Flue gas detector and calibration detection method thereof

Abstract

The invention relates to the technical field of air pollution detection, in particular to a flue gas detector and a calibration detection method thereof, and the flue gas detector comprises a concentration fluctuation analysis module, an abnormal data screening module, an error interval collection module, a parameter trend analysis module and a period stability judgment module. According to the method, abnormal data and equipment drift influence are dynamically screened, an error change interval is actively collected, and through linkage comparison of multi-source parameter trends and separation of working condition disturbance, instrument abnormality and periodic offset signals, automatic screening of concentration data and operation parameters under multiple links is achieved, feature recognition and abnormal partition are associated, and the accuracy of the system is improved. The sampling effectiveness, the data stability and the instrument parameter change trend under the continuous detection working condition are managed in a layered mode, a traceable dynamic calibration chain is formed, the smoke detection consistency and the data exception handling capacity in the complex emission environment are remarkably improved, and an intelligent judgment mechanism of data grading rejection and trend aggregation is perfected.

Description

Technical Field

[0001] This invention relates to the field of air pollution detection technology, and in particular to a flue gas detector and its calibration method. Background Technology

[0002] Air pollution detection refers to the monitoring and analysis of the composition and concentration of pollutants in the air to detect and evaluate environmental quality. It mainly includes monitoring emissions from stationary pollution sources, ambient air quality monitoring, and industrial process emission control. Traditional flue gas detectors are devices used to analyze and measure the composition of flue gas from industrial combustion processes or emission sources. These devices typically undergo calibration using standard gases for comparison or concentration conversion using quantitative dilution devices. They primarily rely on standard gases of known concentrations for flow control and comparison detection, or on fixed-ratio gas mixing devices for instrument calibration.

[0003] Existing technologies employ static calibration methods, comparing test data only under standard or dilution gas conditions. This ignores the actual fluctuations in flue gas emissions and the impact of environmental interference during sampling, making it difficult to promptly screen out sudden anomalies within the data sampling period. Furthermore, there is a lack of correlation analysis between changes in instrument parameters and test data, making it difficult to guarantee data stability and error traceability during periodic calibration. When encountering frequent environmental disturbances or significant emission fluctuations, misjudgments, abnormal data accumulation, and a decline in calibration effectiveness are likely to occur, affecting the comprehensive ability to distinguish the continuity and abnormal states of flue gas emissions. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a flue gas detector and its calibration and testing method. The technical solution is as follows: On the one hand, a flue gas detector is provided, the detector comprising: The concentration fluctuation analysis module is based on the flue gas emission channel. It analyzes the sampling data in each continuous calibration cycle, calculates the differences between adjacent time points, summarizes the time series differences, summarizes the fluctuation characteristics in each cycle, and obtains the concentration fluctuation characteristic sequence. Based on the concentration fluctuation feature sequence, the abnormal data screening module determines the wind speed and humidity corresponding to the sampling period, analyzes their coordinated changes, filters abnormal data related to changes in environmental parameters, calibrates the constant temperature state of the sampling probe, and obtains the abnormal screening dataset. The error interval aggregation module, based on the anomaly screening dataset, compares the difference between the sampling period and the self-test zero-point benchmark, analyzes the data offset in conjunction with temperature changes, groups and organizes the differences, summarizes the deviation of the detection segment, and obtains the interval offset feature group. Based on the interval offset feature group, the parameter trend analysis module analyzes the sensitivity response of the ultraviolet detection channel, filters out the stages with concentrated span changes, compares the channel trends, and obtains the trend change feature quantity. The periodic stability discrimination module analyzes the deviation of the parameter change trend from the mean within the period based on the trend change feature quantity, filters out irregular segments, statistically removes periodic segments, and obtains stability screening statistics.

[0005] On the other hand, the concentration fluctuation feature sequence includes segment identifier, fluctuation mean, and rate of change; the anomaly screening dataset includes screening label, associated time period, and anomaly attribute; the interval offset feature group includes offset type, corresponding interval, and offset degree; the trend change feature quantity includes span trend, sensitivity trend, and periodic feature; and the stability screening statistics include number of removals, stability classification, and anomaly ratio.

[0006] On the other hand, the concentration fluctuation analysis module includes: The difference calculation submodule analyzes the concentration detection data at each sampling time within a continuous calibration cycle based on the flue gas emission channel, compares the detection results of adjacent sampling periods, calculates the difference between the subsequent detection data and the previous detection data, organizes the difference data, and obtains the concentration change amplitude sequence. The time-series aggregation submodule determines the continuity of the detection sampling time based on the concentration change amplitude sequence, removes sampling interruption segments, and reorders the remaining data according to the sampling time to obtain a time-series difference arrangement sequence. The fluctuation induction submodule arranges the sequence based on the time period differences, analyzes the direction of change and the magnitude of data increase or decrease of adjacent data segments, aggregates the correlation parameters of the sampling segments, adjusts the data grouping, and summarizes the dynamic characteristics of each segment to obtain the concentration fluctuation characteristic sequence.

[0007] On the other hand, the abnormal data screening module includes: The environmental judgment submodule analyzes wind speed data and relative humidity data based on the concentration fluctuation characteristic sequence, judges the trend of wind speed change in each sampling period, compares the direction of relative humidity change, determines the synchronicity of wind speed and relative humidity changes in each period, and calculates the frequency of coordinated change of the two to obtain the coordinated change characteristic quantity. The change matching submodule compares the changes in each time period with the coordinated change feature quantity, filters out the time periods in which the concentration change amplitude is consistent with the changes in wind speed and relative humidity, determines the synchronous change interval, and obtains the abnormal joint segment sequence. The isothermal calibration submodule determines the heating status of the sampling probe in the abnormal section based on the abnormal joint segment sequence, filters out non-isothermal operation segments, removes the corresponding abnormal data, and classifies and marks the causes of the abnormality to obtain an abnormal screening dataset.

[0008] On the other hand, the error interval aggregation module includes: The benchmark comparison submodule analyzes the detection results of each sampling period based on the anomaly screening dataset, compares the detection data of each sampling period with the instrument self-test zero-point benchmark, determines the degree of numerical deviation between the two, records the difference distribution state, and obtains the zero-point offset factor. The temperature offset submodule compares the temperature change trend corresponding to the sampling period based on the zero-point offset factor, determines the correlation characteristics between temperature change and the zero-point offset factor, filters out data offset phenomena under the influence of temperature, and obtains the amount of temperature influence change. The grouping and sorting submodule groups the samples based on the temperature change and the time and difference information of each sampling segment, summarizes the degree of deviation of each group during the detection process, and statistically analyzes the changes between groups to obtain the interval deviation feature group.

