,x) of the spatial index position at the current time point is 0.69.

[0037] The fitting method for α and β utilizes an orthogonal experimental design, specifically a three-factor, four-level combination of temperature, flow velocity, and thickness difference variables. Representative catalytic fiber samples were selected to conduct joint detection experiments on microchannel flow response and temperature changes under multiple control conditions. Each set of experiments collected no fewer than 100 time window sample points, with a cumulative sample size exceeding 2,000 sets. After outlier removal and normalization preprocessing, the experimental data was regressed on the response model of the microvalve adjustment coefficient K(t,x) using a least-squares fitting function, resulting in the minimum error range for the α and β fitting values. The values ​​for α ranged from [0.45, 0.60], and for β ranged from [0.35, 0.50]. A priority weight function, fdisturbance(t,x), was selected based on the type of fluid disturbance to adjust the dynamic weight ratio. All experimental data and fitting results were stored in the algorithm configuration file and loaded into the control program.

[0038] The dynamic weight function derived from the algorithm configuration file, which includes the weight ratio obtained from historical experimental fitting and the real-time adjustment factor, is further called to operate on the spatiotemporal composite characteristic curve to calculate the microvalve adjustment coefficient K(t,x), where K is the adjustment coefficient, t is the sampling time point, and x is the spatial position of the adjustable microvalve in the array.

[0039] Considering that local microchannels may produce instantaneous high-amplitude fluctuations due to short-term thermal shock, sudden change in gas density, or mechanical inertia of the adjustable microvalve, after calculating the microvalve adjustment coefficient K(t,x), K(t,x) in the continuous time period is expanded into a spatiotemporal characteristic plane according to the sliding time window N2.

[0040] The window length of N2 is determined by the control program after reading the experimental configuration file and optimizing it through multiple groups of experiments in the early stage, and is a pre-set analysis window parameter. Subsequently, a local curvature analysis algorithm was used to calculate the characteristic planes and extract the instantaneous flow fluctuation pattern for each adjustable microvalve position. Spatial differentiation and time-weighted calculations were performed. The time-weighted coefficient was derived from historical empirical values ​​in the configuration file, and the spatial differentiation spacing was calculated based on the adjustable microvalve spacing in the assembly design. This yielded the short-term flow regulation trend for each position.

[0041] Furthermore, the short-term flow regulation trend is compared with the mean of the historical steady-state segment to screen out the time range with abnormal change amplitude.

[0042] Within the abnormal time range, the fast Fourier decomposition algorithm is called to decompose the microvalve adjustment coefficient K(t,x) into a high-frequency disturbance component and a low-frequency reference component, and the energy proportion of the high-frequency component is calculated. The energy proportion calculation parameters are derived from the built-in formula of signal processing and the collected data.

[0043] When the proportion of high-frequency energy in two consecutive N2 windows exceeds the preset threshold, the microchannel is marked as a short-term disturbance adjustment interval, and a microvalve adjustment mask M is generated. The mask M is in the form of a matrix data structure, each element corresponds to a channel position and marks whether it is prioritized for adjustment. The mask data is stored in the adjustment queue in real time.

[0044] The preset and adjustment thresholds are determined through orthogonal experimentation. Specifically, during the experimental phase, multiple different threshold combinations are preset. Repeated testing is performed for each combination, and the corresponding conversion rate offset, response fluctuation, and adjustment stability indicators are recorded. The optimal parameter range is determined through multi-factor orthogonal analysis. The final selected parameters are then written into the configuration file called by the control program as the preset and adjustment thresholds.

[0045] When performing priority regulation, vector operations are used to calculate the local regulation direction vector by combining the instantaneous rate of change of K(t,x) within the current N2 window with the gradient difference at adjacent spatial indices. The gradient difference is calculated by the difference in the microvalve adjustment coefficients K(t,x) between adjacent microvalves, and the difference data is derived from real-time sampling data in the same time window.

[0046] The adjustment direction vector is then transmitted to the microvalve driver, which performs micron-level rotational step adjustment of the valve disc angle based on the direction vector. The adjustment amplitude is collected in real time by the angle feedback module and fed back to the control program to form a closed-loop control.

[0047] Then, the opening of the corresponding microvalve is adjusted in real time according to the updated microvalve adjustment coefficient K(t,x); the reaction gas that is about to be exported is diverted into multiple sub-channels with uniform flow and isolated from each other, thereby dividing the effective detection area of ​​the catalytic fiber into multiple independent sub-areas at the physical level.

