Testing method of high-precision greenhouse gas concentration observation system

Abstract

The invention provides a test method of a high-precision greenhouse gas concentration observation system, which belongs to the technical field of concentration observation and comprises the following steps: configuring test schemes based on different test performances for the concentration observation system according to the system structure of the concentration observation system; performing measurement medium deployment on the concentration observation system according to the test scheme, performing performance test on a corresponding test target based on the deployed measurement medium, and constructing a corresponding performance matrix; analyzing each performance matrix, and setting an abnormal level label for the corresponding performance matrix in combination with the performance standard of the corresponding test performance; and generating an early warning report of the concentration observation system based on all the abnormal level labels for reminding. According to the method, device deployment is carried out on the concentration observation system through test schemes under various test performances, multi-aspect performance test is realized, and an early warning report for the system is output through matrix analysis and label setting, so that the system performance meets the required observation requirements and the reliability of the system is ensured.

Description

Technical Field

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

[0002] At present, the methods for measuring the concentration of greenhouse gases (CO2 / CH4) on the ground with high precision 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 of high-precision observation, resulting in varying accuracy. Among them, the observation instruments and observation systems must undergo strict performance testing, and only after passing the test can they be put into commercial use by the meteorological department.

[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 a variety of test performances, realize multi-faceted performance testing, and then output an early warning report for the system through matrix analysis and label setting, thereby ensuring that the system performance meets the required observation requirements and ensures 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 to the corresponding performance matrix in combination with 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, test schemes based on different test performances are configured, including:

[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 position 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 matching the target distribution is retrieved.

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

[0016] Extracting respectively from the test scheme 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] Acquire the latest maintenance parameters matching each test target from the historical database, and adjust 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 To set the final standard for the identification process; C y Initial setting of 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 i1 The number of repairs involved in maintaining the target state for the repair state;

[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 quantity; Tc y is the set original acquisition period; Ty 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;

[0025] Obtain the average number of acquisition times of all test targets under the corresponding test performance that meets the preset normal distribution probability, and obtain the final acquisition cycle;

[0026] The deployed measurement medium is controlled to test the corresponding test target according to the final acquisition cycle to construct a performance matrix.

[0027] Preferably, each performance matrix is ​​analyzed separately, 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] Based on all the first analysis values ​​and the second analysis values, an initial function of the corresponding performance matrix is ​​obtained;

[0030] According to the number of targets and test types involved in the corresponding test performance, the test accuracy of the corresponding test performance is determined, and combined with the initial function, the final analysis value and analysis error of the corresponding performance matrix are obtained.

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

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

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

[0034] Determine the distribution dimension in the distribution graph and, based on the distribution quantity under each distribution dimension, set an abnormal probability to the corresponding test target;

[0035] The test target whose abnormal probability is greater than the preset probability is regarded as an abnormal target;

[0036] Locking the prominent elements of each row in the subtraction matrix, and setting the elimination probability for the corresponding row in combination with the final analysis value and the analysis error;

[0037] Based on the deployment location of each abnormal target in the system structure, set a dependency relationship to the corresponding abnormal target;

[0038] Based on the dependency, maintenance status and abnormal target, based on abnormal probability, elimination probability and abnormal characteristics under target characteristics, an abnormal symbol 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, it also includes: determining the abnormal characteristics 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, and performing curve drawing and curve fitting to lock discrete points;

[0042] A discrete function is constructed based on the discrete gradient of each discrete point, and 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 characteristic to obtain the abnormal characteristic.

[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 association of all its own performance topology structures and different anomaly level labels, and combined with the anomaly association 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 beneficial effects:

[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 partly become apparent from the description, or 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 The present invention is a flowchart of a method for testing 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 in conjunction with 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 As 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 to the corresponding performance matrix in combination with 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, the ground high-precision greenhouse gas (CO 2 / CH 4 ) After the concentration observation system is installed, it is necessary to conduct an on-site test of the observation system, and then use steps 1 to 4 to implement the subsequent steps. It should be noted that the standard gas used in the test has been balanced and left 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 when the sample gas flows through the pipelines, joints, connectors, etc. in 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 there is no abnormal increase in the carbon dioxide concentration at the analyzer end, it indicates that the overall air tightness of the analyzer and its joints is good.