[0009] On the other hand, the parameter trend analysis module includes: The span variation submodule acquires the infrared detection channel output data for each detection cycle based on the interval offset feature group, calculates the change of the span parameter in the continuous cycle, determines the change amplitude of the parameter in each cycle, and records the span variation in each cycle to obtain the infrared span fluctuation amount. The sensitivity trend submodule analyzes the response changes of the ultraviolet differential detection channel within the same period based on the infrared span fluctuation amount, compares the fluctuation characteristics of sensitivity data with the period, identifies the concentrated stage of sensitivity change, and obtains the ultraviolet sensitivity sequence. The trend comparison submodule compares the changing trends of the infrared detection channel and the ultraviolet detection channel in each detection cycle based on the ultraviolet sensitivity sequence, judges the consistency of the changes between the two, organizes the trend change stage information, and obtains the trend change feature quantity.

[0010] On the other hand, the periodic stability discrimination module includes: The continuous trend submodule analyzes the growth stage of the period span parameter within a continuous period based on the trend change characteristic quantity, judges the continuity and fluctuation characteristics of parameter changes, counts the continuous growth period, and obtains the number of continuous growth intervals. The mean offset submodule analyzes the degree of deviation between the span parameter of each detection cycle and the cycle mean based on the number of continuous growth intervals, determines the relationship between the parameter change trend in each cycle and the cycle mean, filters out the time periods that are inconsistent with the mean change, and obtains the cycle offset trend number. The anomaly screening submodule determines the deviation of the span parameter change range from the normal change range within the period segment based on the period offset trend number, counts the number of period segments that are screened out, and classifies and summarizes the abnormal period segments to obtain the stability screening statistics.

[0011] On the other hand, the continuous calibration cycle refers to each complete cycle in which the flue gas detector performs multiple continuous automatic or manual calibrations of the detection system at set intervals.

[0012] On the other hand, the synergistic change refers to the trend of wind speed and relative humidity changing together over time.

[0013] On the other hand, a calibration and testing method for a flue gas detector is provided. This method, applied to a flue gas detector, includes the following steps: S1: Based on the flue gas emission channel, analyze the concentration data obtained at each sampling time within the continuous calibration cycle, compare the detection results corresponding to two adjacent samplings, calculate the difference between each group of data by subtraction, collect all differences according to the sampling time sequence, and summarize the data fluctuations within the detection cycle by group to obtain the concentration fluctuation characteristic sequence. S2: Based on the concentration fluctuation feature sequence, determine the wind speed and relative humidity corresponding to each sampling period, analyze the coordinated change law of wind speed and relative humidity, screen out abnormal segments that are related to the change amplitude of concentration difference and the change of environmental parameters, calibrate the constant temperature operation status of the flue gas sampling probe, and obtain the abnormal screening dataset. S3: Based on the anomaly screening dataset, compare the differences between each sampling period and the instrument self-test zero-point reference, analyze the data offset caused by temperature changes during the sampling period, sort out the difference information of all sampling segments in groups, summarize the deviation of each group of detection segments, and obtain the interval offset feature group. S4: Based on the interval offset feature group, calculate the span parameter change of the infrared detection channel output within the continuous detection period, analyze the response change of the ultraviolet differential detection channel sensitivity in the same period, screen the stage with concentrated span change amplitude, and compare the trends of infrared and ultraviolet detection channels to obtain the trend change feature quantity. S5: Based on the trend change characteristic quantity, determine the stage of continuous growth of the span parameter of each detection cycle, analyze the deviation of the change trend of the span parameter of each cycle from the mean, screen out segments with irregular change amplitude within the cycle segment, count the number of cycle segments that are screened out, and obtain the stability screening statistics.

[0014] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: Based on the real-time concentration fluctuation characteristics at the flue gas emission site, the system dynamically screens for abnormal data and the impact of equipment drift, actively collects error variation ranges, and separates operating condition disturbances, instrument anomalies, and periodic offset signals through the linkage comparison of multi-source parameter trends. This enables automatic screening of concentration data and operating parameters under multiple stages, identification of related features, and anomaly zoning. It also enables hierarchical management of sampling effectiveness, data stability, and instrument parameter variation trends under continuous detection conditions, forming a traceable dynamic calibration chain. This significantly improves the consistency of flue gas detection and the ability to handle data anomalies in complex emission environments, and perfects the intelligent discrimination mechanism for data classification and trend aggregation. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the modules of the present invention; Figure 2 This is a schematic diagram of the module framework of the present invention; Figure 3 This is a flowchart of the concentration fluctuation analysis module of the present invention; Figure 4 This is a flowchart of the abnormal data screening module of the present invention; Figure 5 This is a flowchart of the error interval aggregation module of the present invention; Figure 6 This is a flowchart of the parameter trend analysis module of the present invention; Figure 7 This is a flowchart of the periodic stability discrimination module of the present invention; Figure 8 This is a flowchart of the method steps of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] This invention provides a flue gas detector, such as... Figure 1 As shown, the detector includes: The concentration fluctuation analysis module is based on the flue gas emission channel. It analyzes the concentration data obtained at each sampling time within the continuous calibration cycle, compares the detection results corresponding to two adjacent samplings, calculates the difference between each group of data by subtraction, collects all differences according to the sampling time sequence, and summarizes the data fluctuations within the detection cycle by group to obtain the concentration fluctuation characteristic sequence. The abnormal data screening module is based on the concentration fluctuation feature sequence to determine the wind speed and relative humidity corresponding to each sampling period, analyze the coordinated change law of wind speed and relative humidity, screen abnormal segments that are related to the change of concentration difference and environmental parameter changes, calibrate the constant temperature operation status of the flue gas sampling probe, and obtain the abnormal screening dataset. The error interval aggregation module is based on the anomaly screening dataset. It compares the differences between the sampling period and the instrument self-test zero point reference, analyzes the data offset caused by temperature changes during the sampling period, groups and organizes the difference information of all sampling segments, summarizes the deviation of each group of detection segments, and obtains the interval offset feature group. The parameter trend analysis module calculates the span parameter changes of the infrared detection channel output within a continuous detection period based on the interval offset feature group, analyzes the response changes of the ultraviolet differential detection channel sensitivity in the same period, filters out the stages with concentrated span changes, and compares the trends of the infrared and ultraviolet detection channels to obtain the trend change feature quantity. The periodic stability discrimination module uses trend change characteristics to determine the stage of continuous growth of parameters in each detection period, analyzes the deviation of the change trend of parameters in each period from the mean, filters out segments with irregular change amplitudes within the period segment, counts the number of period segments that are filtered out, and obtains the stability screening statistics.

