A pan-tilt linear beam combustible gas detection system and device

Abstract

The application relates to the technical field of gas detection, in particular to a gimbal type linear light beam combustible gas detection system and equipment, which comprises a data acquisition module, a spectrum analysis module, a positioning and identification module and a trend identification module. The wavelength scanning range of the to-be-detected gas is determined by collecting the starting wavelength and the ending wavelength of the to-be-detected gas under the current scene. The absorbance spectrum graph is formed according to the wavelength scanning range of the to-be-detected gas and the absorbance values of each point in the wavelength scanning range. The matching items of the spectrum points of the absorbance spectrum graph are extracted one by one, and the matching gas items of each spectrum point in the absorbance spectrum graph are identified. The phase difference of the matching gas items of each spectrum point at adjacent points is used to locate the coordinates of the leakage points according to the linear relationship between the absorbance and the gas concentration on the leakage points. The response strategies of each leakage point are located according to the change trend of the gas concentration on each leakage point. The precision and positioning accuracy of combustible gas detection are realized.

Description

Technical Field

[0001] This invention relates to the field of gas detection technology, specifically a pan-tilt linear beam combustible gas detection system and device. Background Technology

[0002] Gas detection is a technology used to detect, measure, and monitor the concentration of specific gases in an environment. It can detect a variety of gases, from hazardous gases to those present in the general environment. In fields such as petrochemicals and natural gas transportation, high accuracy and response speed are required for detecting flammable gas leaks. Pan-tilt gas detectors allow for detector rotation and scanning, expanding the detection range. Linear beam technology covers a large area with a single scan, improving the efficiency of flammable gas detection.

[0003] For example, Chinese Patent Publication No. CN116046721A discloses a through-beam open-circuit laser gas detector and a real-time calibration method for gas detection. The transmitting device alternately emits signal light and calibration light. A portion of the laser signal emitted by the laser is controlled by the signal modulation wave and emitted as signal light to the open-circuit environment. The other portion is converted into a calibration modulation wave after passing through a standard gas chamber and controls the laser to generate calibration light. The receiving device receives the optical signal and calculates and calibrates the product of the concentration of the gas to be measured and the distance in the open-circuit environment in real time based on the voltage waveforms of the signal light and the calibration light and the product of the reference gas concentration and the distance.

[0004] For example, Chinese Patent Publication No. CN118190873A discloses a gas detection method, electronic device, detection laser, and sensor. The method includes: obtaining an initial operating temperature, an ambient temperature range, and the absorption center wavelength of the gas to be detected, wherein the ambient temperature range and the initial operating temperature are respectively the operating ambient temperature range and the initial operating temperature of the detection laser used for gas detection; obtaining a target operating temperature based on the ambient temperature range, wherein the target operating temperature is greater than the upper limit of the ambient temperature range; obtaining a wavelength offset based on the initial operating temperature and the target operating temperature; obtaining an initial wavelength range based on the wavelength offset and the absorption center wavelength; determining a target laser chip from multiple preset laser chips based on the initial wavelength range; encapsulating the target laser chip with a preset heating resistor to obtain the detection laser; and heating the detection laser to the target operating temperature using the heating resistor to lock its operating wavelength at the absorption center wavelength for detecting the gas to be detected.

[0005] Existing technologies describe calibration processing through the absorption spectral center to prevent the influence of open-circuit environments; they also point out that heating to the target operating temperature can reduce spectral shifts. However, existing technologies process gas detection based on a standard gas chamber and cannot dynamically detect changes in the concentration of combustible gases within the current area according to the size of the detected area. This results in an inability to respond to the concentration of combustible gases and other gases, thus failing to achieve multi-type identification and prevention of combustible gases and improve the prediction and classification response to combustible gas leaks. Summary of the Invention

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a gimbal-type linear beam combustible gas detection system, comprising: a data acquisition module, used to acquire the starting wavelength and ending wavelength of the gas to be tested in the current scene, and determine the wavelength scanning range of the gas to be tested; based on the wavelength scanning range of the gas to be tested, an absorption spectrum is formed using the absorbance values ​​of each point within the wavelength scanning range.

[0007] The spectral analysis module is used to extract features from the spectral points of the absorption spectrum, extract the matching items for each spectral point one by one, and identify the matching gas items for each spectral point in the absorption spectrum according to the matching value of each matching item.

[0008] The location identification module is used to locate the coordinates of the leak point based on the phase difference between the matching gas terms at adjacent points of each spectral point and the linear relationship between absorbance and gas concentration at the leak point.

[0009] The trend recognition module is used to invert the gas concentration at the leak point and determine the changing trend of the gas concentration at each leak point; based on the changing trend of the gas concentration at each leak point, the response strategy for each leak point is determined.