[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 looseness, and the air inlet pipe and the tower body are firmly fixed; for the pipeline under the tower, use a high-concentration carbon dioxide spraying device to spray around the pipeline and at the joints, and observe that the carbon dioxide concentration at the analyzer end does not increase abnormally, which indicates that the pipeline is airtight.

[0069] (4) Pipeline connecting 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's air outlet and joints, as well as the joints on the gas 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 gas path is airtight.

[0070] (5) Hidden interfaces / valve port connection pipelines: Open the outer panels of equipment such as cold traps and valve boxes, 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 there is no abnormal increase in the carbon dioxide concentration at the analyzer end or no bubbles appear at the joints, it indicates that the air tightness of the hidden interfaces is good.

[0071] 2. Sample gas flow test:

[0072] (1) Set different injection flow rates, the flow rate range is between 200-800 mL / min, and the values ​​are taken 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 measured by spraying high concentration CO around the air inlet. 2 , stopwatch timing observation analyzer CO 2 The time to obtain peak concentration should not exceed 2 minutes.

[0076] (2) The retention time of the standard gas to the analyzer main unit can be measured by switching the valve position to the standard gas inlet gas path and observing the analyzer CO 2 The peak concentration time should not exceed 1 minute.

[0077] Water removal performance test:

[0078] (1) The water removal performance test is performed when 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 computer. 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 H corresponding to the cold trap switching period 2 O、CO 2 and CH 4 Second-level concentration data, H 2 O、CO 2 or CH 4 The time span for the concentration to recover to the concentration level before the cold trap tube was switched should not exceed 1 minute.

[0084] 5. System performance test:

[0085] 5.1 Overall performance test

[0086] (1) After removing the sampling pump, connect the outlet connector to the target standard gas. To avoid wasting the standard gas, close the bypass exhaust port on the rear end of the sampling pump.

[0087] (2) Set the injection 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 alternately measured 3 times as a cycle. The test should be no less than 3 cycles.

[0088] (3) Take the mean of the continuous 3-minute data with the smallest standard deviation of the working standard gas and the target standard gas as their respective measurement results, and use the working standard gas measurement results and nominal concentration to construct a linear regression equation to calculate the CO in the target standard gas. 2 and CH 4 Actual concentration.

[0089] (4) According to the target standard gas CO 2 and CH 4 The deviation between the actual concentration and the nominal concentration is used to evaluate whether the overall performance meets the requirements. 2 and CH 4 The deviations should be less than 0.1ppm and 2ppb respectively; for polluted areas, CO 2 and CH 4 The deviation should be less than 0.2ppm and CH 4 ≤5ppb.

[0090] 5.2 Linearity test:

[0091] (1) Using 5 bottles containing atmospheric greenhouse gas CO 2 and CH 4 The concentration gradient standard gas within the concentration fluctuation range constitutes a standard gas sequence of different concentrations for linear testing.

[0092] (2) By setting the injection sequence, 5 bottles of gradient standard gas are allowed to enter the measurement unit in sequence. The injection and measurement time of each bottle of standard gas is 5 minutes, and the measurement is repeated 3 times.

[0093] (3) Take the average value of the continuous 3-minute data with the smallest standard deviation for each bottle of standard gas as the mean value of each measurement, and take the arithmetic mean of the 3 measurement means as the measurement result for each bottle of standard gas.

[0094] (4) Using 5 bottles of standard gas CO 2 and CH 4 The measured results were linearly fitted with their corresponding nominal concentrations, CO 2 and CH 4 The goodness of fit should be greater than 0.999, and the fitting residuals should be evenly distributed without obvious outliers.

[0095] 5.3 Repeatability test:

[0096] (1) Under the same measurement conditions, continuous measurement shall be carried out using standard gas for a measurement period 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. 2 <0.1ppm、CH 4 <2ppb; for polluted areas, CO 2 <0.2ppm、CH 4 <5ppm.