[0023] Concentration fluctuation feature sequences include segment identifiers, fluctuation mean, and rate of change; anomaly screening datasets include screening labels, associated time periods, and anomaly attributes; interval offset feature groups include offset type, corresponding interval, and offset degree; trend change feature quantities include span trend, sensitivity trend, and periodic features; and stability screening statistics include the number of removals, stability classification, and anomaly ratio.

[0024] In the concentration fluctuation analysis module, the continuous calibration cycle refers to each complete cycle in which the flue gas detector performs multiple continuous automatic or manual calibrations of the detection system at set intervals. Each cycle typically includes sampling, detection, and internal parameter correction processes. The detection result refers to the actual concentration measurement data of flue gas pollutants read by the sensor (or detection unit) at each sampling moment, which can be a digital display or a concentration reading stored internally by the system. The sequential subtraction method refers to subtracting the concentration detection results of two adjacent samplings in chronological order to obtain the concentration change between each pair of sampling moments, reflecting the dynamic fluctuation of flue gas emissions. The difference between each set of data refers to the concentration change between each pair of adjacent sampling points, that is, the concentration difference between two points before and after a set of sampling data, used to evaluate emission fluctuations. The aggregation of all differences refers to the aggregation of all calculated concentration changes in chronological order to form a complete change sequence, reflecting the dynamic change process of concentration within a continuous cycle. The data fluctuation refers to the frequency and amplitude of concentration changes in the difference sequence, used to reveal the instability and transient change characteristics of concentration during flue gas emissions.

[0025] In the abnormal data screening module, wind speed and relative humidity refer to the wind speed and relative humidity of the gas in the pipeline simultaneously measured on-site during each flue gas sampling period, serving as environmental parameters reflecting changes in sampling conditions. Coordinated change pattern refers to the trend of these two environmental parameters, wind speed and relative humidity, changing together over time. By comparing the synchronous fluctuations of these two items, it helps determine whether environmental anomalies occurred during the sampling period. Concentration difference change amplitude refers to the magnitude of concentration change (difference) within each time period obtained from the concentration fluctuation analysis module, providing direct data for determining whether flue gas emission is stable or experiencing abnormal fluctuations. Environmental parameter variation refers to drastic fluctuations in external conditions such as wind speed and humidity during certain periods, which can affect the reliability of concentration data. Abnormal segments refer to data segments where the detected concentration change amplitude is correlated with environmental parameter changes and exhibits unconventional or abrupt characteristics; these need to be excluded to ensure the accuracy of subsequent analysis. Constant temperature operation refers to the state where the flue gas sampling probe operates continuously and stably at a set temperature; only data collected under these conditions is considered valid, excluding abnormal sampling periods affected by factors such as insufficient heating.

[0026] In the error interval aggregation module, the instrument self-test zero-point reference refers to the system baseline data obtained during the flue gas detector's self-test. It represents the theoretical zero concentration reference when the instrument does not detect pollutants and is used for subsequent deviation correction of actual detection values. The data deviation refers to the difference between the sampled data and the self-test zero-point reference, reflecting the systematic error caused by factors such as environment, equipment aging, or drift in the detection data. The difference information refers to the set of all differences between the sampled concentration and the zero-point reference, which truly records the error performance of each sampling segment under different conditions. The deviation of the detection segment refers to the degree of deviation of each group of sampled data, which is used to distinguish data stability and potential error sources, and further support calibration and traceability.

[0027] In the parameter trend analysis module, the infrared detection channel refers to the dedicated sensing channel in the flue gas detector used for infrared absorption detection of pollutants (such as CO and CO2); the span parameter change refers to the change in key instrument parameters (such as measurement range) output by the infrared detection channel over time in different detection cycles, used to identify fluctuations in instrument performance; the ultraviolet differential detection channel refers to the channel in the flue gas detector used for ultraviolet absorption detection of specific pollutants (such as SO2 and NO). X The dedicated sensing channel of the ultraviolet differential detection channel; response change refers to the dynamic change of the detection sensitivity or response intensity of the ultraviolet differential detection channel with the detection cycle, reflecting the stability of the system; the stage of concentrated span change refers to the time period in which the span parameter changes are relatively dense or continuous, and the stage is prone to abnormal performance of instrument parameter fluctuation.

[0028] In the periodic stability discrimination module, the phase of continuous increase in span parameter refers to the time period in which the instrument span parameter shows a sustained upward trend within several detection cycles. This phase indicates zero drift or range instability of the instrument. The deviation of the change trend from the mean refers to the difference between the actual change trend of the span parameter in the detection cycle and the average trend of each cycle. This is used to judge the reliability of the cycle operation. The irregular segment refers to the time period in which the parameter change significantly deviates from the average trend or the overall change pattern. This is the key target for periodic stability discrimination and abnormal data removal. The number of filtered period segments refers to the total number of data segments that are removed in the periodic stability discrimination due to non-compliance with the change pattern. This data is used for subsequent equipment maintenance or parameter library optimization.

[0029] like Figure 2 and Figure 3 As shown, the concentration fluctuation analysis module includes: The difference calculation submodule analyzes the concentration detection data at each sampling time within a continuous calibration cycle based on the flue gas emission channel, compares the detection results of adjacent sampling periods, calculates the difference between the subsequent detection data and the previous detection data, organizes the difference data, and obtains the concentration change amplitude sequence. First, sampling is performed once per minute within a continuous calibration cycle. A total of 30 sets of concentration data are collected within a 30-minute cycle. Each set of data is obtained through a concentration detection unit deployed in the flue gas emission channel. The sampled data represents the actual pollutant concentration values; for example, the concentration in the first minute is 21.3, in the second minute it is 20.6, in the third minute it is 22.1, and in the fourth minute it is 21.5. The data is then processed. During the process, the detection value of the next minute is extracted and subtracted from the detection value of the previous minute to obtain the concentration change between two adjacent samplings. For example, the difference between the second minute and the first minute is - The difference between the 2nd and 3rd minutes is +1.5, and the difference between the 3rd and 4th minutes is -0.6. This process proceeds step by step in chronological order, calculating the difference between each pair of adjacent sampling periods. After each calculation, the difference is recorded in the dataset and sorted according to the original sampling time order to form a continuous sequence of concentration change amplitudes, such as [-0.7, +1.5, -0.6, +0.9, ...]. The difference values ​​are integrated into a complete time series through the difference data processing step to reflect the change characteristics of concentration data in the continuous sampling period. This series will serve as the input data for subsequent processing modules.