[0010] The beneficial effects of this invention are as follows: First, this invention determines the wavelength scanning range by querying the absorption spectrum library based on the type of gas to be tested, and achieves dynamic wavelength adjustment by setting step values; it calculates the wavelength offset by combining temperature difference and pressure difference to compensate for the influence of environmental factors on the spectrum, and forms an absorption spectrum based on absorbance values ​​to provide basic data for subsequent analysis; then, the spectral points are divided by band, and spectral features are matched by preset templates, and overlapping absorption peaks are judged by cross-analysis, and gas matching items are determined by combining slope extreme points or maximum absorbance values; this can reduce the influence of environmental factors on the current detection results, improve the accuracy of detection, and prevent misidentification and false identification.

[0011] Second, this invention uses the phase difference and concentration gradient that appear during spectral point measurement for joint localization. It sequentially calculates the phase difference and concentration gradient between the spectral point and its 8 neighboring nodes to construct a phase difference matrix, phase vector direction, and concentration matrix. Combined with the linear relationship between absorbance and gas concentration, the leak point is located linearly through phase difference convergence, concentration gradient matching, and trend curve. This can improve the accuracy of leak point location and ensure the reliability of the location.

[0012] Third, this invention determines the state and trend of a leak point after leakage, and describes the gas concentration change trend at the leak point in the form of state changes. It can identify the prediction and classification response of the leakage trend. Using gas concentration, concentration change rate, rise time, fall time, and plateau time as state variables, it calculates state probabilities and sets response strategies in combination with trend classification. This reduces the problem that the state changes of various gases cannot correspond to each other in scenarios where multiple gases are matched together, and completes the accurate identification of combustible gases in multiple scenarios. Attached Figure Description

[0013] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0014] Figure 1 This is a system framework diagram of a pan-tilt linear beam combustible gas detection system.

[0015] Figure 2 This is a schematic diagram of a pan-tilt linear beam combustible gas detection system.

[0016] Figure 3 This is a schematic diagram of a pan-tilt linear beam combustible gas detection system.

[0017] Figure 4 This is a schematic diagram of a pan-tilt linear beam combustible gas detection system.

[0018] Figure 5 This is a schematic diagram of a pan-tilt linear beam combustible gas detection system.

[0019] Figure 6 This is a schematic diagram of an electronic device. Detailed Implementation

[0020] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Where specific techniques or conditions are not specified in the embodiments, they shall be performed in accordance with the techniques or conditions described in the literature in the art or in accordance with the product manual.

[0021] See Figure 1A pan-tilt linear beam combustible gas detection system includes: a data acquisition module, a spectral analysis module, a positioning and identification module, and a trend identification module; wherein, the output end of the data acquisition module is connected to the spectral analysis module, the output end of the spectral analysis module is connected to the positioning and identification module, and the output end of the positioning and identification module is connected to the trend identification module.

[0022] The data acquisition module is used to acquire the starting and ending wavelengths of the gas to be tested in the current scene, and determine the wavelength scanning range of the gas to be tested; based on the wavelength scanning range of the gas to be tested, an absorption spectrum is formed using the absorbance values ​​at each point within the wavelength scanning range.

[0023] The spectral analysis module is used to extract features from the spectral points of the absorption spectrum, extract the matching items for each spectral point one by one, and identify the matching gas items for each spectral point in the absorption spectrum according to the matching value of each matching item.

[0024] The location identification module is used to locate the coordinates of the leak point based on the phase difference between the matching gas terms at adjacent points of each spectral point and the linear relationship between absorbance and gas concentration at the leak point.

[0025] The trend recognition module is used to invert the gas concentration at the leak point and determine the changing trend of the gas concentration at each leak point; based on the changing trend of the gas concentration at each leak point, the response strategy for each leak point is determined.

[0026] In the data acquisition module, since the scene contains various combustible gases, each with different absorption intensities in the near-infrared region, they will exhibit different trends in the acquired spectrum. Therefore, the starting and ending wavelengths for wide-range gas scanning are set according to the current scene. For example, when acquiring combustible gas detection results, the wavelengths that the combustible gas can absorb after infrared light emission are used as the range. The wavelength ranges of all combustible gases that can appear in the current scene are marked, and then the starting and ending wavelengths are set. The identified wavelength range can be selected from pre-collected data in the database to determine the wavelength to be set for the currently measured combustible gas.