[0099] 5.4 Drift test:

[0100] (1) Under the same measurement conditions, continuous measurement is carried out using standard gas, and the measurement time is not less than 24 hours.

[0101] (2) Exclude the measurement data of the first hour and use the measurement data with stable readings of the remaining instruments 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 hourly measurement results should meet the following requirements: For clean areas, CO 2 <0.1ppm、CH 4 <2ppb; for polluted areas, CO 2 <0.2ppm、CH 4 <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. The alternating measurement is 3 times as one cycle. The test should be 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 CO concentration in the gradient standard gas of each cycle is calculated by linear fitting the measurement results of the working standard gas of each cycle and its nominal concentration. 2 and CH 4 actual concentration.

[0108] (5) According to all gradient standard gas CO 2 and CH 4 The deviation between the actual concentration and the nominal concentration is used to evaluate whether the accuracy test meets the requirements. 2 <0.1ppm、CH 4 <2ppb; for polluted areas, CO 2 <0.2ppm、CH 4 <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 observations.

[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 setting structure 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 relevant values ​​measured once, and there are N measurements, where N is a variable.

[0116] In this embodiment, the abnormal level label is another form for 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 abnormal maintenance.

[0118] The beneficial effect of the above technical solution is: through the test schemes under various test performances, the devices of the concentration observation system are deployed 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 to ensure its reliability.

[0119] The present invention provides a test method for a high-precision greenhouse gas concentration observation system, which configures a test scheme based on different test 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 position 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 comparison table, a test plan matching the target distribution is retrieved.

[0124] In this embodiment, the system performance requirements refer to the overall requirements of the concentration observation system, including air tightness requirements, flow requirements, time requirements, deviation requirements, drift requirements, linearity requirements, etc. Therefore, several test requirements can be obtained by directly disassembling them.

[0125] In this embodiment, the test structure mapping analysis refers to certain parts of the concentration observation system that are tested as required by the corresponding test requirements, such as the air tightness test - the joints (certain parts) of each connecting pipeline, and the target distribution is the deployment of the locations of the joints.

[0126] In this embodiment, the distribution-test comparison table includes different target distribution conditions and test schemes corresponding to the distributions, and the test schemes are, for example, detecting whether there is leakage of gas in corresponding parts.

[0127] The beneficial effect of the above technical solution is: by decomposing the requirements and analyzing the structure mapping, the test target is effectively locked, and then the test plan is retrieved from the comparison table, providing a basis for the subsequent construction of the performance matrix.

[0128] The present invention provides a test method for a high-precision greenhouse gas concentration observation system, and deploys a measurement medium for the concentration observation system according to the test scheme, including:

[0129] Extracting respectively from the test scheme the deployment information of the measurement medium based on each test target in the corresponding target distribution situation;

[0130] The corresponding measurement medium is deployed according to the deployment information.

[0131] In this embodiment, the deployment information includes, for example, the location of the targeted connector.

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

[0133] The beneficial effect of the above technical solution is: by extracting the deployment information, the reasonable deployment of the measurement medium is achieved, which facilitates the measurement of the numerical values ​​of the relevant performance.

[0134] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, and constructs a corresponding performance matrix, including:

[0135] Acquire the latest maintenance parameters matching each test target from the historical database, and adjust 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 To set the final standard for the identification process; C y Initial setting of 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 i1 The number of repairs involved in maintaining the target state for the repair state;

[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 quantity; 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 of 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 replacement of targets and maintenance on the target body, so the adjustment results of the threshold 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 the average value obtained by averaging all the times that meet the normal distribution probability within 80% among all the acquisition times.

[0146] In this embodiment, the final acquisition period=ceiling(Tz / average number of times), where ceiling() is a rounding-up function.

[0147] In this embodiment,

[0148]

[0149] In this embodiment, the setting standard may be a related value.