[0030] The time-series aggregation submodule determines the continuity of the detection sampling time based on the concentration change amplitude sequence, removes sampling interruption segments, and reorders the remaining data according to the sampling time to obtain the time-series difference arrangement sequence; The actual sampling time corresponding to each difference data point is obtained. By traversing the difference data index and comparing it with the sampling time recorded by the sampling control system, it is possible to identify whether there is a sampling interruption. For example, if the sampling time is recorded as 1, 2, 3, 7, and 8 minutes, it can be determined that the data from the 4th to the 6th minute is missing, which is a sampling interruption segment. The elimination method is to delete all difference values ​​associated with the interruption segment and at the same time delete the edge data segments that are discontinuous in time due to the interruption. For example, the difference item between the 3rd minute and the 7th minute is cleared. The remaining data is a complete and continuous record. Then, the remaining difference values ​​are reordered according to time sequence, that is, starting from the earliest valid sampling time and arranged in ascending order of minutes, to ensure that each difference data has a strict progressive relationship on the time axis. After sorting, the concentration difference data constitutes a time-segment difference sequence, such as the first item being the difference from 1 minute to 2 minutes, the second item being the difference from 2 minutes to 3 minutes, the third item being the difference from 7 minutes to 8 minutes, etc. This sequence structure clearly reflects the concentration changes during the sampling process and provides a basis for accurate time for subsequent fluctuation analysis.

[0031] The fluctuation induction submodule arranges the sequence based on time-period differences, analyzes the direction of change and the magnitude of data increase and decrease in adjacent data segments, aggregates the correlation parameters of sampling segments, adjusts data grouping, and summarizes the dynamic characteristics of each segment to obtain the concentration fluctuation characteristic sequence. For each data difference, direction determination and magnitude classification are performed. Direction determination identifies whether the difference is positive or negative, indicating an increase or decrease in concentration. For example, a difference of +1.2 in segment 1 is classified as an increase, and -0.6 in segment 2 as a decrease. The direction of change between adjacent segments is compared; if the directions are the same, they are classified as the same-direction segment; if the directions are opposite, they are classified as opposite-direction segments. Then, the magnitude of change is classified according to the absolute value of each difference: low fluctuation is defined as a difference between 0.0 and 0.5, medium fluctuation as a difference between 0.5 and 1.5, and high fluctuation as a difference above 1.5. For example, a difference of 0.3 is classified as low fluctuation, 1.1 as medium fluctuation, and 2.7 as high fluctuation. Combined with simultaneously recorded wind speed and relative humidity information, the environmental conditions of each data segment are analyzed. The comparison extracts the wind speed and humidity changes for each sampling period. If the wind speed change between two consecutive sampling points exceeds 0.8 or the relative humidity change exceeds 6, the segment is marked as an environmental disturbance segment and classified separately. All remaining segments are grouped according to the fluctuation level and direction. For example, three consecutive upward segments with medium fluctuations are grouped together, and three consecutive downward segments with high fluctuations are grouped together. Finally, all groups are marked as concentration fluctuation feature information. The information records the group number, start and end time, fluctuation direction, fluctuation level, and corresponding environmental parameters, which are used to construct the concentration fluctuation feature sequence. For example, a group is identified as number 1, with an upward direction, a medium fluctuation level, a wind speed of 3.2, and a humidity of 76%. The concentration fluctuation feature sequence is a collection of multiple sets of information. Each set of information has a unified structure, which is beneficial for subsequent modules to further analyze and utilize the fluctuation pattern.

[0032] like Figure 2 and Figure 4 As shown, the abnormal data screening module includes: The environmental assessment submodule analyzes wind speed and relative humidity data based on the concentration fluctuation characteristic sequence, determines the trend of wind speed change in each sampling period, compares the direction of relative humidity change, determines the synchronicity of wind speed and relative humidity changes in each period, and calculates the frequency of coordinated change between the two to obtain the coordinated change characteristic quantity. First, the raw wind speed records for each time period are extracted and arranged in chronological order of sampling time. Then, the incremental change in wind speed between consecutive time points is calculated. For example, if the wind speed is 3.1 in the first minute, 3.6 in the second minute, and 3.4 in the third minute, the change between the second and first minutes is +0.5, and the change between the third and second minutes is -0.2. These changes are recorded as a trend sequence. The positive and negative changes in the trend sequence are then categorized by direction: a positive change is marked as increasing, a negative change as decreasing, and a change less than 0.2 is marked as unchanged. Simultaneously, the relative humidity data for the corresponding sampling time period is extracted and processed in the same way to determine the direction of humidity change within adjacent time periods. For example, if the humidity is 75% in the first minute, 78% in the second minute, and 76% in the third minute, the change between the second and first minutes is +3. For an increase, the change in wind speed relative to the change in humidity is -2 in the 3rd minute; for a decrease, the direction of wind speed change is compared with the direction of humidity change. If the two directions are consistent within a certain period, it is considered a synchronous change. A total of 30 sets of sampling data are set within a 30-minute cycle. After performing the above judgment, a label indicating synchronous change in a certain period can be obtained. For example, the 2nd minute is marked as synchronous, and the 3rd minute is not synchronous. Then, the number of time periods marked as synchronous changes in the entire cycle is counted. If it exceeds 40% of the total time periods, i.e., 12 periods, the frequency of coordinated change is 12. This value represents the frequency of occurrence of coordinated changes between the two parameters. If this value exceeds the set coordination threshold, for example, set to 10 times, it indicates that there is a high degree of synchronicity between wind speed and humidity in the current cycle. The final output of the coordinated change feature quantity records are the number of synchronizations, the total number of sampling time periods, and whether the coordination judgment threshold has been reached.

[0033] The change matching submodule compares the changes in each time period with the co-change feature quantity, filters the time periods in which the concentration change amplitude is consistent with the changes in wind speed and relative humidity, determines the synchronous change interval, and obtains the abnormal joint segment sequence. By comparing the concentration changes recorded in each sampling period with the trends of wind speed and humidity during that period, the concentration change values ​​are extracted and their direction of change is determined as either upward or downward. This is then compared with the directions of wind speed and humidity changes in the current sampling period. If an increase in concentration corresponds to an increase in both wind speed and humidity, or a decrease in concentration corresponds to a decrease in both wind speed and humidity, then that period is marked as having all three trends moving in the same direction, meaning the concentration fluctuation is synchronized with the changes in the two environmental parameters. If any two of the three trends are in the same direction, it can also be marked as partially synchronized. If all three trends are inconsistent, it is marked as asynchronous. A filtering threshold is set where the number of times the trends are consistent exceeds two. Synchronous segments, such as the 5th, 6th, and 8th minutes within a cycle, which exhibit the characteristics of rising concentration, rising wind speed, and rising humidity, are identified as synchronous variation intervals. Subsequently, all time periods marked as synchronous variations are extracted to form a joint segment sequence. The start and end times, concentration value changes, wind speed changes, humidity changes, and direction labels of each segment are recorded, and segment numbers are assigned for index management. This sequence is used for subsequent calibration of probe operating status and anomaly troubleshooting steps. The final anomaly joint segment sequence is a set of structured time segments used to identify potential abnormal periods of synchronous fluctuation between concentration and the environment.