[0048] It's important to note that an adjustment threshold is set within each sampling cycle. When the real-time calculated K(t,x) deviates from the baseline value by more than the adjustment threshold, the actuator unit is immediately driven to adjust the valve disc angle. After the adjustment is completed, the sensor unit recollects data in the next sampling cycle and substitutes the new data into the K(t,x) calculation process, forming a continuous feedback correction process.

[0049] Among them, the adjustment threshold value comes from the adjustment strategy configuration table and is determined by preliminary experiments.

[0050] To further enhance local detection accuracy and adjustment capabilities, a miniature sampling probe is installed at the exit of each physically divided subchannel. This sampling probe integrates a miniature heat flow sensor unit and a local gas composition sensor unit. The sensor elements simultaneously sense gas flow rate, temperature, and methane and carbon dioxide concentrations at the instant of sampling. The sensor output is converted to digital and transmitted to a synchronous acquisition unit. This unit samples and buffers the multidimensional data at predetermined sampling intervals within the control program, generating independent gas response curves for each subchannel.

[0051] The sampling interval value is derived from the system detection configuration file and is set in combination with the sensor response time; Furthermore, the gas concentration signals obtained from each sub-region at the same time point are normalized with the corresponding local temperature and flow rate signals.

[0052] Among them, the normalization parameters are derived from the standard operating condition data stored in the calibration phase to ensure that the calculation results are comparable at the same scale.

[0053] The standard operating condition data is obtained by calibration using standard gases before the experiment. During the calibration process, standard gases are introduced into the detection unit through sampling lines at a fixed flow rate and known concentration. Peak area data for methane and carbon dioxide are collected repeatedly. A linear fit is used to establish a corresponding relationship between peak area and molar concentration. This results in a baseline gas concentration value under standard operating conditions, which is used as the standard operating condition data in the normalization calculation.

[0054] By comparing the normalized data, the additional conversion rate offset caused by thickness deviation, surface roughness, and carbon deposit accumulation in each sub-region is calculated. The statistical mean of the historical steady-state sampling data is then used as the overall baseline value. The offset of each sub-region is then compared with the overall baseline value. A multi-dimensional feature vector is generated by combining the offset magnitude, data stability indicators, and response fluctuation amplitude.

[0055] The multidimensional feature vector is specifically defined as: V(x)=[Rc(x),Fv(x),Tp(x)]; Rc(x) represents the normalized conversion rate offset of the sub-area, which comes from the gas concentration of the sampling probe. Fv(x) represents the normalized value of the local gas flow rate, which comes from the flow rate of the micro heat flow sensing unit. Tp(x) represents the normalized value of the local temperature, which comes from the temperature data of the micro temperature sensing unit.

[0056] Furthermore, dynamic clustering is used to cluster the multidimensional feature vectors and generate independent performance level labels for each sub-region; It should be noted that this embodiment uses the K-means clustering algorithm due to its fast convergence and stable results when processing continuous physical characteristic data. Specifically, the number of cluster centers, k, is determined by performing silhouette coefficient analysis on the distribution characteristics of historical experimental data. The initial centers are automatically initialized using the k-means++ method. The iteration process aims to minimize the intra-cluster squared error. The final performance level label is output when the change in cluster centers between two consecutive iterations is less than the convergence threshold specified in the configuration file.

[0057] At the same time, the historical sample data used for K-means clustering training comes from the multi-dimensional feature vector library accumulated during the long-term detection process. The total number of samples exceeds two hundred groups. Each group of data corresponds to the conversion rate offset, flow rate and temperature data of different sub-areas in multiple sampling windows. The performance level labels are manually annotated through historical records and standard performance test results.

[0058] To ensure the stability and representativeness of the k value, a silhouette coefficient threshold of 0.5 was set. Clustering effects were tested for k values ​​in the range [2, 10]. The k corresponding to the maximum silhouette coefficient was selected as the final number of cluster centers. The historical sample library exceeded 200 groups, covering subregional data from different temperature ranges and conditioning conditions. The eigenvector dimensions were unified as V(x) = [Rc(x), Fv(x), Tp(x)]. During the iteration process, the k-means++ method was used to initialize the center points, and a cluster center change convergence threshold of 0.01 was set. The clustering process was terminated if the change was less than this value for two consecutive times. These clustering hyperparameter settings are recorded in the algorithm control file and dynamically loaded at runtime.

[0059] Furthermore, a multi-dimensional correlation is performed between each sub-region's performance level label and its corresponding time series gas response curve, the real-time calculated microvalve adjustment coefficient K(t,x), and the historical temperature and flow rate feature vectors to construct a comprehensive data matrix. This matrix is ​​then analyzed to identify sub-regions where performance levels remain low or fluctuate frequently over multiple consecutive sampling windows. The microvalve adjustment trajectory data for these sub-regions, stored in the historical record module, is then used to perform local time series backtracking analysis, combined with real-time flow regulation feedback. Specifically, the adjustment coefficient K(t,x) is correlated and compared with the performance level label changes over multiple sampling windows, extracting the coupled change relationship from the historical adjustment data and the current state data.