[0150] The beneficial effects of the above technical solution are: by obtaining the latest maintenance status of maintenance parameters of different targets to adjust the threshold, and then determine the number of collections, and in order to ensure that all targets can achieve full-cycle data collection as much as possible, the final collection cycle is obtained based on the average number of times that meet the preset normal distribution probability, and then the effective number of times is measured, and a performance matrix is ​​constructed to provide a basis for analyzing the reliability of the system.

[0151] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, which analyzes each performance matrix separately, including:

[0152] 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;

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

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

[0155] In this embodiment, row analysis refers to calculating the average value and variance of the corresponding row vector, and adding the average value and variance to obtain a first analysis value.

[0156] Column analysis refers to calculating the mean and variance of the corresponding column vector, and adding the two together to obtain the second analysis value.

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

[0158] In this embodiment, the target number 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, analysis error=1-variance of all first analysis values+1-variance of all second analysis values.

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

[0165] The present invention provides a test 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 a corresponding test performance, including:

[0166] A performance vector is constructed based on the performance standard of the corresponding test performance, and each row vector in the performance matrix under the corresponding test performance is sequentially subtracted 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 diagram based on each cluster set;

[0168] Determine the distribution dimension in the distribution graph and, based on the distribution quantity under each distribution dimension, set an abnormal probability to the corresponding test target;

[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 the 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 abnormal target in the system structure, set a dependency relationship to the corresponding abnormal 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 symbol 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 preset, 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: a clustered distribution less than 0, a clustered distribution equal to 0, and a clustered distribution greater than 0. The distribution dimension at this time can be regarded as 3, and the distribution quantity 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, wherein 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 - analysis error) / (the number of all prominent elements × the final analysis value)), and the square root is taken.

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

[0185] In this embodiment, the expression-exception comparison table includes the abnormal symbol expression and the initial exception that matches the abnormal symbol expression, which is pre-set and can be directly matched.

[0186] In this embodiment, the initial abnormality refers to a problem with the corresponding performance, for example, there is a gas leak at the connecting pipeline a1.

[0187] In this embodiment, the dependency relationship refers to whether there is an influence relationship between the abnormal target and the surrounding targets. If so, it is considered that there is a dependency relationship, otherwise it is considered that there is no dependency relationship;

[0188] That is, the existence of a dependency relationship is indicated by the symbol 1, and the non-existence of a dependency relationship is indicated by the symbol 0.

[0189] In this embodiment, the column vector corresponding to the initial anomaly is the column vector of the subtraction matrix corresponding to the corresponding abnormal target, and there may be multiple initial anomalies in a performance matrix. The column variance is the calculated variance of all elements in the corresponding column vector, and the first label set for the corresponding abnormal target is: the average value of the absolute values ​​of all values ​​in the corresponding column vector + the column variance, that is, the first label is the set value.

[0190] In this embodiment, the abnormal level label = the value of the first label of all abnormal targets × the sum of the test weights.

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

[0192] The beneficial effects of the above technical solution are: based on the subtraction result of the performance vector and the performance matrix, a distribution map is constructed to set the abnormality probability, and by comparing all locked abnormal targets, and then deeply locking the prominent elements to effectively determine the elimination probability, and then combining the dependency relationship and abnormal characteristics to obtain the initial abnormality, thereby ensuring the reliability of setting the label and thus ensuring the reliability of the system performance test.

[0193] The present invention provides a testing method for a high-precision greenhouse gas concentration observation system, further comprising: determining abnormal characteristics of an abnormal target based on abnormal probability, elimination probability and target characteristics, specifically including:

[0194] Obtaining a column vector corresponding to an abnormal target from the subtraction matrix, and performing curve drawing and curve fitting to lock discrete points;

[0195] A discrete function is constructed based on the discrete gradient of each discrete point, and a probability function is constructed based on the abnormal 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 characteristic to obtain the abnormal characteristic.

[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 prior art 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, probability function=G(abnormal probability, elimination probability).

[0200] In this embodiment, the target characteristic refers to the attribute of the corresponding measurement target, such as pipeline attributes, 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 leakage, etc.

