A high-precision greenhouse gas concentration observation system test method

Abstract

The application provides a kind of high-precision greenhouse gas concentration observation system test method, belongs to concentration observation technical field, comprising: according to the system structure of concentration observation system, the test scheme based on different test performance is configured to concentration observation system;According to the test scheme, the deployment of measuring medium is carried out to concentration observation system, and the performance test is carried out to the corresponding test target based on the deployed measuring medium, and the corresponding performance matrix is constructed;Each performance matrix is analyzed respectively, and the abnormal level label is set to the corresponding performance matrix in combination with the performance standard of corresponding test performance;The warning report of concentration observation system is generated based on all abnormal level labels to remind. Through the test scheme under multiple test performances, the device deployment is carried out to concentration observation system, the performance test in multiple aspects is realized, and then through the matrix analysis and label setting, the warning report for system is output, to ensure that the system performance meets the required observation requirements and ensures its reliability.

Description

Technical Field

[0001] The present invention relates to the field of concentration observation technology, and in particular to a testing method for a high-precision greenhouse gas concentration observation system. Background Art

[0002] Currently, high-precision ground-based greenhouse gas (CO2 / CH4) concentration measurement methods include cavity ring-down spectroscopy (CRDS), off-axis integrated cavity output spectroscopy (OA-ICOS), optical feedback-cavity-enhanced non-dispersive infrared absorption (NDIR), Fourier transform infrared spectroscopy (FTIR), and gas chromatography (GC). However, there are no specific requirements for whether the observation instruments and observation systems used in these measurement methods meet the requirements for high-precision observation, resulting in varying degrees of accuracy. Observation instruments and observation systems must undergo rigorous performance testing before they can be put into operational use by meteorological departments.

[0003] Therefore, the present invention proposes a testing method for a high-precision greenhouse gas concentration observation system. Summary of the Invention

[0004] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, which is used to deploy components of the concentration observation system through test schemes under multiple test performances, realize multi-faceted performance testing, and then output an early warning report for the system through matrix analysis and label setting to ensure that the system performance meets the required observation requirements and ensure its reliability.

[0005] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, comprising:

[0006] Step 1: configuring a test scheme based on different test performances to the concentration observation system according to the system structure of the concentration observation system;

[0007] Step 2: Deploy the measurement medium of the concentration observation system according to the test plan, and perform performance tests on the corresponding test targets based on the deployed measurement medium to construct a corresponding performance matrix;

[0008] Step 3: Analyze each performance matrix separately, and set an abnormal level label for the corresponding performance matrix based on the performance standard of the corresponding test performance;

[0009] Step 4: Generate an early warning report for the concentration observation system based on all abnormal level labels to provide reminders.

[0010] Preferably, configuration of test solutions based on different test performances includes:

[0011] Decompose the system performance requirements to obtain several test requirements;

[0012] Perform test structure mapping analysis on each test requirement and the system structure, and lock the test target based on the test requirement;

[0013] Based on the deployment location of the test target in the system structure, the target distribution based on the corresponding test requirements is obtained;

[0014] Based on the distribution-test comparison table, a test plan that matches the target distribution is retrieved.

[0015] Preferably, deploying the measurement medium of the concentration observation system according to the test plan includes:

[0016] Extracting respectively from the test plan the deployment information of the measurement medium based on each test target in the corresponding target distribution situation;

[0017] The corresponding measurement medium is deployed according to the deployment information.

[0018] Preferably, a corresponding performance matrix is ​​constructed, including:

[0019] Obtaining the latest maintenance parameters matching each test target from a historical database, and adjusting the identification process of the corresponding test target under the corresponding test performance according to the maintenance status of the latest maintenance parameters;

[0020]

[0021] Among them, Z y The final setting standard for the identification process; C y Initially set the criteria for the identification process; y is the accuracy loss factor when the maintenance status is the overall replacement status; For nth i1 Precision loss factor under maintenance; γ i1 is the precision loss factor under the i1th maintenance; γ max For all γ i1 The maximum value in ; For all γ i1 Variance of n i1 The number of repairs involved in maintaining the target state of repair;

[0022] Determine the number of acquisitions for the corresponding test target based on the preliminary and final set standards of the identification process;

[0023]

[0024] Among them, Δ y is the corresponding adjustment unit; Tc y is the set original acquisition period; Ty a collection period corresponding to a determined test target; a ceiling symbol; a floor symbol; Tz is a total running time of the concentration observation system; N y a collection number corresponding to a test performance;

[0025] obtaining an average number of collection numbers of all test targets under a corresponding test performance satisfying a preset normal distribution probability, to obtain a final collection period;

[0026] controlling the measurement medium after deployment to test the corresponding test target according to the final collection period, to construct a performance matrix.

[0027] Preferably, each performance matrix is analyzed respectively, including:

[0028] performing row analysis and column analysis on the corresponding performance matrix, to obtain a first analysis value of each row vector and a second analysis value of each column;

[0029] obtaining an initial function of the corresponding performance matrix based on all first analysis values and second analysis values;

[0030] determining a test precision of the corresponding test performance according to the number of targets and the type of test involved under the corresponding test performance, and obtaining a final analysis value and an analysis error of the corresponding performance matrix in combination with the initial function.