[0027] like Figure 2As shown, the data acquisition module also includes: obtaining the types of combustible gases in the current scene, querying the absorption spectral library based on the type of combustible gas, and determining the wavelength list of the gas to be measured. The absorption spectral library represents information such as the main absorption peak position, absorption intensity, linewidth, and transition probability of each gas. At this point, the wavelength of the absorption peak center of different combustible gases is mainly queried, and a wavelength list of all gas absorption peaks is output. Then, the initial wavelength and termination wavelength are set according to their absorption peak ranges to determine the main gas to be identified during laser scanning. For example, when identifying methane, its absorption peak center is 1653nm, and the absorption peak center of ethylene is 1620nm. The absorbance of these gases is most obvious at these wavelengths, and the concentration of the gas can be directly identified. The scanning wavelength queried at this time is generally extracted directly from the absorption peak center of different gases to indicate the relatively sensitive position of their absorbance and gas concentration.

[0028] Extract the absorption peak coverage range of the gas to be tested from the wavelength list, and use the upper and lower limits of the absorption peak coverage range as the start wavelength and end wavelength, respectively.

[0029] Determine the wavelength offset of the start and end wavelengths in the current scene, and set the step value based on the absorption peak range corresponding to the wavelength offset. Use the step value, start wavelength, and end wavelength as the output scanning wavelength range.

[0030] The wavelength offset here represents the impact of current temperature and pressure on the emitted laser wavelength. This value is based on a comparison between the wavelength values ​​measured at the current temperature and pressure and those measured at normal temperature and pressure to indicate the existing wavelength offset. Then, the absorption peaks corresponding to the wavelength offset for different combustible gases at the current temperature are identified. The width of these absorption peaks is then set with a step value to adjust the laser's measurement of combustible gases within the scanning wavelength range according to a specific step size. For example, the scanning range for methane is set to 1650nm-1660nm with a wavelength step of 0.1nm; this is used as the current output wavelength scanning range.

[0031] The initial step value is set to 0.05 nm to obtain the most detailed spectral data. Alternatively, the step value can be adjusted according to the temperature and pressure included in the current environmental parameters. The average step value used at the corresponding frequency of occurrence of the current temperature and pressure in historical data can be selected to divide the relevant absorbance.

[0032] The implementation method of setting the step value based on the absorption peak range corresponding to the wavelength offset includes: calculating the wavelength offset based on the temperature and pressure differences between two adjacent measurements of the gas to be measured. When measuring gas concentration using lasers, thermal expansion and Doppler broadening can easily cause a shift in the center wavelength of the identified absorption peak. In this case, the wavelength offset with respect to temperature is obtained by multiplying the material's coefficient of thermal expansion, the temperature difference, and the initial wavelength value. The coefficient of thermal expansion will be selected differently depending on the machine or environment; for example, the typical value for a DFB laser is approximately 0.1 nm / ℃.

[0033] The wavelength offset corresponding to the pressure difference indicates that the change in refractive index caused by the pressure change affects the effective optical path, resulting in a wavelength shift. The wavelength offset is calculated by multiplying the initial wavelength value, the pressure coefficient, and the pressure difference. The pressure coefficient is set according to typical pressure conditions, such as approximately 3 × 10⁻⁶ for air. -6 nm / Pa, at this point, these calculated initial wavelength values ​​are selected according to the different values ​​of the coefficients used to determine the wavelength offset under the current conditions; and the wavelength offsets corresponding to the temperature difference and pressure difference are added together to obtain the current wavelength offset. The initial wavelength value here represents the center wavelength of the absorption peak found in the wavelength list, that is, it represents the nominal center wavelength of the gas absorption peak. For example, under standard conditions, the center of the absorption peak range of methane is 1653 nm.

[0034] It should be noted that temperature and pressure are obtained through temperature sensors and pressure sensors, respectively.

[0035] Preferably, the obtained start wavelength and end wavelength are affected by the wavelength offset, that is, the start wavelength and end wavelength are adjusted by the wavelength offset, and the step value is selected according to the adjusted start wavelength and end wavelength.

[0036] Here, the linewidth corresponding to the wavelength offset is quantified using Doppler broadening to illustrate the temperature-dependent Doppler broadening after the absorption peak range is reached when a wavelength offset occurs. Represented as, ;in, The center frequency of the gas absorption line, measured in Hz, is the same as the nominal center wavelength. satisfy, ; This represents the Boltzmann constant as 1.38 × 10⁻⁶. -23 J / K; Indicates ambient temperature. Indicates the mass of gas molecules. This represents the speed of light. The value is then determined under the spectral propagation conditions corresponding to Doppler broadening, and subsequently converted into wavelength form. ;in, This represents the Doppler broadening in wavelength form, indicating how the absorption line broadening changes under temperature-dependent conditions. This yields the wavelength broadening under a wavelength offset. After determining the pressure-dependent broadening, the adjusted absorption peak range can be determined, ultimately revealing the values ​​of the starting and ending wavelengths.