[0201] The beneficial effect of the above technical solution is: by drawing 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 association of all its own performance topology structures and different anomaly level labels, and combined with the anomaly association 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 and different types of abnormalities have corresponding colors of different significance. For example, blue is used to represent airtightness abnormalities. Specifically, light blue is used to represent abnormality level one, dark blue is used to represent abnormality level two, and so on.

[0208] In this embodiment, rendering is for coloring to facilitate obtaining a 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 graph.

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

[0211] In this embodiment, the abnormal association refers to the abnormal impact between labels of different abnormal levels, some of which have an impact and some do not.

[0212] In this embodiment, all the performance topological structures, the abnormal associations of different abnormal level labels, and the abnormal association topological structures of each deployment point are combined and input into the report generation model to automatically generate an early warning report, wherein the report generation model is obtained by training a neural network model based on samples of topological structures with different combinations and abnormal analysis reports based on the structures.

[0213] The beneficial effect of the above technical solution is: the abnormal level labels are rendered in the same color according to the deployment points, and the two topological structures are obtained by combining the alignment processing operation and its own overall analysis, which is convenient for generating alarm reports, realizing timely improvement of the system, and ensuring the reliability of system observation.

[0214] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A testing method for 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 to the corresponding performance matrix in combination with 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.

2. The testing method of the high-precision greenhouse gas concentration observation system according to claim 1 is 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 position 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 matching the target distribution is retrieved.

3. The testing method of the high-precision greenhouse gas concentration observation system according to claim 1 is characterized in that: Deploying the concentration observation system to measure the medium according to the test plan includes: Extracting respectively from the test scheme 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 testing method of the high-precision greenhouse gas concentration observation system according to claim 1 is characterized in that: Construct the corresponding performance matrix, including: Acquire the latest maintenance parameters matching each test target from the historical database, and adjust the identification process of the corresponding test target under the corresponding test performance according to the maintenance status of the latest maintenance parameters; Among them, Z y To set the final standard for the identification process; C y Initial setting of 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 i1 The number of repairs involved in maintaining the target state of repair; Determine the number of acquisitions for the corresponding test target based on the preliminary and final set standards of the identification process; Among them, Δ y is the corresponding adjustment unit quantity; 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; Obtain the average number of acquisition times of all test targets under the corresponding test performance that meets the preset normal distribution probability, and obtain the final acquisition cycle; The deployed measurement medium is controlled to test the corresponding test target according to the final acquisition cycle to construct a performance matrix.

5. The testing method of the high-precision greenhouse gas concentration observation system according to claim 4 is 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; According to the number of targets and test types involved in the corresponding test performance, the test accuracy of the corresponding test performance is determined, and combined with the initial function, the final analysis value and analysis error of the corresponding performance matrix are obtained.

6. The testing method of the high-precision greenhouse gas concentration observation system according to claim 5 is characterized in that: In combination with the performance criteria of the corresponding test performance, level labels are set to the corresponding performance matrix, including: A performance vector is constructed based on the performance standard of the corresponding test performance, and each row vector in the performance matrix under the corresponding test performance is sequentially subtracted 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 diagram based on each cluster set; Determine the distribution dimension in the distribution graph and, based on the distribution quantity under each distribution dimension, set an abnormal probability to the corresponding test target; 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 the elimination probability for the corresponding row in combination with the final analysis value and the analysis error; Based on the deployment location of each abnormal target in the system structure, set a dependency relationship to the corresponding abnormal target; Based on the dependency, maintenance status and abnormal target, based on abnormal probability, elimination probability and abnormal characteristics under target characteristics, an abnormal symbol 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.

7. The testing method of the 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, and performing curve drawing and curve fitting to lock discrete points; A discrete function is constructed based on the discrete gradient of each discrete point, and 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 characteristic to obtain the abnormal characteristic.

8. The testing method of the high-precision greenhouse gas concentration observation system according to claim 1 is 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 association of all its own performance topology structures and different anomaly level labels, and combined with the anomaly association topology structure of each deployment point.

9. The method for testing a high-precision greenhouse gas concentration observation system according to claim 1, characterized in that: 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.

Images