[0031] Preferably, in combination with a performance standard of the corresponding test performance, a level label is set to the corresponding performance matrix, including:

[0032] constructing a performance vector based on the performance standard of the corresponding test performance, and sequentially subtracting each row vector in the performance matrix under the corresponding test performance, to obtain a subtraction matrix;

[0033] performing cluster analysis on each column in the subtraction matrix to obtain a cluster set of the corresponding test target, and drawing a distribution graph based on each cluster set;

[0034] determining a distribution dimension in the distribution graph and setting an abnormal probability to the corresponding test target based on the number of distributions under each distribution dimension;

[0035] regarding the test target with an abnormal probability greater than a preset probability as an abnormal target;

[0036] locking a prominent element of each row in the subtraction matrix, and setting an elimination probability to the corresponding row in combination with the final analysis value and the analysis error;

[0037] setting a dependency relationship to the corresponding abnormal target based on the deployment position of each abnormal target in the system structure;

[0038] Based on the dependency, maintenance status, and abnormal target, based on abnormal probability, elimination probability, and abnormal characteristics under target characteristics, an abnormal symbolic expression for the abnormal target is generated, and matched with the expression-abnormality comparison table to obtain an initial abnormality;

[0039] Based on all initial anomalies and the column variance of the column vector corresponding to each initial anomaly, a first label is set to the corresponding anomaly target, and in combination with the test weight of the anomaly target, an anomaly level label is set to the corresponding performance matrix.

[0040] Preferably, the method further includes: determining abnormal features of the abnormal target based on the abnormal probability, elimination probability, and target characteristics, specifically including:

[0041] Obtaining a column vector corresponding to an abnormal target from the subtraction matrix, performing curve drawing and curve fitting, and locking discrete points;

[0042] A discrete function is constructed based on the discrete gradient of each discrete point. At the same time, a probability function is constructed based on the abnormal probability and the elimination probability.

[0043] The discrete function and the probability function are input into a function analysis model that matches the corresponding target characteristics to obtain abnormal characteristics.

[0044] Preferably, an early warning report of the concentration observation system is generated based on all abnormal level labels, including:

[0045] Based on each abnormal level label, the deployment point of the concentration observation system is rendered using a significant color consistent with the corresponding abnormal level label to obtain a corresponding first deployment map;

[0046] Obtain the performance topology of each first deployment diagram, and align all first deployment diagrams to obtain the abnormal correlation topology of each deployment point;

[0047] Generate an early warning report based on the anomaly correlation of all its own performance topology structures and different anomaly level labels, and combined with the anomaly correlation topology structure of each deployment point.

[0048] Preferably, the test performance includes: air tightness test based on concentration observation system, sample gas flow test, retention time test, dehumidification performance test, cold trap tube switching impact test, system performance test, greenhouse gas concentration linearity test, greenhouse gas concentration drift test and greenhouse gas deviation test.

[0049] Compared with the prior art, the present invention has the following advantages:

[0050] Through test schemes under various test performances, the concentration observation system is deployed to implement multi-faceted performance testing. Then, through matrix analysis and label setting, an early warning report for the system is output to ensure that the system performance meets the required observation requirements and ensures its reliability.

[0051] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0052] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0054] Figure 1 This is a flow chart of a testing method for a high-precision greenhouse gas concentration observation system in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0056] The present invention provides a method for testing a high-precision greenhouse gas concentration observation system. Figure 1 Shown, including:

[0057] Step 1: configuring a test scheme based on different test performances to the concentration observation system according to the system structure of the concentration observation system;

[0058] Step 2: Deploy the measurement medium of the concentration observation system according to the test plan, and perform performance tests on the corresponding test targets based on the deployed measurement medium to construct a corresponding performance matrix;

[0059] Step 3: Analyze each performance matrix separately, and set an abnormal level label for the corresponding performance matrix based on the performance standard of the corresponding test performance;

[0060] Step 4: Generate an early warning report for the concentration observation system based on all abnormal level labels to provide reminders.

[0061] Preferably, the test performance includes: air tightness test based on concentration observation system, sample gas flow test, retention time test, water removal performance test, full-automatic cold trap switching impact test, system performance test (overall performance test, equipment linearity test, equipment drift test, repeatability test and accuracy test).

[0062] In this embodiment, after the installation of the ground high-precision greenhouse gas (CO2 / CH4) concentration observation system is completed, the observation system needs to be tested on-site, and then steps 1 to 4 are used to implement the subsequent steps. It should be noted that the standard gas used in the test has been balanced and allowed to stand, and the concentration should be stable and the water vapor content should be less than 5ppm.

[0063] In this embodiment, the performance test is as follows:

[0064] 1. Air tightness test:

[0065] After the observation system is installed, a sample air tightness test should be carried out to check that there is no air leakage in the pipelines, joints, connectors, etc. where the sample gas flows through the observation system.

[0066] (1) Standard gas connection pipeline: Use soap bubble leak detection liquid to check for leaks at the joints of the cylinder, pressure reducing valve, gas line, pressure reducing gauge, etc. If obvious bubbles appear, it indicates that there is a leak at that place and the connection needs to be reinstalled.