[0037] As for pressure-related Doppler widening Represented as, ;in, Indicates the pressure broadening factor (unit: cm) -1 (e.g., methane in air typically has a concentration of 0.1 cm³ / atm). -1 / atm, This represents ambient pressure (unit: atm). Pressure broadening indicates that pressure changes cause molecular collisions, leading to broadening of the absorption lines and thus changes in the spectrum. It is then converted to wavelength form. The scanning bandwidth corresponding to the wavelength offset at this time can be expressed as: At this point, the relevant data of the absorption peak under the corresponding bias amount are obtained. Then, by adding or subtracting half of the scanning bandwidth from the adjusted wavelength bias amount, the starting wavelength and ending wavelength can be obtained, thus obtaining the current wavelength scanning range.

[0038] If the step value is always less than one-fifth of the scan bandwidth, then the scan bandwidth is divided by the number of scan points to obtain the step value.

[0039] Preferably, the absorbance described above... Represented as, ;in, This indicates the light intensity after passing through the gas being measured. The absorbance is proportional to the gas concentration, representing the incident light intensity. However, when the gas concentration changes, the curve showing the relationship between the measured gas concentration and absorbance will exhibit flickering points and multiple peak changes, leading to identification errors when matching the corresponding gas.

[0040] The processing method at this time uses a DFB laser driver. After covering the set wavelength range through current-temperature coordinated tuning, modulation technology is used to convert the emitted laser into spectral data, which is used to analyze the combustible gas identified at the corresponding wavelength and the corresponding concentration of the combustible gas. This part mainly uses a photodetector to convert the emitted laser signal into a corresponding electrical signal to determine the spectrum of the laser after identifying the gas.

[0041] In one embodiment of the present invention, the matching gas item is used to identify the gas form measured at a position such as the absorption peak represented by the current spectral point of the gas to be measured, and to further adjust the measured spectral point data according to the wavelength of the gas under theoretical conditions. Then, the linear relationship between the absorbance and the gas concentration represented by the corresponding spectral point is analyzed to find the specific concentration of the gas to be measured.

[0042] Preferably, the output matching gas item will indicate which gas the currently identified spectral point most likely represents as its absorption peak.

[0043] like Figure 3 As shown, the implementation of the spectral analysis module includes: sorting the matching items of spectral points to form a spectral linked list, and mapping the spectral points in a chain structure to illustrate the order of each spectral point in the current absorption spectrum.

[0044] Examine the spectral points in the spectral chain, divide the spectral points into multiple continuous bands according to the length of the absorption spectrum, and set an initial remaining bit width for each band.

[0045] At this point, the spectral points are divided into multiple bands according to the position of the corresponding absorption peaks in the spectrum. The wavelength range, mean absorbance, and standard deviation of each band are used as matching terms to match each band one by one. The initial remaining bit width indicates how much remaining matching length exists in the absorption spectrum, which means how many unmatched features remain after matching the current band.

[0046] The spectral points in each band are matched with the preset template to obtain the matching value of each spectral point. The matching value is calculated by comparing the data in the spectrum with the preset template using methods such as cosine similarity or Pearson correlation coefficient. The wavelength and absorbance corresponding to the spectral point are calculated with the data in the preset template, and the corresponding values ​​of the calculation results are output to obtain the matching value. At this time, data standardization processing is used to normalize each value before calculating the matching value.

[0047] When matching, each band will be matched with the preset template in turn according to the different matching items selected. The matching value of each spectral point will be mapped to the spectral list. The gas matching item of the current band will be identified by the matching value and position of each spectral point.

[0048] The methods for matching spectral points in each band with preset templates include: traversing the spectral list, determining whether the current band is locked, and skipping the current band if it is locked.

[0049] If no match is found, check if the width of the absorption peak in the current band is less than the initial remaining bit width. If it is less, match the current band; otherwise, continue traversing the next band.

[0050] When a data match is found in a band, the current band is locked, the initial remaining bit width of the current band is updated, and the position, matching value and corresponding type of the match are recorded.

[0051] When a match fails, release the current band and backtrack to the beginning of the spectral list to re-traverse the corresponding bands until all bands are matched.

[0052] The above-described spectral list manages all bands to be matched, with each band carrying a lock status indicator. The matching process first checks whether the current band is locked. Setting a lock is to prevent other matching tasks from occupying the current analysis process, which could lead to concurrent modifications and incorrect gas identification.

[0053] The restriction that the absorption peak width should be less than the remaining bit width is to determine whether a complete peak exists in the current band. Generally, the initial remaining bit width is greater than the absorption peak width. In this case, the matching process is only initiated when the absorption peak width is less than the remaining bit width to avoid matching failure due to insufficient space. Simultaneously, when updating the initial remaining bit width of a band, the remaining bit width is updated with the matched portion, ensuring that the remaining bit width set for each band represents unmatched data items. When the unmatched data item is less than the width of an absorption peak, the features obtained solely from this portion of data are severely lacking, and the absorption peak center may even be unidentifiable; therefore, the corresponding content is not processed.