[0034] The isothermal calibration submodule determines the heating status of the sampling probe in the abnormal section based on the abnormal joint section sequence, filters the non-isothermal operation section, removes the corresponding abnormal data, and classifies and marks the cause of the abnormality to obtain the abnormal screening dataset. For each marked time period, retrieve the temperature control recording data of the sampling probe within the corresponding time range to determine whether the probe is in a constant heating state during that time period. First, obtain the set constant temperature reference value. For example, if the probe heating set temperature is 180 degrees Celsius, and the allowable fluctuation range is ±3 degrees Celsius, then the judgment condition is that the temperature is maintained between 177 and 183 degrees Celsius to be in a constant temperature state; otherwise, it is in a non-constant temperature state. For example, if the probe temperature is 179 degrees Celsius in the 5th minute, 175 degrees Celsius in the 6th minute, and 178 degrees Celsius in the 8th minute, then the 5th and 8th minutes meet the constant temperature condition, while the 6th minute does not. Mark the sampling time period that does not meet the constant temperature condition. The data segment is marked as non-constant temperature and removed from the concentration data. The removal method is to mark the time segment as invalid in the data record and skip the subsequent analysis steps. At the same time, the reason for the removal of the segment is recorded, and anomaly data classification labels are established. For example, the anomaly reason for the 6th minute is marked as "insufficient heating", the synchronization of wind speed and humidity is marked as "synchronization anomaly", and the concentration fluctuation level is marked as "high amplitude". Finally, an anomaly screening dataset is generated, which records the sampling time segment number, removal reason, wind speed and humidity changes, concentration fluctuation amplitude level and synchronization label of all removed samples, for subsequent statistical analysis and anomaly attribution processing.

[0035] like Figure 2 and Figure 5 As shown, the error interval aggregation module includes: The benchmark comparison submodule analyzes the detection results of each sampling period based on the anomaly screening dataset, compares the detection data of each sampling period with the instrument self-test zero-point benchmark, determines the degree of numerical deviation between the two, records the difference distribution status, and obtains the zero-point offset factor. Extract the pollutant detection concentration values ​​for each time period and retrieve the self-test zero-point reference concentration values ​​recorded by the instrument within that cycle. Compare the numerical differences between the detection results of each sampling time period and the zero-point reference value. During the process, the zero-point reference concentration is initially set to 0.2. If the concentration measured in a certain sampling time period is 0.8, then the offset for that time period is 0.6. This process is repeated to calculate the offset for all valid sampling segments, and these offsets are recorded in chronological order as an offset sequence. Then, the amplitude of each set of difference data is differentiated. Based on the set deviation threshold, the degree of offset is divided into three levels: slight offset is less than 0.5, moderate offset is 0.5 to 1.0, and severe offset is greater than 1.0, for example, an offset value of 0. A score of 0.3 is considered minor, 0.8 is moderate, and 1.2 is severe. Each time period is marked with its offset level, and the frequency and distribution of each level throughout the entire period are statistically analyzed. Then, a trend is marked according to the offset direction of each segment. If three consecutive segments are positive offsets, the trend is upward; if they are negative, the trend is downward; if the directions are intersecting, the trend is mixed. The combination of the above offset levels and trends forms a differential distribution record. For example, in 30 data segments, 8 segments are moderate offsets with an upward trend, and 10 segments are mild offsets with a mixed trend. The three types of information, offset magnitude, offset trend, and frequency of occurrence, are combined into a zero-point offset factor. This factor is output as a label sequence containing differential levels and changing trends.

[0036] The temperature offset submodule compares the temperature change trend corresponding to the sampling period based on the zero-point offset factor, determines the correlation characteristics between temperature change and the zero-point offset factor, filters out data offset phenomena under the influence of temperature, and obtains the amount of temperature influence change. The system retrieves each sampling segment number, its corresponding offset value, and trend marker recorded in the zero-point offset factor. Simultaneously, it retrieves temperature monitoring records for the same time period. First, it extracts temperature values ​​according to the sampling sequence. Then, it performs incremental calculations on the temperature values ​​of adjacent time periods to determine the temperature trend. For example, if the temperature is 25.4 in the first minute, 26.1 in the second minute, and 26.5 in the third minute, then the temperature in the second minute increases by 0.7 compared to the first minute, and in the third minute it increases by 0.4 compared to the second minute, indicating a continuous temperature rise. When the corresponding offset factors are +0.6 and +0.9, it is determined that the temperature rise is consistent with the direction of the offset increase. Subsequently, the temperature trend in all sampling segments is analyzed. The time period in which the trend and the offset trend are consistent is marked as a correlated segment. If the offset trend is opposite to the direction of temperature change, it is marked as an anti-correlated segment. Then, all segments marked as correlated are filtered. If multiple consecutive segments have the same direction and the offset is large, they are further judged as segments dominated by temperature influence. The threshold for judgment is set to 40% of the total number of correlated segments in which the offset and the temperature trend are consistent. For example, if 15 out of 30 segments meet the same direction and the offset is greater than 0.6, they are judged as strongly correlated temperature influence. The segment number, temperature change amplitude and offset direction label are recorded. The output is the change in temperature influence. This change in temperature indicates the range and degree of influence of temperature on the zero-point drift of the detection system.

[0037] The grouping and sorting submodule is based on the change in temperature effect, and groups the samples according to the time and difference information of each sampling segment. It summarizes the degree of deviation of each group during the detection process, counts the difference in change between groups, and obtains the interval deviation feature group. The system receives the sampling segment number and corresponding offset data identified by the temperature change. First, the sampling segments are initially segmented chronologically, with each 10-minute interval forming a detection group. Then, within each group, the offset values ​​of the segments are categorized and organized. Sampling segments with similar offset amplitudes and consistent trend directions are grouped into the same subgroup. For example, if a group has three segments with offset values ​​of +0.7, +0.8, and +0.9, all trending upwards, they are grouped into the "Upward Offset - General Amplitude" subgroup. If there are two other segments with offset values ​​of -1.2 and -1.4, they are grouped into the "Upward Offset - General Amplitude" subgroup. The "Decrease in Offset - Severity Amplitude" subgroup forms a multi-level classification structure for this group. Then, the differences in the average offset values ​​between different groups are compared. The average offset values ​​of each segment within each group are calculated and subtracted to calculate the difference between groups. If the difference exceeds 0.5, it is determined that there is a significant difference between groups. The group number, the average offset within the group, and the difference between groups are recorded. The interval offset feature group is output. The content of the feature group includes the offset level distribution within each sampling group, the offset trend direction, the offset difference with adjacent groups, and the comprehensive offset statistics for all sampling intervals.