[0067] (2) Analyzer connection pipeline: Use high-concentration carbon dioxide to check the overall air tightness of the analyzer. Spray it on the analyzer's connection joints and chassis vents. If the carbon dioxide concentration at the analyzer end does not increase abnormally, it indicates that the analyzer and its joints are in good overall air tightness.

[0068] (3) Pipeline connecting the sampling port to the vacuum pump: Inspection of the negative pressure section of the pipeline between the sampling port and the vacuum pump. For the sampling port on the tower, use a telescope to visually check that the air inlet filter on the tower is firmly installed and has no signs of loosening, and the air inlet pipe and the tower body are firmly fixed; for the pipeline below the tower, use a high-concentration carbon dioxide spraying device to spray around the pipeline and at the joints. If the carbon dioxide concentration at the analyzer end does not increase abnormally, it indicates that the pipeline is airtight.

[0069] (4) Connection line from the vacuum pump to the analyzer: Check the positive pressure section of the pipeline from the vacuum pump to the air inlet of the analyzer main unit and the components through which the sample flows. Use a high-concentration carbon dioxide spray device to spray high-concentration carbon dioxide around the vacuum pump to check the air tightness of the vacuum pump; use a special soap bubble leak detection liquid to check the air tightness of the vacuum pump outlet and joints, as well as the joints on the air path. Observe that there is no abnormal increase in the carbon dioxide concentration at the analyzer end or no bubbles appear at the joints, which indicates that the air path is airtight.

[0070] (5) Concealed interfaces / valve port connection pipelines: Open the outer panels of the cold trap instrument, valve box and other equipment, and check whether all gas pipeline interfaces, condenser pipe interfaces, valve port pipeline interfaces, etc. inside the equipment are intact and free of cracks; use a high-concentration carbon dioxide spray device or a special leak detection liquid to spray or apply it on the visible pipe joints inside the water removal equipment and valve box, and the valve ports of the selector valves. If the carbon dioxide concentration at the analyzer end does not increase abnormally or there are no bubbles at the joints, it indicates that the airtightness of the concealed interfaces is good.

[0071] 2. Sample gas flow test:

[0072] (1) Set different injection flow rates, with the flow rate range between 200-800 mL / min, and take values ​​in intervals of 50 mL / min.

[0073] (2) For different injection flow rates, the standard gas is introduced for 5 minutes each time, and the standard deviation of the data in the last 3 minutes is calculated. When the flow rate continues to increase and reaches a certain set value, and the standard deviation corresponding to the flow rate basically no longer changes and tends to be stable, it is used as the injection flow rate of the observation system.

[0074] 3. Retention time test:

[0075] (1) The retention time from the air inlet to the analyzer host can be obtained by spraying high-concentration CO2 around the air inlet and observing the peak time of CO2 concentration in the analyzer with a stopwatch. It should not exceed 2 minutes.

[0076] (2) The retention time of the standard gas to the analyzer main unit can be achieved by switching the valve position to the standard gas sampling gas path and observing the CO2 concentration peak time of the analyzer with a stopwatch. It should not exceed 1 minute.

[0077] Water removal performance test:

[0078] (1) The water removal performance test is performed under meteorological conditions where the relative humidity in the ambient air is greater than 80%.

[0079] (2) Adjust the observation system and water removal equipment to normal measurement status, observe and analyze the water vapor content measured by the host, and the water vapor content in the air sample after water removal should be less than 40ppm.

[0080] 4. Automatic cold trap tube switching impact test:

[0081] (1) Set the injection sequence to sample gas injection to avoid valve position switching when switching the cold trap tube.

[0082] (2) Set the automatic switching time of the cold trap tube to the shortest interval for testing, and complete at least 3-5 cold trap tube switches.

[0083] (3) Select the H2O, CO2 and CH4 second-level concentration data corresponding to the cold trap switching period, the time span for the H2O, CO2 or CH4 concentration to recover to the concentration level before the cold trap switching should not exceed 1 minute.

[0084] 5. System performance test:

[0085] 5.1 Overall performance test

[0086] (1) After the sampling pump is removed, the outlet joint is connected to the target standard gas. In order to avoid waste of standard gas, the bypass exhaust port on the back end of the sampling pump gas path needs to be closed.

[0087] (2) Set the sample sequence to alternately measure the working standard gas and the target standard gas. The measurement time for each bottle of standard gas is 5 minutes. Each bottle of standard gas is measured alternately for 3 times as a cycle. Test not less than 3 cycles.

[0088] (3) Take the average of the minimum standard deviation of the continuous 3 minutes of data of the working standard gas and the target standard gas as the measurement result of each. Use the working standard gas measurement result and the nominal concentration to construct a linear regression equation to calculate the actual concentration of CO2 and CH4 in the target standard gas.

[0089] (4) According to the deviation of the actual concentration of CO2 and CH4 in the target standard gas from the nominal concentration, evaluate whether the overall performance meets the requirements. For clean areas, the CO2 and CH4 deviations should be less than 0.1 ppm and 2 ppb respectively; for polluted areas, the CO2 and CH4 deviations should be less than 0.2 ppm and CH4≤5 ppb respectively.