[0054] The method for identifying gas matching terms in the current spectral band includes: performing cross-analysis on each spectral point to determine whether there are overlapping absorption peaks in each band; if so, determining the gas matching term for the current band by averaging the matching values ​​of the extreme slope points within the current band. In this case, the extreme slope points can help determine whether there are multiple absorption laser intensity components within the overlapping absorption peaks, and the gas corresponding to the average matching value is considered the currently detected gas.

[0055] If it does not exist, the gas matching item for the current band is determined by the matching value of the maximum absorbance in the current band. This means that if it is a single peak and there are no adjacent overlapping absorption peaks, the gas corresponding to the maximum value of the gas matching value can be directly used as the matching gas item output at this time to describe the gas currently identified.

[0056] In the case of overlapping absorption peaks, the first derivative of the absorption spectrum is calculated to obtain the derivative spectrum. Then, the extreme points are identified, and the presence of overlapping peaks is determined by the number of peaks in the derivative spectrum. If multiple adjacent peaks exist, such as in a bimodal structure, it is determined to be an overlapping absorption peak. If no overlapping peaks exist, it proceeds to the single-peak treatment.

[0057] In derivative spectroscopy, local maxima correspond to the inflection points of absorption peaks in the original spectrum. Identifying the slope extrema at these points can represent the absorption characteristics of different gases. For example, when the absorption peaks of CO2 and CH4 partially overlap, the derivative spectrum will show two extrema. The bands with overlapping absorption peaks are divided into multiple sub-intervals according to the slope extrema. Assuming each sub-interval corresponds to a potential gas component, the data of each sub-interval is matched with a preset template. Then, the matching value of the corresponding spectral point in this sub-interval is calculated. The matching values ​​in this sub-interval are then weighted and averaged to obtain the matching average value. The weighting can be based on the ratio of the slope value to the sum of the slope values ​​in the sub-interval. The calculation is performed sequentially for each sub-interval. Then, the gas type corresponding to the matching average value is used as the output matching gas item for the current band. The contents of the matching average value in the preset table in the database are checked to find the corresponding gas type.

[0058] For a single scene, the absorbance values ​​of its spectral points are iterated to find the point with the maximum absorbance and the gas type corresponding to the matching degree calculated by the preset template. If the matching degree is greater than 0.95, the corresponding gas is output. This completes the identification of the gas types that appear in the alignment. Then, by identifying the location where the corresponding gas type appears, the relevant information of the combustible gas measured in the current scene is determined.

[0059] In one embodiment of the present invention, during the location identification module, multiple scanned absorption spectra are acquired, the positions of the matching gas items in the spectra are marked, and the presence of phase differences between adjacent points is identified. The phase differences are then used to locate the coordinates of the leak point and the corresponding gas concentration.

[0060] At the same time, based on the absorbance value, the gas concentration needs to be calculated using the Lambert-Beer law. By matching the relative linear relationship between the gas concentration at the location of the gas term and the adjacent points, it can be determined whether there is a correlation between the current leak point and the gas concentration.

[0061] The absorbance and gas concentration are used as analytical conditions to identify the gas concentration measured at each absorbance, forming an absorbance-gas curve. The absorbance is used as the abscissa and the gas concentration as the ordinate to represent the trend of measurement at different absorbances. The trend values ​​of each point in the trend are used to view the matching gas item of the current spectral point to obtain the relationship between the coordinates of the leak point and the gas concentration under the corresponding description.

[0062] The gas concentration and absorbance are directly proportional, such as ;in, Indicates gas concentration. This represents the molar absorptivity of a gas, used to indicate how a gas absorbs light. The optical path length represents the distance the laser travels from emission to reception. After identifying the matching gas, the corresponding gas concentration can be retrieved based on the absorbance value at the corresponding wavelength. Then, by analyzing its trend and phase difference, the presence of a leak at the identified combustible gas location can be determined.

[0063] Preferably, at this time, absorption spectra at multiple locations are obtained using a gimbal-type linear beam, and each spectral point in the absorption spectra is labeled with its wavelength, absorbance, gimbal coordinates, and matching gas.

[0064] like Figure 4 As shown, the implementation of the positioning and identification module includes: calculating the phase difference between the spectral points labeled as the same matching gas item and their 8 neighboring nodes, setting the phase difference matrix corresponding to each spectral point, and recording the phase vector direction corresponding to each spectral point.