[0060] After obtaining the coupling change relationship, the local trend fitting algorithm is called to extract the dominant factors affecting the performance level change based on the fitting model parameters stored in the algorithm configuration file, and the expected performance recovery target curve of the sub-area is calculated.

[0061] The target curve is differentially calculated with the current microvalve adjustment coefficient K(t,x) to obtain a new microvalve adjustment correction value Kk(t,x) and transmit it to the corresponding microvalve control unit. The control unit is driven to prioritize the gas flow regulation of this sub-area in the next sampling cycle to shorten the performance recovery time.

[0062] Furthermore, the data from all sub-regions is integrated at the cluster level. Specifically, by analyzing the spatial proximity of each sub-region and the similarity of their performance level labels, the sub-regions are grouped into several performance clusters. Within each performance cluster, the microvalve adjustment correction value Kk(t,x) is weighted averaged to calculate the cluster-level adjustment coefficient Kc(t). The weight parameters are derived from the historical response stability data and current fluctuation amplitude indicators of each sub-region within the cluster.

[0063] In the subsequent sampling period, the cluster-level adjustment coefficient Kc(t) is used to synchronously drive the opening adjustment of each microvalve in the cluster, so that the flow and reaction state of all sub-areas in the cluster gradually tend to be consistent, thereby further reducing local performance differences and achieving reaction balance and performance improvement of the overall catalytic fiber.

[0064] During the water vapor resistance test, a constant temperature liquid container is used as a carrier. The container is filled with deionized water output by the pure water preparation equipment, and the built-in constant temperature control unit performs closed-loop regulation according to the set temperature value pre-written in the control program. When the reaction gas enters the container, it passes through the porous bubbling pipe to form stable bubbles under the liquid surface. During the rising process, the gas carries water vapor through the liquid.

[0065] The temperature value comes from the experimental configuration file.

[0066] After leaving the liquid, the gas enters the mixing section, where spoilers and a homogenizing grid are arranged. The fluid homogenizing structure implements multi-directional, staggered diversion and remixing of the gas carrying water vapor, achieving uniform distribution of the water vapor within the gas. An online humidity monitoring probe with a built-in capacitive humidity sensor is installed at the outlet of the mixing section to collect humidity signals in real time. This signal is compared with the humidity range set in the control parameter file. If the signal deviates from this range, feedback correction is implemented by adjusting the heating power of the bubbling device and the air flow control valve until the humidity value stabilizes.

[0067] The humidity feedback control formula is as follows: Uctrl=Kh×(Hset-Hmeas); Where Hset represents the target humidity set in the control parameter file, Hmeas represents the real-time measured humidity value, and Kh represents the humidity adjustment gain coefficient, which is derived from the humidity calibration experiment record.

[0068] Then, the electric heating unit inside the furnace is driven according to the heating curve parameters in the experimental configuration file to achieve programmed heating of the starting temperature, heating rate and target temperature.

[0069] The reaction gas at the outlet enters the gas chromatography analysis unit through the sampling pipeline. The chromatographic system separates the gas through a capillary column and uses a hydrogen flame ionization detector to collect the peak area data of methane and carbon dioxide. The detection data is processed into actual concentration values ​​by the response factor conversion module.

[0070] Among them, the response factor comes from the calibration step completed with standard gas before the test. The control program compares the calculated conversion rate with the historical benchmark value under anhydrous conditions and outputs the quantitative evaluation results of the water vapor resistance performance.

[0071] Furthermore, a dual-channel mass flow controller is called to adjust the intake flow of methane and air respectively. The thermal flow sensor inside the controller collects the flow in real time and compares it with the set value of the kinetic range. The proportional valve is adjusted to achieve gas supply in the low conversion rate range.

[0072] After the reaction channel is maintained in the kinetic control state, the above-mentioned water-containing atmosphere detection process is repeated, and gas chromatography data are continuously collected at each temperature section. The control program calculates and records the reaction rate at different temperature sections.

[0073] Next, to evaluate the stability of the catalytic fiber under long-term operating conditions, the furnace was switched to constant temperature operation mode, with the temperature set by the long-term operation profile. Automatic sampling was periodically triggered at intervals N1, and the outlet gas entered the gas chromatography analysis unit through a sampling line. The test results were converted into concentration data based on the response factor.