[0090] 5.2 Linearity test:

[0091] (1) Use 5 bottles of concentration gradient standard gas covering the concentration fluctuation range of atmospheric greenhouse gases CO2 and CH4 to construct different concentration standard gas sequences for linearity test.

[0092] (2) By setting the sample sequence, 5 bottles of gradient standard gas enter the measurement unit in turn. Each bottle of standard gas is measured for 5 minutes, and is measured alternately for 3 times.

[0093] (3) Take the average of the minimum standard deviation of the continuous 3 minutes of data of each bottle of standard gas as the average of each measurement. Take the arithmetic mean of the average of 3 measurements as the measurement result of each bottle of standard gas.

[0094] (4) Linear fitting is performed on the CO2 and CH4 measurement results of the 5 bottles of standard gas and their corresponding nominal concentrations. The goodness of fit of CO2 and CH4 should be greater than 0.999, and the fitting residuals should be uniformly distributed without obvious outliers.

[0095] 5.3 Reproducibility test:

[0096] (1) Under the same measurement conditions, continuous measurement shall be carried out using standard gas for a measurement time of not less than 10 hours.

[0097] (2) After the instrument reading stabilizes, record the instrument output value (display value) and calculate the standard deviation every 5 minutes.

[0098] (3) Evaluate whether the instrument accuracy meets the requirements based on the average value of the standard deviation. For clean areas, CO2 < 0.1ppm, CH4 < 2ppb; for polluted areas, CO2 < 0.2ppm, CH4 < 5ppm.

[0099] 5.4 Drift test:

[0100] (1) Under the same measurement conditions, continuous measurement shall be carried out using standard gas for a measurement period of not less than 24 hours.

[0101] (2) Eliminate the measurement data of the first hour and use the measurement data of the remaining instruments with stable readings as valid test data. Calculate the average value of each hour as the hourly measurement result.

[0102] (3) In the valid test data, the deviation between the maximum and minimum values ​​of the instrument's hourly measurement results should meet the following requirements: for clean areas, CO2 < 0.1ppm, CH4 < 2ppb; for polluted areas, CO2 < 0.2ppm, CH4 < 5ppm.

[0103] 5.5 Accuracy test:

[0104] (1) The working standard gas (high concentration standard gas and low concentration standard gas) and five bottles of concentration gradient standard gas covering the fluctuation range of greenhouse gas concentration in the atmosphere are connected to the observation system at the same time.

[0105] (2) Alternately measure the working standard gas and 5 bottles of gradient standard gas. The measurement time for each bottle of standard gas is 5 minutes. Three alternating measurements constitute one cycle, and the test should be carried out for no less than 3 cycles.

[0106] (3) Take the mean of the continuous 3-minute data with the smallest standard deviation of the working standard gas and the gradient standard gas as the measurement mean, and take the arithmetic mean of the three measurement means of each bottle of standard gas as the measurement result of one cycle.

[0107] (4) The actual concentrations of CO2 and CH4 in the gradient standard gas of each cycle are calculated by linear fitting the measurement results of the working standard gas of each cycle with its nominal concentration.

[0108] (5) Evaluate whether the accuracy test meets the requirements based on the deviation between the actual concentration of CO2 and CH4 and the nominal concentration of all gradient standard gases. For clean areas, CO2 < 0.1ppm, CH4 < 2ppb; for polluted areas, CO2 < 0.2ppm, CH4 < 5ppm.

[0109] Generally speaking, when all results meet the observation requirements, the instrument is judged to have passed the acceptance test and can enter the meteorological department's operational observation.

[0110] In this embodiment, there are multiple measurement results for each test performance, and each measurement result constitutes a row vector. The multiple measurement structures constitute a matrix, that is, a performance matrix.

[0111] In this embodiment, the system structure is intuitively obtained based on the configuration diagram of the concentration observation system.

[0112] In this embodiment, the test plan is mainly to clarify which tests need to be performed on the system and which parts need to be tested under the corresponding test performance, so as to facilitate deployment and installation and achieve test acquisition of relevant parameters.

[0113] In this embodiment, the matrix analysis includes analysis of rows and analysis of columns.

[0114] In this embodiment, the performance standard is a pre-set performance requirement. For example, the test concentration standard of the measurement medium 1 for the test target 01 should be b.

[0115] In this embodiment, each row in the performance matrix represents all correlation values ​​measured once, and there are N measurements, where N is a variable.

[0116] In this embodiment, the abnormality level label is another form of reflecting the abnormality of the matrix.

[0117] In this embodiment, the early warning report is a reasonable analysis of all abnormal level tags to facilitate subsequent abnormality maintenance.

[0118] The beneficial effects of the above technical solution are: through test schemes under various test performances, the concentration observation system is deployed with devices to achieve multi-faceted performance testing, and then through matrix analysis and label setting, an early warning report for the system is output to ensure that the system performance meets the required observation requirements and ensures its reliability.

[0119] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, which configures a testing scheme based on different testing performances, including:

[0120] Decompose the system performance requirements to obtain several test requirements;

[0121] Perform test structure mapping analysis on each test requirement and the system structure, and lock the test target based on the test requirement;

[0122] Based on the deployment location of the test target in the system structure, the target distribution based on the corresponding test requirements is obtained;

[0123] Based on the distribution-test correspondence table, a test scheme matching the target distribution condition is called.