[0065] By using the absorbance corresponding to each spectral point, the gas concentration measured at each absorbance is identified, and an absorbance gas curve is formed. By using the trend value of each point in the change trend of the absorbance gas curve, the matching gas term of each spectral point is viewed. Using the absorbance corresponding to each spectral point, the gas concentration measured at each absorbance is identified to form a concentration matrix, and the concentration gradient corresponding to each spectral point is obtained.

[0066] Based on the phase vector direction and concentration gradient corresponding to each spectral point, multiple conditions are determined, and data that meet all conditions are marked as leakage points.

[0067] Preferably, when setting the phase difference matrix as described above, the phase of the data corresponding to each spectral point is extracted by performing a Fourier transform. Then, the phase difference between this point and the nodes in the eight adjacent directions is compared. After combining these points according to their relative positions into a phase difference matrix, the direction and intensity of the phase difference can be directly known from the phase difference value. The direction of the phase vector can be indicated by combining the phase differences in the eight directions in the calculated phase difference matrix into a gradient component.

[0068] At the same time, when setting its phase difference matrix, its phase unwrapping and neighborhood points are also standardized to prevent discontinuities in the phase difference interval. Then, the neighborhood points are standardized to prevent the problem of inconsistent dimensions caused by missing matrices.

[0069] Since the phase vector direction is obtained by combining the phase differences of 8 neighboring nodes, it is necessary to identify the gradients in the horizontal and vertical directions. The horizontal gradient is calculated by subtracting the phase difference between the current spectral point and the eastern point from its phase difference with the western point, then subtracting the phase differences in the northwest and southwest directions from the phase differences in the northeast and southeast directions, and summing these two sets of phase differences. Similarly, the vertical gradient is obtained by subtracting the phase difference in the north from the south, and adding the phase differences in the southeast and southwest, and the phase differences in the northeast and northwest directions. Finally, the arctangent function is used, with the ratio of the horizontal to the vertical gradient as input, to obtain the angle of the current phase vector direction. It should be noted that the eight directions mentioned here are used to describe the east, west, etc., directions after establishing a standard plane coordinate system by displaying the current measurement area in a planar form.

[0070] Preferably, when examining the matching gas term for each spectral point using the trend values ​​at each point in the changing trend of the absorbance gas curve, the linear equation corresponding to the current curve is fitted using the values ​​of the absorbance gas curve and the calculation method of the coefficient of determination, that is, letting... ;in, This represents the slope of the gas absorption curve. This represents the intercept term. After fitting, if the coefficient of determination is greater than 0.95, it is considered that there is a linear relationship between the currently retrieved gas concentration and absorbance, and then the process of aligning the concentration gradient is triggered. Then, the concentration gradient at the location of each spectral point is identified. When identifying the concentration gradient, the gradient components in the horizontal and vertical directions are directly obtained, and the direction angle is obtained in the same way as the phase vector direction, which facilitates the subsequent determination of the leak point it points to.

[0071] Preferably, when making multi-condition judgments based on the phase vector direction and concentration gradient corresponding to each spectral point, the implementation method includes: using the judgment methods of phase difference convergence, concentration gradient matching and trend curve linearity, when there are candidate leakage points that all meet the corresponding judgment methods, the corresponding candidate leakage point is output as the leakage point.

[0072] Phase difference convergence marks the convergence point of the phase vector direction corresponding to each spectral point as the candidate leak coordinate. When the phase vector direction pointing to the candidate leak coordinate is greater than a preset direction threshold, the current candidate leak coordinate is regarded as a candidate leak point. At this time, the preset direction threshold is set according to the number of arrows represented by the phase vector direction. For example, the candidate leak coordinate is considered as a candidate leak point only when the number of arrows is greater than or equal to 3. Alternatively, the average number of arrows represented by the phase vector direction when a leak point is identified in historical data can be used as the basis for screening candidate leak points at this time.

[0073] Concentration gradient matching compares the angle between the concentration gradient direction of a candidate leak point and the phase vector direction. If the angle is less than the included angle threshold, the candidate leak point is considered valid. The included angle threshold is used to determine whether the concentration gradient direction and the phase vector direction are consistent. That is, the included angle threshold is set to 15°, or the average value of the concentration gradient direction and the phase vector direction in historical data is used for identification.

[0074] The linearity of the trend curve is determined by verifying the coefficient of determination of the light-absorbing gas curve. When the coefficient of determination is greater than 0.95, the current candidate leak point is considered valid.

[0075] In one embodiment of the present invention, the trend related to the state change at each leak point is determined by taking the gas concentration at the leak point as its state and correcting it with the state probability, and the change value of its state probability is taken as the current output change trend.

[0076] like Figure 5 As shown, the implementation of the trend recognition module includes: using the gas concentration and concentration change rate of each leak point as state variables, and using the rise time, fall time and plateau time of gas concentration change as trend morphology variables, to obtain the state probability of each leak point.