[0074] Finally, a conversion rate curve was plotted with time as the horizontal axis, and the slope and fluctuation range of the curve were calculated to determine the maintenance of catalytic fiber activity and performance degradation trends. Activity retention and stability indicators under long-term operating conditions were obtained.

[0075] The calculation formula of activity retention rate A-rate is as follows: A-rate=(Rend / Rinit)×100%; Among them, Rend represents the average conversion rate measured after long-term operation, and Rinit represents the average conversion rate measured in the initial operation stage. Both are derived from the data after the aforementioned gas chromatography detection and normalization processing.

[0076] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0077] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0078] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0079] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0080] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0081] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. 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 method for detecting the methane catalytic performance of catalytic fibers, It is characterized by including: The catalytic fiber is pretreated in a controlled atmosphere furnace by monitoring oxygen concentration, controlling the atmosphere, heating according to a heating curve and maintaining the temperature constant; The catalytic fiber is subjected to size collection, CNC cutting, flatness verification, and shape and size adjustment in the finalized tooling; The methane and air flow rates are set by mass flow controllers to complete gas mixing and perform gas chromatography detection according to the temperature rise curve; A microchannel array is processed in the outlet area and an adjustable microvalve is assembled to collect local data and perform adjustment and sampling analysis.

2. The method for detecting the methane catalytic performance of a catalytic fiber according to claim 1, characterized in that: The catalytic fiber is pretreated in a controlled environment before testing. After heating and constant temperature treatment according to the set heating curve, its size is collected, CNC cut, flatness and thickness are checked, and its shape and size are adjusted in the finalized tooling to meet the installation requirements.

3. The method for detecting the methane catalytic performance of a catalytic fiber according to claim 2, characterized in that: The flow rates of methane and air are set and controlled separately. The two gases are introduced into the mixing section and the homogenizing pipe section for mixing. Component detection is performed at the outlet detection point. The starting temperature, target temperature, holding time and heating rate are set, and a heating curve file is generated. A simulation run is performed and the temperature curves are compared. Standard gases are selected to establish a chromatographic response relationship. The reaction outlet gas is sampled and tested in each temperature section, and the methane conversion rate is calculated.

4. A method for detecting methane catalytic performance of catalytic fibers according to claim 3, characterized in that; An interlaced microchannel array is formed in the outlet area, and an installation slot is processed at the starting end of each microchannel and an adjustable microvalve is assembled. Multi-point pressure and temperature sensing units are arranged to collect data at time intervals, generate a time-space parameter matrix, and calculate the microvalve adjustment coefficient.

5. The method for detecting the methane catalytic performance of a catalytic fiber according to claim 4, characterized in that: The short-term flow regulation trend is extracted, the abnormal time range is screened, the short-term disturbance adjustment interval is marked and a microvalve adjustment mask is generated. The local adjustment direction vector is calculated and the microvalve is driven to perform the adjustment, so that the gas is diverted into multiple independent sub-channels.

6. The method for detecting the methane catalytic performance of a catalytic fiber according to claim 5, characterized in that: A sampling probe is configured at the outlet of each sub-channel to collect gas flow rate, temperature and concentration data, and a gas response curve of each sub-area is formed within a preset sampling interval.

7. The method for detecting the methane catalytic performance of a catalytic fiber according to claim 6, characterized in that: After normalizing the collected data, the conversion rate offset of each sub-area is calculated, and a multidimensional feature vector is generated and clustered to obtain a performance level label. The performance level label is associated with the time series data and the microvalve adjustment coefficient to form a comprehensive data matrix. The data is analyzed to filter out sub-areas with abnormal performance and the adjustment trajectory is backtracked. The expected performance recovery target curve is calculated and the microvalve adjustment correction amount is obtained. The correction amount is transmitted to the corresponding microvalve control unit for adjustment. The data of all sub-areas are integrated to calculate the cluster-level adjustment coefficient and synchronously drive the microvalves in the cluster for adjustment.

8. A method for detecting methane catalytic performance of catalytic fibers according to claim 7, characterized in that; Water vapor is introduced through a constant temperature liquid container and the gas is evenly mixed through a mixing section. The humidity is detected and feedback corrected using a humidity monitoring probe at the outlet.

9. The method for detecting the methane catalytic performance of a catalytic fiber according to claim 8, characterized in that: The furnace is driven to heat according to the heating curve and the reaction gas is sent to the gas chromatograph for sampling and analysis. The mass flow controller is called to adjust the intake flow of methane and air. The reaction channel is kept under kinetic control to repeatedly detect and record the reaction data of different temperature sections. The furnace is switched to the constant temperature operation mode to automatically sample and record the test results at time intervals. The conversion rate change curve is drawn and the activity retention rate and stability index under long-term operation conditions are calculated.

Images