[0124] In this embodiment, the system performance requirement refers to the overall requirement of the concentration observation system, including air tightness requirement, flow requirement, time requirement, deviation requirement, drift requirement, linearity requirement, etc., so that a plurality of test requirements can be obtained by direct disassembly.

[0125] In this embodiment, the test structure mapping analysis refers to the test of some parts of the concentration observation system corresponding to the test requirement, such as air tightness test - joint of each connecting pipeline (some parts), and the target distribution condition is the deployment condition of the position of the joint.

[0126] In this embodiment, the distribution-test correspondence table contains different target distribution conditions and the test scheme corresponding to the distribution, and the test scheme is, for example, detecting whether the gas at the corresponding part leaks or not.

[0127] The beneficial effects of the above technical solutions are: through requirement disassembly and structure mapping analysis, the test target is effectively locked, and then the test scheme is called from the correspondence table, providing a basis for subsequent construction of the performance matrix.

[0128] The present application provides a kind of high-precision greenhouse gas concentration observation system test method, according to the test scheme, the deployment of measuring medium of the concentration observation system is carried out, comprising:

[0129] From the test scheme, the deployment information of the measuring medium based on each test target in the corresponding target distribution condition is extracted respectively;

[0130] According to the deployment information, the corresponding measuring medium is deployed.

[0131] In this embodiment, the deployment information is, for example, the position of the connecting head.

[0132] In this embodiment, for example, the measuring medium for air tightness test can be special leak detection liquid for leak detection at each joint, and for flow test, it can be a qualified flow meter.

[0133] The beneficial effects of the above technical solutions are: through the extraction of deployment information, the reasonable deployment of measuring medium is realized, and the numerical value of the measurement related performance is facilitated.

[0134] The present application provides a kind of high-precision greenhouse gas concentration observation system test method, constructs corresponding performance matrix, comprising:

[0135] Obtaining the latest maintenance parameters matching each test target from a historical database, and adjusting the identification process of the corresponding test target under the corresponding test performance according to the maintenance status of the latest maintenance parameters;

[0136]

[0137] Among them, Z y The final setting standard for the identification process; C y Initially set the criteria for the identification process; y is the accuracy loss factor when the maintenance status is the overall replacement status; For nth i1 Precision loss factor under maintenance; γ i1 is the precision loss factor under the i1th maintenance; γ max For all γ i1 The maximum value in ; For all γ i1 Variance of n i1 The number of repairs involved in maintaining the target state of repair;

[0138] Determine the number of acquisitions for the corresponding test target based on the preliminary and final set standards of the identification process;

[0139]

[0140] Among them, Δ y is the corresponding adjustment unit; Tc y is the set original acquisition period; T y The collection period corresponding to the determined test target; is the rounding symbol; is the rounding down symbol; Tz is the total operation time of the concentration observation system; N y is the corresponding number of collections;

[0141] Obtain the average number of acquisition times for all test targets under the corresponding test performance that meets the preset normal distribution probability, and obtain the final acquisition cycle;

[0142] The deployed measurement medium is controlled to test the corresponding test target according to the final acquisition cycle to construct a performance matrix.

[0143] In this embodiment, the history database contains historical maintenance data of different test targets, including operations such as target replacement and maintenance on the target itself. Therefore, the threshold adjustment results are discussed in two categories.

[0144] In this embodiment, the maintenance parameters are statistical results of parameters involved in the corresponding maintenance process.

[0145] In this embodiment, the average number of times refers to an average value obtained by averaging all the times that meet the normal distribution probability within 80%.

[0146] In this embodiment, the final collection period = ceiling (Tz / average number of times), and ceiling() is a ceiling function.

[0147] In this embodiment,

[0148]

[0149] In this embodiment, the set standard can be a correlation value.

[0150] The beneficial effects of the above technical solution are: by adjusting the threshold value by obtaining the maintenance state of the latest maintenance parameters of different targets, the collection number is determined, and in order to ensure that all targets can realize full-cycle data collection as much as possible, the average number of times based on the preset normal distribution probability is used to obtain the final collection period, and then the effective number of times is measured, and a performance matrix is constructed, which provides a basis for analyzing the reliability of the system.

[0151] The present application provides a kind of high-precision greenhouse gas concentration observation system test method, each performance matrix is analyzed respectively, including:

[0152] The corresponding performance matrix is analyzed in row and column, to obtain the first analysis value of each row vector and the second analysis value of each column;

[0153] Based on all the first analysis value and the second analysis value, the initial function of the corresponding performance matrix is obtained;

[0154] According to the number of targets involved in the corresponding test performance and the test type, the test precision of the corresponding test performance is determined, and the final analysis value and analysis error of the corresponding performance matrix are obtained in combination with the initial function.

[0155] In this embodiment, row analysis refers to calculating the average value and variance of the corresponding row vector, and the sum of the two is the first analysis value.

[0156] Column analysis refers to calculating the average value and variance of the corresponding column vector, and the sum of the two is the second analysis value.

[0157] In this embodiment, the initial function = C0 (all first analysis values and all second analysis values involved in the corresponding performance matrix).