[0038] like Figure 2 and Figure 6 As shown, the parameter trend analysis module includes: The span variation submodule acquires the infrared detection channel output data for each detection cycle based on the interval offset feature group, calculates the change of the span parameter in the continuous cycle, determines the change amplitude of the parameter in each cycle, records the span change in each cycle, and obtains the infrared span fluctuation amount. The output concentration data of the infrared detection channel in each cycle is extracted, and its representative output parameters are selected as span reference values. For example, the upper limit of the infrared channel concentration in cycle 1 is 95, the lower limit is 2, and its span is recorded as 93. The upper limit of cycle 2 is 97, the lower limit is 3, and the span is 94. The span values ​​of each cycle are recorded in chronological order to form a span sequence. Then, the span change between consecutive cycles is calculated. Each item is subtracted from the previous item. For example, the change in cycle 2 compared to cycle 1 is +1, cycle 3 is -2, and cycle 4 is +3. Then, the change amplitude is divided into levels according to the set span fluctuation classification criteria. The change amplitude between -2 and +2 is set as a stable segment, greater than 2 is an increasing fluctuation, and less than -2 is a decreasing fluctuation. If the span change of a certain cycle is +3, it is determined to be an increasing fluctuation segment, and +1 is a stable segment. Then, the change amplitude of each cycle is recorded to form an infrared span fluctuation dataset. The dataset contains the detection cycle number, span value, change amount and its corresponding fluctuation level, which is used for subsequent trend analysis.

[0039] The sensitivity trend submodule analyzes the response changes of the ultraviolet differential detection channel within the same period based on the infrared span fluctuation, compares the fluctuation characteristics of sensitivity data with the period, identifies the concentrated stage of sensitivity change, and obtains the ultraviolet sensitivity sequence. The response value of the corresponding UV differential detection channel is matched cycle by cycle. This response value is the measurement signal intensity of the UV channel for the standard pollutant gas. For example, the response value is 52 for cycle 1, 55 for cycle 2, 57 for cycle 3, and decreases to 54 for cycle 4. All cycle response data are formed into a sensitivity time series. The direction and amplitude of the response values ​​of every two consecutive cycles are judged and compared in turn. If the change in response value between two cycles is within ±1, it is marked as no fluctuation. If the change is greater than 1 and is rising, it is marked as enhanced sensitivity. If the change is greater than 1 and is falling, it is marked as weakened sensitivity. A change of +3 between cycle 2 and cycle 1 indicates enhancement, and a change of -3 between cycle 4 and cycle 3 indicates weakening. The cycle number, sensitivity direction, and amplitude of all marked fluctuation segments are recorded. The segments that show the same direction of change in multiple consecutive cycles are selected. The minimum number of consecutive cycles is set to 3 as the threshold for the determination of concentrated fluctuation segments. For example, if the continuous enhancement is from cycle 5 to 7, it is classified as a concentrated enhancement segment. The start and end cycles, response value change range, and direction label of this stage are recorded and summarized into a UV sensitivity sequence.

[0040] The trend comparison submodule is based on the ultraviolet sensitivity sequence. It compares the changing trends of the infrared detection channel and the ultraviolet detection channel in each detection cycle, judges the consistency of the changes between the two, organizes the trend change stage information, and obtains the trend change feature quantity. To determine the consistency of the changes between the two, the following formula is used: ; Calculate the consistency index, organize information on trend change stages, and obtain trend change characteristic quantities. Among them, Representing the Consistency index of the changing trends of the infrared detection channel and the ultraviolet detection channel during the detection cycle. The infrared detection channel represents the first The cycle relative to the first The difference in the span of the cycle, The ultraviolet detection channel represents the first The cycle relative to the first The periodicity is poor.

[0041] Consistency Indicators This refers to the degree of similarity or synchronicity in the changing trends of the infrared and ultraviolet detection channels between the i-th detection period and the previous detection period; the closer the direction and magnitude of the normalized changes of the infrared and ultraviolet detection channels are within the same detection period, the better. The numerical characteristics reflect that the more consistent the trends of the two channels (i.e., the more synchronous the changes in the two channels), the greater the difference in the magnitude of the changes, the opposite the direction, or the one moving and the other still. The numerical characteristics are reflected in the greater the inconsistency between the two trends (i.e., the channel changes are disconnected); it is used to measure whether the changes in the infrared and ultraviolet channels in each cycle are synchronized - the more synchronized (whether they rise together, fall together, or remain stable together). High consistency indicates good alignment, while low consistency indicates poor alignment. This is used to measure whether the infrared and ultraviolet detection channels are "aligned" during the calibration phase, thereby helping to determine the accuracy and reliability of the calibration and identify detection cycles with deviations or anomalies.

[0042] Extract the span variation of the infrared detection channel within each detection cycle, and sequentially call the cycle numbers as follows: and The infrared detection output data is used to calculate the infrared span difference. Simultaneously, the response data of the ultraviolet detection channel under the same period are called to calculate the ultraviolet sensitivity difference. The two parameters mentioned above represent the trend changes of the two detection channels within adjacent periods, and then the infrared normalized difference is obtained through normalization processing. UV normalization difference Substitute the two data points into the consistency index formula and perform term-by-term calculations. First, calculate the numerator: ; The denominator is then calculated as follows: ; The consistency index is: ; The consistency index is divided into three intervals to describe the degree of consistency between the infrared and ultraviolet detection channels in each detection cycle, as follows: when When this occurs, it is classified as a "trend inconsistency phase," which indicates that the two channels change in opposite directions or the magnitude of the changes differs too much, lacking trend coordination characteristics. when At this time, it is classified as the "general consistency stage". At this time, there is a certain degree of matching between the trends of infrared and ultraviolet channels, but a stable coordinated change has not yet been formed. It is characterized by the consistent direction of change in some cycles, but the difference in amplitude is still obvious. when When this period is classified as the "highly consistent stage," it means that the direction and magnitude of change in the infrared and ultraviolet channels are relatively similar, showing obvious trend synchronization characteristics, and can be regarded as a trend stable segment.

[0043] Consistency index is The result indicates that the infrared and ultraviolet detection channels have strong consistency in normalized changes within the current detection cycle. The formula, by introducing two different physical channel parameter change terms and participating in the absolute value and subtraction square processing respectively, achieves the ability to simultaneously measure the direction and magnitude of change. Furthermore, by using square root operations to construct a normalized reference basis, the consistency and robustness of trend judgment between channels are improved.