[0077] By associating and mapping each leak point with the data proportion corresponding to the state probability of each leak point, and classifying the leak points according to the trend after mapping, the response strategy for each leak point is set.

[0078] Preferably, the rise time, fall time, and plateau time of gas concentration change are determined by the values ​​of gas concentration and concentration change rate. Then, when constructing the state probability, the change of the current trend morphology variable in any form of rise, fall, or plateau within the corresponding time period is taken as one condition, and the change of gas concentration to another gas concentration within the corresponding time period is taken as another condition. The state probability of the current leak point is expressed in the form of conditional probability.

[0079] Preferably, the rise time represents the length of time during which the concentration change rate is greater than the concentration change rate threshold, indicating an increase in the gas concentration at the leak point. The length of time during which the concentration change rate is less than the negative value of the concentration change rate threshold is considered the fall time, and the length of time during which the absolute value of the concentration change rate is less than the concentration change rate threshold is considered the plateau time. In this case, the concentration change rate threshold can be 0.05 ppm / s, or the average value of the thresholds used to judge the rise and fall trends in historical data.

[0080] After obtaining the value of its state probability, the data with the same state probability value are compared with all the data collected so far to determine the change stage experienced by the current leak point. Then, the state probability and data ratio are mapped to the leak point, and each leak point is classified. The classified label is used as the output trend classification.

[0081] When obtaining the trend classification, the state probability and data proportion corresponding to each leakage point are used to form a feature vector. The feature vector is then standardized and used as a sample point. Starting from any unvisited sample point, all points within its neighborhood radius are searched. When the number of samples within the neighborhood radius is greater than the minimum number of neighborhood points, a new cluster is created, and the point and its neighborhood points are marked as the same class. All sample points are processed recursively until no more points can be expanded. This process is repeated until all sample points corresponding to the data processing of all leakage points have been accessed.

[0082] At this point, the neighborhood radius is calculated using Euclidean distance to measure the distances between all sample point pairs. After sorting the samples in ascending order, a curve representing the clustering and cumulative percentage is plotted. The point where the slope abruptly changes is selected as the current neighborhood radius. The cumulative percentage represents the number of sample point pairs existing at the corresponding distance relative to the total distance of all sample point pairs. The calculated distances between sample point pairs are obtained using Euclidean distance. The minimum neighborhood number is set based on twice the dimension of the feature vector corresponding to the current sample point. For example, if the selected leak point uses the proportion of its rise time, fall time, and plateau time as the feature dimension, the minimum neighborhood number would be 6, indicating a relatively stable clustering situation. The trend classification of the corresponding leak point during clustering is then output. Based on the response strategy set in the database according to the trend classification, the response strategy for the corresponding leak point is selected. For example, if the trend classification is active leak, the response strategy would be emergency shutdown + evacuation. After completing the trend classification, multiple sets of values ​​corresponding to the current leak point are examined to determine the current response strategy.

[0083] like Figure 6 As shown, the present invention also provides an electronic device, including: at least one processor, at least one memory, a communication interface, and a bus; wherein the processor, memory, and communication interface communicate with each other through the bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to implement the system as described above.

[0084] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered within the protection scope of the present invention.

Claims

1. A pan-tilt linear beam combustible gas detection system, characterized in that, include: The data acquisition module is used to acquire the starting and ending wavelengths of the gas to be tested in the current scene, and determine the wavelength scanning range of the gas to be tested; based on the wavelength scanning range of the gas to be tested, an absorption spectrum is formed using the absorbance values ​​at each point within the wavelength scanning range. The spectral analysis module is used to extract features from the spectral points of the absorption spectrum, extract the matching items for each spectral point one by one, and identify the matching gas items for each spectral point in the absorption spectrum according to the matching value of each matching item. The location identification module is used to locate the coordinates of the leak point based on the phase difference of the matching gas terms at adjacent points of each spectral point and the linear relationship between absorbance and gas concentration at the leak point. The trend recognition module is used to invert the gas concentration at the leak point and determine the changing trend of the gas concentration at each leak point; Based on the changing trends of gas concentration at each leak point, determine the response strategy for each leak point; The implementation of the location identification module includes: calculating the phase difference between spectral points labeled as matching gas items and their 8 neighboring nodes, setting a phase difference matrix corresponding to each spectral point, and recording the phase vector direction corresponding to each spectral point; identifying the gas concentration measured at each absorbance using the absorbance corresponding to each spectral point to form an absorbance gas curve; examining the trend values ​​of each point in the trend of the absorbance gas curve to view the matching gas items for each spectral point, identifying the gas concentration measured at each absorbance using the absorbance corresponding to each spectral point to form a concentration matrix, and obtaining the concentration gradient corresponding to each spectral point; performing multi-condition judgment based on the phase vector direction and concentration gradient corresponding to each spectral point, and marking the data that meets all conditions as leak points. When performing multi-condition judgment based on the phase vector direction and concentration gradient corresponding to each spectral point, the implementation methods include: using phase difference convergence, concentration gradient matching, and trend curve linearity judgment methods, when all candidate leak points meet the corresponding judgment methods, the corresponding candidate leak point is output as a leak point; phase difference convergence marks the convergence point of the phase vector direction corresponding to each spectral point as the candidate leak coordinate; when the phase vector direction pointing to the candidate leak coordinate is greater than a preset direction threshold, the current candidate leak coordinate is regarded as a candidate leak point; concentration gradient matching compares the angle between the concentration gradient direction of the candidate leak point and the phase vector direction, and when it is less than the included angle threshold, the candidate leak point is considered valid; trend curve linearity obtains the determination coefficient by verifying the absorbance gas curve, and when the determination coefficient is greater than 0.95, the current candidate leak point is considered valid.