[0158] In this embodiment, the number of targets refers to the number of test targets under the test performance, and the test type refers to the type of test performance, such as air tightness type, flow type, etc.

[0159] In this embodiment,

[0160]

[0161] Where σ1 is the quantity-precision conversion coefficient of the corresponding test type, and ln is the sign of the logarithmic function.

[0162] In this embodiment, the final analysis value=the average value of all analysis values ​​involved in the initial function+the variance of all analysis values×the test accuracy.

[0163] In this embodiment, the analysis error=1-the variance of all first analysis values+1-the variance of all second analysis values.

[0164] The beneficial effect of the above technical solution is: by performing row analysis and column analysis on the matrix to construct the initial function, and combining the analysis of the accuracy, the final analysis value and analysis error are effectively determined, providing a basis for subsequent label setting.

[0165] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, which sets a level label to a corresponding performance matrix in combination with a performance standard of the corresponding test performance, including:

[0166] Constructing a performance vector based on the performance standard of the corresponding test performance, and sequentially subtracting the performance vector from each row vector in the performance matrix under the corresponding test performance to obtain a subtraction matrix;

[0167] Performing cluster analysis on each column in the subtraction matrix to obtain a cluster set corresponding to the test target, and drawing a distribution map based on each cluster set;

[0168] Determining a distribution dimension in the distribution graph and setting an abnormality probability for a corresponding test target based on the distribution quantity under each distribution dimension;

[0169] The test target whose abnormal probability is greater than the preset probability is regarded as an abnormal target;

[0170] Locking the prominent elements of each row in the subtraction matrix, and setting an elimination probability for the corresponding row in combination with the final analysis value and the analysis error;

[0171] Based on the deployment location of each exception target in the system structure, set dependencies to the corresponding exception target;

[0172] Based on the dependency, maintenance status, and abnormal target, based on abnormal probability, elimination probability, and abnormal characteristics under target characteristics, an abnormal symbolic expression for the abnormal target is generated, and matched with the expression-abnormality comparison table to obtain an initial abnormality;

[0173] Based on all initial anomalies and the column variance of the column vector corresponding to each initial anomaly, a first label is set to the corresponding anomaly target, and in combination with the test weight of the anomaly target, an anomaly level label is set to the corresponding performance matrix.

[0174] In this embodiment, the highlighted elements are, for example, elements greater than b1 are selected from those greater than 0, and b1 is greater than 0, or elements less than b2 are selected from those less than 0, and b2 is less than 0.

[0175] In this embodiment, the performance standard is pre-set, which refers to the standard value of each measurement target under different performances, and thus the performance vector = {standard values ​​of different measurement targets under the corresponding test performance}.

[0176] In this embodiment, subtraction refers to subtracting the same element of each row vector in the matrix from the corresponding performance vector.

[0177] In this embodiment, cluster analysis refers to clustering all values ​​in the column vector, for example, clustering values ​​less than 0, clustering values ​​equal to 0, and clustering values ​​greater than 0 together.

[0178] In this embodiment, the cluster set includes results that are clustered together according to different criteria.

[0179] In this embodiment, the distribution graph includes, for example: clustered distribution less than 0, clustered distribution equal to 0, and clustered distribution greater than 0. The distribution dimension at this time can be regarded as 3, and the distribution number refers to the number of values ​​less than 0, the number of values ​​equal to 0, and the number of values ​​greater than 0.

[0180] In this embodiment, the abnormal probability refers to the number of abnormalities / the total number of distributions, where the number of abnormalities, for example, refers to the number less than 0 + the number greater than 0.

[0181] In this embodiment, the preset probability is pre-set and has a value of 0.8.

[0182] In this embodiment, the prominent element refers to a value in a row of the corresponding matrix that is far above the performance standard or far below the performance standard.

[0183] Elimination probability = ((the number of values ​​far above the performance standard + the number of values ​​far below the performance standard) / total number of distributions) × ((the difference between the average of the absolute values ​​of all prominent elements and the absolute value of the final analysis value - the analysis error) / (the number of all prominent elements × the final analysis value)), and taking the square root.

[0184] In this embodiment, the abnormal symbol expression is: dependency relationship—maintenance status—abnormal characteristics.

[0185] In this embodiment, the expression-exception table includes the exception symbol expression and the initial exception matching the exception symbol expression, and is pre-set and directly matched.

[0186] In this embodiment, the initial exception refers to the existing problem of the corresponding performance, for example, the air leakage at the connecting pipeline a1.

[0187] In this embodiment, the dependency relationship refers to whether there is an influence relationship between the exception target and the neighbor target, if yes, it is considered that there is a dependency relationship, otherwise, it is considered that there is no dependency relationship.

[0188] That is, the dependency relationship is represented by symbol 1, and the dependency relationship is represented by symbol 0.

[0189] In this embodiment, the column vector corresponding to the initial exception is the column vector of the subtraction matrix corresponding to the corresponding exception target, and there may be multiple initial exceptions in a performance matrix, the column variance is the variance of all elements in the corresponding column vector, and the first label set by the corresponding exception target is the average value of the absolute values of all values in the corresponding column vector+column variance, that is, the first label is the set value.

[0190] In this embodiment, the exception level label = the sum of the values of the first labels of all exception targets x test weights.