[0044] like Figure 2 and Figure 7 As shown, the periodic stability determination module includes: The continuous trend submodule analyzes the growth phase of the cycle span parameter within a continuous cycle based on the trend change characteristic quantity, judges the continuity and fluctuation characteristics of parameter changes, counts the continuous growth period, and obtains the number of continuous growth intervals. The process extracts span values ​​periodically and forms a span sequence in chronological order. During execution, it first determines whether the span values ​​within any three consecutive periods show a unidirectional increasing relationship. If so, it is recorded as a growth segment. The judgment criteria are that the span value is continuously greater than the previous period, and the single increment is not less than 1. For example, the span value in the first period is 93, the second period is 95, and the third period is 98. By comparing them sequentially, the growth amounts are +2 and +3. The condition is met twice in a row, and this segment is marked as a continuous growth stage. If the fourth period is 97, showing a downward trend, the growth segment is interrupted. Then, it is re-judged from the fifth period to determine whether a new continuous growth segment has been formed. This process is repeated until all periods have been traversed. The number of all period segments that meet the continuous growth condition is counted, and the starting period and total span change value of each segment are recorded. The number of continuous growth intervals output is the total number of validly marked growth segments, which helps to identify the cumulative trend deviation of the device during operation.

[0045] The mean offset submodule analyzes the degree of deviation between the span parameters of each detection cycle and the cycle mean based on the number of continuously increasing intervals, determines the relationship between the parameter change trend in each cycle and the cycle mean, filters out the time periods that are inconsistent with the mean change, and obtains the cycle offset trend number. Calculate the average span of each period segment, setting the comparison window to a 5-period sliding window mode. Calculate the average span value within each window group, and then perform a difference calculation between the span parameter corresponding to each detection period and the average value within its window to determine whether the period deviates from the average level. The deviation judgment standard is set as follows: if the difference between the single-period span value and the average value is greater than 2, it is recorded as a deviation period. For example, the average span in the 6th to 10th period window is 95, and the span in the 8th period is 98, with a deviation of +3. Therefore, the 8th period is judged as a deviation period, and the deviation direction and magnitude are recorded. Subsequently, the number of deviation periods in all detection periods is counted, and the consistency between their changing trends and the overall average trend is further analyzed. For example, if the overall average trend is decreasing in a certain segment, but the span value of a certain period increases and deviates from the average, it is marked as a period with inconsistent trends. Finally, all period segments with inconsistent trends are screened out, and their numbers, deviation values, and deviation directions are recorded to form the period deviation trend number. This value represents the number of all abnormal trends found in the overall trend consistency judgment.

[0046] The anomaly screening submodule uses the period offset trend number to determine the deviation of the span parameter change range from the normal change range within the period segment, counts the number of period segments that are screened out, and classifies and summarizes the abnormal period segments to obtain the stability screening statistics. The variation range of each span parameter is judged. First, the normal variation range threshold is set to within ±2. The judgment method is to compare the difference between the span values ​​of two consecutive periods and see if it is within this range. If the variation range of a certain period exceeds this range, it is judged as an abnormal variation period. For example, the span of the 11th period is 95 and the span of the 12th period is 99, with a variation range of +4, which exceeds the threshold and is classified as an abnormal period segment. All abnormal period segments are numbered and recorded, and the intersection is processed with the trend inconsistency segments recorded in the period offset trend number. If a period segment belongs to both trend inconsistency and span variation range exceeds the standard, it is classified as a multiple abnormal period segment. The number of all screened period segments is counted and divided into single abnormal and multiple abnormal categories. The proportion and position index of each type of abnormal segment are summarized, and the stability screening statistics are output.

[0047] like Figure 8 As shown, a calibration and testing method for a flue gas detector includes the following steps: S1: Based on the flue gas emission channel, analyze the concentration data obtained at each sampling time within the continuous calibration cycle, compare the detection results corresponding to two adjacent samplings, calculate the difference between each group of data by subtraction, collect all differences according to the sampling time sequence, and summarize the data fluctuations within the detection cycle by group to obtain the concentration fluctuation characteristic sequence. S2: Based on the concentration fluctuation feature sequence, determine the wind speed and relative humidity corresponding to each sampling period, analyze the coordinated change law of wind speed and relative humidity, screen out abnormal segments that are related to the change amplitude of concentration difference and the change of environmental parameters, calibrate the constant temperature operation status of the flue gas sampling probe, and obtain the abnormal screening dataset. S3: Based on the anomaly screening dataset, compare the differences between the sampling period and the instrument self-test zero point reference, analyze the data offset caused by temperature changes during the sampling period, sort out the difference information of all sampling segments in groups, summarize the deviation of each group of detection segments, and obtain the interval offset feature group. S4: Based on the interval offset feature group, calculate the span parameter change of the infrared detection channel output within the continuous detection period, analyze the response change of the ultraviolet differential detection channel sensitivity in the same period, screen the stage with concentrated span change amplitude, and compare the trends of infrared and ultraviolet detection channels to obtain the trend change feature quantity. S5: Based on the trend change characteristic, determine the stage of continuous growth of the span parameter of each detection cycle, analyze the deviation of the change trend of the span parameter of each cycle from the mean, screen out segments with irregular change amplitude within the cycle segment, count the number of cycle segments that are screened out, and obtain the stability screening statistics.

[0048] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A flue gas detector, characterized in that, The detector includes: The concentration fluctuation analysis module is based on the flue gas emission channel. It analyzes the sampling data in each continuous calibration cycle, calculates the differences between adjacent time points, summarizes the time series differences, summarizes the fluctuation characteristics in each cycle, and obtains the concentration fluctuation characteristic sequence. Based on the concentration fluctuation feature sequence, the abnormal data screening module determines the wind speed and humidity corresponding to the sampling period, analyzes their coordinated changes, filters abnormal data related to changes in environmental parameters, calibrates the constant temperature state of the sampling probe, and obtains the abnormal screening dataset. The error interval aggregation module, based on the anomaly screening dataset, compares the difference between the sampling period and the self-test zero-point benchmark, analyzes the data offset in conjunction with temperature changes, groups and organizes the differences, summarizes the deviation of the detection segment, and obtains the interval offset feature group. Based on the interval offset feature group, the parameter trend analysis module analyzes the sensitivity response of the ultraviolet detection channel, filters out the stages with concentrated span changes, compares the channel trends, and obtains the trend change feature quantity. The periodic stability discrimination module analyzes the deviation of the parameter change trend from the mean within the period based on the trend change feature quantity, filters out irregular segments, statistically removes periodic segments, and obtains stability screening statistics.