2. The pan-tilt linear beam combustible gas detection system according to claim 1, characterized in that, The implementation methods of the data acquisition module also include: Obtain the types of combustible gases in the current scene, query the absorption spectral library based on the types of combustible gases, and determine the wavelength list of the gas to be tested; Extract the absorption peak coverage range of the gas to be tested from the wavelength list, and use the upper and lower limits of the absorption peak coverage range as the start wavelength and end wavelength, respectively. Determine the wavelength offset of the start and end wavelengths in the current scene, and set the step value based on the absorption peak range corresponding to the wavelength offset. Use the step value, start wavelength, and end wavelength as the output scanning wavelength range.

3. The pan-tilt linear beam combustible gas detection system according to claim 2, characterized in that, The implementation methods for setting the step value based on the absorption peak range corresponding to the wavelength offset include: The wavelength offset with respect to temperature is obtained by multiplying the material expansion coefficient, temperature difference, and initial wavelength value by the temperature difference and pressure difference between two adjacent measurements of the gas to be measured; the wavelength offset with respect to pressure is obtained by multiplying the initial wavelength value, pressure coefficient, and pressure difference; the wavelength offsets corresponding to temperature difference and pressure difference are added together to obtain the wavelength offset currently in use. The starting and ending wavelengths are adjusted by wavelength offset, and the step value is selected based on the adjusted starting and ending wavelengths.

4. The pan-tilt linear beam combustible gas detection system according to claim 1, characterized in that, The implementation methods of the spectral analysis module include: Sort the matching items of the spectral points to form a spectral linked list; Examine the spectral points in the spectral chain, divide the spectral points into multiple continuous bands according to the length of the absorption spectrum, and set an initial remaining bit width for each band; Match the spectral points in each band with the preset template to obtain the matching value for each spectral point; The matching value of each spectral point is mapped to a spectral list, and the matching gas item for the current band is identified based on the matching value and position of each spectral point.

5. A pan-tilt linear beam combustible gas detection system according to claim 4, characterized in that, The methods for matching spectral points in each band with preset templates include: Traverse the spectral list and determine if the current band is locked. If it is locked, skip the current band. If no match is found, check if the width of the absorption peak in the current band is less than the initial remaining bit width. If it is less, match the current band; otherwise, continue to traverse the next band. When a data match is successful in a band, the current band is locked, the initial remaining bit width of the current band is updated, and the position, matching value and corresponding type of the match are recorded. When a match fails, release the current band and backtrack to the beginning of the spectral list to re-traverse the corresponding bands until all bands are matched.

6. A pan-tilt linear beam combustible gas detection system according to claim 4, characterized in that, Identifying the matching gas term for the current band can be achieved through the following methods: Cross-analysis is performed on each spectral point to determine whether there are overlapping absorption peaks in each band. If they exist, the matching gas term for the current band is determined by the average value of the matching slope extreme points in the current band; if they do not exist, the matching gas term for the current band is determined by the matching value of the maximum absorbance in the current band.

7. A pan-tilt linear beam combustible gas detection system according to claim 1, characterized in that, The implementation methods of the trend recognition module include: Using the gas concentration and concentration change rate at each leak point as state variables, and the rise time, fall time and plateau time of gas concentration change as trend morphology variables, the state probability of each leak point is calculated. By associating and mapping each leak point with the data proportion corresponding to the state probability of each leak point, and classifying the leak points according to the trend after mapping, the response strategy for each leak point is set.

8. An electronic device, characterized in that, include: The system includes at least one processor, at least one memory, a communication interface, and a bus; wherein the processor, memory, and communication interface communicate with each other via the bus. The memory stores program instructions that can be executed by the processor, which invokes the program instructions to implement the functions of the system as described in any one of claims 1 to 7.

Images