[0191] In this embodiment, the test weight of each target is pre-set and can be directly used, and the sum of the test weights of all targets involved under different performances is 1.

[0192] The beneficial effects of the above technical scheme are: based on the subtraction result of the performance vector and the performance matrix to construct the distribution graph and set the exception probability, and by comparing all the locked exception targets, the elimination probability is effectively determined by deeply locking the highlight elements, and then the initial exception is obtained by combining the dependency relationship and the exception feature, the reliability of the set label is guaranteed, and the reliability of the system performance test is guaranteed.

[0193] The application provides a test method of a high-precision greenhouse gas concentration observation system, and further comprises: determining an exception target based on an exception probability, an elimination probability and an exception feature under a target characteristic, specifically comprising:

[0194] Obtain the column vector of the corresponding exception target from the subtraction matrix, and perform curve drawing and curve fitting to lock the discrete points;

[0195] Construct a discrete function based on the discrete gradient of each discrete point, and at the same time, construct a probability function based on the exception probability and the elimination probability;

[0196] The discrete function and the probability function are input into a function analysis model that matches the corresponding target characteristics to obtain abnormal characteristics.

[0197] In this embodiment, the curve is drawn by taking the number of occurrences as the horizontal coordinate and the value of the element in the column vector as the vertical coordinate, and the curve fitting belongs to the existing technology and can directly obtain discrete points.

[0198] In this embodiment, the discrete gradient refers to the discrete direction and discrete size, which can be directly determined, and the discrete function = L (the discrete gradient of each discrete point).

[0199] In this embodiment, the probability function = G (abnormality probability, elimination probability).

[0200] In this embodiment, the target characteristic refers to the attribute of the corresponding measurement target, such as pipeline attributes, air outlet attributes, etc., and the function analysis model is obtained based on the matching of the characteristic-model database, which contains unique function analysis models based on different characteristics, and the model is based on the combination of various discrete functions and probability functions under different characteristics and the abnormal analysis of the combination, which can also be fundamentally trained, thereby effectively obtaining abnormal characteristics, such as continuous air leakage, etc.

[0201] The beneficial effect of the above technical solution is: by plotting and fitting the column vector, it is easy to construct a discrete function based on discrete points, and the probability function constructed by combining the two probabilities can effectively obtain the abnormal characteristics, providing a basis for generating abnormal symbolic expressions.

[0202] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, which generates an early warning report of the concentration observation system based on all abnormal level labels, including:

[0203] Based on each abnormal level label, the deployment point of the concentration observation system is rendered using a significant color consistent with the corresponding abnormal level label to obtain a corresponding first deployment map;

[0204] Obtain the performance topology of each first deployment diagram, and align all first deployment diagrams to obtain the abnormal correlation topology of each deployment point;

[0205] Generate an early warning report based on the anomaly correlation of all its own performance topology structures and different anomaly level labels, and combined with the anomaly correlation topology structure of each deployment point.

[0206] In this embodiment, the deployment point is a location point based on the concentration observation system.

[0207] In this embodiment, different levels correspond to different anomaly types, and different anomaly types correspond to different significant colors, for example, blue is used to represent the air tightness anomaly, specifically, light blue is used to represent the anomaly level one, and dark blue is used to represent the anomaly level two.

[0208] In this embodiment, rendering is just for coloring, so as to obtain the deployment diagram.

[0209] In this embodiment, the self-performance topology structure refers to the distribution structure of the deployment points rendered in the first deployment diagram.

[0210] In this embodiment, the anomaly correlation topology structure refers to the structure combination distribution structure of the same deployment point under different rendering results.

[0211] In this embodiment, the anomaly correlation refers to the influence of different anomaly level labels, some exist and some do not.

[0212] In this embodiment, all the self-performance topology structure, the anomaly correlation of different anomaly level labels, and the anomaly correlation topology structure of each deployment point are combined and input into the report generation model to automatically generate the early warning report, wherein the report generation model is obtained by training the neural network model based on the topology structure of different combinations and the anomaly analysis report as a sample.

[0213] The beneficial effects of the above technical solutions are: the anomaly level labels are rendered in the same layer according to the deployment points, and two topology structures are obtained by combining the alignment processing operation and the self overall analysis, so as to facilitate the generation of the alarm report, realize the timely improvement of the system, and ensure the reliability of the system observation.

[0214] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the claims of the present application and the equivalent technology thereof, the present application also intends to include these modifications and variations.