2. The flue gas detector according to claim 1, characterized in that, The concentration fluctuation feature sequence includes segment identifier, fluctuation mean, and rate of change; the anomaly screening dataset includes screening label, associated time period, and anomaly attribute; the interval offset feature group includes offset type, corresponding interval, and offset degree; the trend change feature includes span trend, sensitivity trend, and periodic feature; and the stability screening statistics include number of removals, stability classification, and anomaly ratio.

3. The flue gas detector according to claim 1, characterized in that, The concentration fluctuation analysis module includes: The difference calculation submodule analyzes the concentration detection data at each sampling time within a continuous calibration cycle based on the flue gas emission channel, compares the detection results of adjacent sampling periods, calculates the difference between the subsequent detection data and the previous detection data, organizes the difference data, and obtains the concentration change amplitude sequence. The time-series aggregation submodule determines the continuity of the detection sampling time based on the concentration change amplitude sequence, removes sampling interruption segments, and reorders the remaining data according to the sampling time to obtain a time-series difference arrangement sequence. The fluctuation induction submodule arranges the sequence based on the time period differences, analyzes the direction of change and the magnitude of data increase or decrease of adjacent data segments, aggregates the correlation parameters of the sampling segments, adjusts the data grouping, and summarizes the dynamic characteristics of each segment to obtain the concentration fluctuation characteristic sequence.

4. The flue gas detector according to claim 1, characterized in that, The abnormal data screening module includes: The environmental judgment submodule analyzes wind speed data and relative humidity data based on the concentration fluctuation characteristic sequence, judges the trend of wind speed change in each sampling period, compares the direction of relative humidity change, determines the synchronicity of wind speed and relative humidity changes in each period, and calculates the frequency of coordinated change of the two to obtain the coordinated change characteristic quantity. The change matching submodule compares the changes in each time period with the coordinated change feature quantity, filters out the time periods in which the concentration change amplitude is consistent with the changes in wind speed and relative humidity, determines the synchronous change interval, and obtains the abnormal joint segment sequence. The isothermal calibration submodule determines the heating status of the sampling probe in the abnormal section based on the abnormal joint segment sequence, filters out non-isothermal operation segments, removes the corresponding abnormal data, and classifies and marks the causes of the abnormality to obtain an abnormal screening dataset.

5. The flue gas detector according to claim 1, characterized in that, The error interval aggregation module includes: The benchmark comparison submodule analyzes the detection results of each sampling period based on the anomaly screening dataset, compares the detection data of each sampling period with the instrument self-test zero-point benchmark, determines the degree of numerical deviation between the two, records the difference distribution state, and obtains the zero-point offset factor. The temperature offset submodule compares the temperature change trend corresponding to the sampling period based on the zero-point offset factor, determines the correlation characteristics between temperature change and the zero-point offset factor, filters out data offset phenomena under the influence of temperature, and obtains the amount of temperature influence change. The grouping and sorting submodule groups the samples based on the temperature change and the time and difference information of each sampling segment, summarizes the degree of deviation of each group during the detection process, and statistically analyzes the changes between groups to obtain the interval deviation feature group.

6. The flue gas detector according to claim 1, characterized in that, The parameter trend analysis module includes: The span variation submodule acquires the infrared detection channel output data for each detection cycle based on the interval offset feature group, calculates the change of the span parameter in the continuous cycle, determines the change amplitude of the parameter in each cycle, and records the span variation in each cycle to obtain the infrared span fluctuation amount. The sensitivity trend submodule analyzes the response changes of the ultraviolet differential detection channel within the same period based on the infrared span fluctuation amount, compares the fluctuation characteristics of sensitivity data with the period, identifies the concentrated stage of sensitivity change, and obtains the ultraviolet sensitivity sequence. The trend comparison submodule compares the changing trends of the infrared detection channel and the ultraviolet detection channel in each detection cycle based on the ultraviolet sensitivity sequence, judges the consistency of the changes between the two, organizes the trend change stage information, and obtains the trend change feature quantity.

7. The flue gas detector according to claim 1, characterized in that, The periodic stability determination module includes: The continuous trend submodule analyzes the growth stage of the period span parameter within a continuous period based on the trend change characteristic quantity, judges the continuity and fluctuation characteristics of parameter changes, counts the continuous growth period, and obtains the number of continuous growth intervals. The mean offset submodule analyzes the degree of deviation between the span parameter of each detection cycle and the cycle mean based on the number of continuous growth intervals, determines the relationship between the parameter change trend in each cycle and the cycle mean, filters out the time periods that are inconsistent with the mean change, and obtains the cycle offset trend number. The anomaly screening submodule determines the deviation of the span parameter change range from the normal change range within the period segment based on the period offset trend number, counts the number of period segments that are screened out, and classifies and summarizes the abnormal period segments to obtain the stability screening statistics.

8. The flue gas detector according to claim 1, characterized in that, The continuous calibration cycle refers to each complete cycle in which the flue gas detector performs multiple continuous automatic or manual calibrations of the detection system at set intervals.

9. The flue gas detector according to claim 1, characterized in that, The synergistic change refers to the trend of wind speed and relative humidity changing together over time.

10. A calibration and testing method for a flue gas detector, the method being used to implement the flue gas detector as described in any one of claims 1-9, characterized in that, Includes the following steps: S1: Based on the flue gas emission channel, analyze the concentration data obtained at each sampling time within the continuous calibration cycle, compare the detection results corresponding to two adjacent samplings, calculate the difference between each group of data by subtraction, collect all differences according to the sampling time sequence, and summarize the data fluctuations within the detection cycle by group to obtain the concentration fluctuation characteristic sequence. S2: Based on the concentration fluctuation feature sequence, determine the wind speed and relative humidity corresponding to each sampling period, analyze the coordinated change law of wind speed and relative humidity, screen out abnormal segments that are related to the change amplitude of concentration difference and the change of environmental parameters, calibrate the constant temperature operation status of the flue gas sampling probe, and obtain the abnormal screening dataset. S3: Based on the anomaly screening dataset, compare the differences between each sampling period and the instrument self-test zero-point reference, analyze the data offset caused by temperature changes during the sampling period, sort out the difference information of all sampling segments in groups, summarize the deviation of each group of detection segments, and obtain the interval offset feature group. S4: Based on the interval offset feature group, calculate the span parameter change of the infrared detection channel output within the continuous detection period, analyze the response change of the ultraviolet differential detection channel sensitivity in the same period, screen the stage with concentrated span change amplitude, and compare the trends of infrared and ultraviolet detection channels to obtain the trend change feature quantity. S5: Based on the trend change characteristic quantity, determine the stage of continuous growth of the span parameter of each detection cycle, analyze the deviation of the change trend of the span parameter of each cycle from the mean, screen out segments with irregular change amplitude within the cycle segment, count the number of cycle segments that are screened out, and obtain the stability screening statistics.

Images