Claims

1. A method for testing a high-precision greenhouse gas concentration observation system, characterized in that: include: Step 1: configuring a test scheme based on different test performances to the concentration observation system according to the system structure of the concentration observation system; Step 2: Deploy the measurement medium of the concentration observation system according to the test plan, and perform performance tests on the corresponding test targets based on the deployed measurement medium to construct a corresponding performance matrix; Step 3: Analyze each performance matrix separately, and set an abnormal level label for the corresponding performance matrix based on the performance standard of the corresponding test performance; Step 4: Generate an early warning report for the concentration observation system based on all abnormal level labels to provide reminders; Among them, building the corresponding performance matrix includes: Obtaining the latest maintenance parameters matching each test target from a historical database, and adjusting the identification process of the corresponding test target under the corresponding test performance according to the maintenance status of the latest maintenance parameters; ； in, To set the standard for the final stage of the identification process; Initially set criteria for the identification process; is the accuracy loss factor when the maintenance status is the overall replacement status; For the Accuracy loss factor under maintenance; For the Accuracy loss factor under maintenance; For all The maximum value in ; For all variance; for The number of repairs involved; Determine the number of acquisitions for the corresponding test target based on the preliminary and final set standards of the identification process; ； ； in, is the corresponding adjustment unit amount; is the set original collection period; The collection period corresponding to the determined test target; is the rounding symbol; is the floor rounding symbol; is the total operation time of the concentration observation system; is the corresponding number of collections; Obtain the average number of acquisition times for all test targets under the corresponding test performance that meets the preset normal distribution probability, and obtain the final acquisition cycle; Controlling the deployed measurement medium to test the corresponding test target according to the final acquisition cycle to construct a performance matrix; Among them, the test performance includes: air tightness test based on concentration observation system, sample gas flow test, retention time test, dehumidification performance test, cold trap tube switching impact test, system performance test, greenhouse gas concentration linearity test, greenhouse gas concentration drift test and greenhouse gas deviation test.

2. The method for testing a high-precision greenhouse gas concentration observation system according to claim 1, characterized in that: Configure test scenarios based on different test capabilities, including: Decompose the system performance requirements to obtain several test requirements; Perform test structure mapping analysis on each test requirement and the system structure, and lock the test target based on the test requirement; Based on the deployment location of the test target in the system structure, the target distribution based on the corresponding test requirements is obtained; Based on the distribution-test comparison table, a test plan that matches the target distribution is retrieved.

3. The method for testing a high-precision greenhouse gas concentration observation system according to claim 2, characterized in that: Deploying the concentration observation system to measure the medium according to the test plan includes: Extracting respectively from the test plan the deployment information of the measurement medium based on each test target in the corresponding target distribution situation; The corresponding measurement medium is deployed according to the deployment information.

4. The method for testing a high-precision greenhouse gas concentration observation system according to claim 1, characterized in that: Each performance matrix is ​​analyzed separately, including: Performing row analysis and column analysis on the corresponding performance matrix to obtain a first analysis value of each row vector and a second analysis value of each column; Based on all the first analysis values ​​and the second analysis values, an initial function of the corresponding performance matrix is ​​obtained; The test accuracy of the corresponding test performance is determined based on the number of targets and test types involved in the corresponding test performance, and combined with the initial function, the final analysis value and analysis error of the corresponding performance matrix are obtained.

5. The method for testing a high-precision greenhouse gas concentration observation system according to claim 4, characterized in that: In combination with the performance standards of the corresponding test performance, level labels are set to the corresponding performance matrix, including: Constructing a performance vector based on the performance standard of the corresponding test performance, and sequentially subtracting the performance vector from each row vector in the performance matrix under the corresponding test performance to obtain a subtraction matrix; Performing cluster analysis on each column in the subtraction matrix to obtain a cluster set corresponding to the test target, and drawing a distribution map based on each cluster set; Determining a distribution dimension in the distribution graph and setting an abnormality probability for a corresponding test target based on the distribution quantity under each distribution dimension; The test target whose abnormal probability is greater than the preset probability is regarded as an abnormal target; Locking the prominent elements of each row in the subtraction matrix, and setting an elimination probability for the corresponding row in combination with the final analysis value and the analysis error; Based on the deployment location of each exception target in the system structure, set dependencies to the corresponding exception target; Based on the dependency, maintenance status, and abnormal target, based on abnormal probability, elimination probability, and abnormal characteristics under target characteristics, an abnormal symbolic expression for the abnormal target is generated, and matched with the expression-abnormality comparison table to obtain an initial abnormality; Based on all initial anomalies and the column variance of the column vector corresponding to each initial anomaly, a first label is set to the corresponding anomaly target, and in combination with the test weight of the anomaly target, an anomaly level label is set to the corresponding performance matrix.

6. The method for testing a high-precision greenhouse gas concentration observation system according to claim 5, characterized in that: Also includes: The determination of abnormal targets is based on abnormal probability, elimination probability, and abnormal characteristics under target characteristics, including: Obtaining a column vector corresponding to an abnormal target from the subtraction matrix, performing curve drawing and curve fitting, and locking discrete points; A discrete function is constructed based on the discrete gradient of each discrete point. At the same time, a probability function is constructed based on the abnormal probability and the elimination probability. The discrete function and the probability function are input into a function analysis model that matches the corresponding target characteristics to obtain abnormal characteristics.

7. The method for testing a high-precision greenhouse gas concentration observation system according to claim 1, characterized in that: Generates an early warning report for the concentration observation system based on all abnormal level labels, including: Based on each abnormal level label, the deployment point of the concentration observation system is rendered using a significant color consistent with the corresponding abnormal level label to obtain a corresponding first deployment map; Obtain the performance topology of each first deployment diagram, and align all first deployment diagrams to obtain the abnormal correlation topology of each deployment point; Generate an early warning report based on the anomaly correlation of all its own performance topology structures and different anomaly level labels, and combined with the anomaly correlation topology structure of each deployment point.

Images