Flue gas emission monitoring method and system for thermal power plant

Abstract

The invention discloses a flue gas emission monitoring method and system for a thermal power plant, and relates to the technical field of flue gas emission monitoring, and the method comprises the steps: carrying out the flue gas emission data collection of a target thermal power plant, and generating a flue gas emission monitoring data time sequence; performing fluctuation state analysis on each flue gas element in the flue gas to generate each fluctuation index; configuring monitoring dual channels based on each fluctuation index; extracting flue gas emission data of the target thermal power plant and performing long-term abnormal standard exceeding judgment to generate a long-term abnormal standard exceeding judgment result; and generating a flue gas emission monitoring result according to the long-term abnormal standard exceeding judgment result. The technical problems that in the prior art, flue gas emission monitoring data is low in accuracy and reliability, and abnormal standard exceeding early warning is not timely are solved, and the technical effects that monitoring accuracy, reliability and timeliness are improved, it is guaranteed that flue gas emission of the thermal power plant meets the environmental protection requirement, and potential standard exceeding risks are found in time are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of flue gas emission monitoring, and in particular to a flue gas emission monitoring method and system for a thermal power plant. Background Art

[0002] In today's society, environmental protection requirements are becoming increasingly stringent. As an important energy production unit, the flue gas emissions of thermal power plants have a profound impact on environmental quality. Traditional flue gas emission monitoring technology faces many challenges. On the one hand, the accuracy of monitoring data is difficult to guarantee. Due to the limitations of sensor accuracy, single data acquisition frequency and data processing algorithms, the acquired flue gas emission data has large errors. For example, in a complex flue gas environment, a variety of interference factors will cause the sensor measurement value to deviate from the true value, and the traditional technology lacks an effective error correction mechanism, which greatly reduces the accuracy of the data. On the other hand, the reliability of monitoring data is also worrying. The stability and consistency of the data are poor. The data collected in different batches or different time periods may have large deviations and cannot truly reflect the actual situation of flue gas emissions. More seriously, the problem of untimely warning of abnormal exceeding of the standard is common. Most of the existing technologies are based on simple threshold comparison to determine whether the standard is exceeded, and lack in-depth analysis and prediction capabilities for the changing trend of emission data. When flue gas emissions show slow abnormal changes and gradually tend to exceed the standard, it is impossible to detect and warn in advance.

[0003] Existing technologies have technical problems such as low accuracy and reliability of flue gas emission monitoring data, and untimely warning of abnormal exceeding of standards. Summary of the invention

[0004] The present application provides a flue gas emission monitoring method and system for a thermal power plant, which is used to solve the technical problems in the prior art of low accuracy and reliability of flue gas emission monitoring data, and untimely warning of abnormal exceeding of standards.

[0005] In view of the above problems, the present application provides a flue gas emission monitoring method and system for a thermal power plant.

[0006] In a first aspect of the present application, a method for monitoring flue gas emissions in a thermal power plant is provided, the method comprising:

[0007] The flue gas emission data of the target thermal power plant is collected during a preset monitoring period to generate a time series of flue gas emission monitoring data; based on the time series of flue gas emission monitoring data, the fluctuation state of each flue gas element in the flue gas is analyzed to generate each fluctuation index; a dual monitoring channel is configured based on each fluctuation index; the flue gas emission data of the target thermal power plant is extracted through the dual monitoring channel and a long-term abnormal excess judgment is performed to generate a long-term abnormal excess judgment result; the flue gas emission monitoring result is generated based on the long-term abnormal excess judgment result.

[0008] A second aspect of the present application provides a flue gas emission monitoring system for a thermal power plant, the system comprising:

[0009] A flue gas emission monitoring data timing generation module, the flue gas emission monitoring data timing generation module is used to collect flue gas emission data of the target thermal power plant during a preset monitoring period and generate a flue gas emission monitoring data timing; a fluctuation index generation module, the fluctuation index generation module performs fluctuation state analysis on each flue gas element in the flue gas based on the flue gas emission monitoring data timing and generates each fluctuation index; a monitoring dual-channel configuration module, the monitoring dual-channel configuration module configures the monitoring dual channel based on each fluctuation index; an abnormal exceeding standard judgment result generation module, the abnormal exceeding standard judgment result generation module is used to extract the flue gas emission data of the target thermal power plant through the monitoring dual channel and perform long-term abnormal exceeding standard judgment to generate a long-term abnormal exceeding standard judgment result; a flue gas emission monitoring result generation module, the flue gas emission monitoring result generation module is used to generate a flue gas emission monitoring result based on the long-term abnormal exceeding standard judgment result.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] During the preset monitoring period, the target thermal power plant is subjected to flue gas emission data collection to generate a flue gas emission monitoring data time series; based on the flue gas emission monitoring data time series, the flue gas elements in the flue gas are subjected to fluctuation state analysis to generate various fluctuation indicators; based on the various fluctuation indicators, a dual monitoring channel is configured; the flue gas emission data of the target thermal power plant is extracted and long-term abnormal exceeding standard judgment is performed to generate a long-term abnormal exceeding standard judgment result; the flue gas emission monitoring result is generated based on the long-term abnormal exceeding standard judgment result. The technical effect of improving the accuracy, reliability and timeliness of monitoring is achieved, which helps to ensure that the flue gas emissions of thermal power plants meet environmental protection requirements and timely discover potential exceeding standard risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 A schematic flow chart of a flue gas emission monitoring method for a thermal power plant provided in an embodiment of the present application;

[0014] Figure 2 A schematic diagram of the structure of a flue gas emission monitoring system for a thermal power plant provided in an embodiment of the present application.

[0015] Explanation of the reference numerals: monitoring data time series generation module 10 , fluctuation index generation module 20 , monitoring dual-channel configuration module 30 , abnormal exceeding standard judgment result generation module 40 , flue gas emission monitoring result generation module 50 . DETAILED DESCRIPTION

[0016] The present application provides a flue gas emission monitoring method and system for a thermal power plant, which is used to solve the technical problems in the prior art of low accuracy and reliability of flue gas emission monitoring data, and untimely warning of abnormal exceeding of standards.

[0017] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] Embodiment 1, as Figure 1 As shown, the present application provides a method for monitoring flue gas emissions of a thermal power plant, the method comprising:

[0019] Step S100: collecting flue gas emission data from a target thermal power plant during a preset monitoring period to generate a flue gas emission monitoring data time series.

[0020] Specifically, during the preset monitoring period, a high-precision sensor network deployed at key locations of the flue gas emission pipeline of the target thermal power plant is used to collect key data such as the concentration, flow rate, temperature, etc. of various components in the flue gas, such as sulfur dioxide, nitrogen oxides, particulate matter, and other flue gas elements in real time, continuously, and accurately. These sensors will arrange and integrate the massive amount of raw data collected in an orderly manner according to precise time stamps. Each time point corresponds to a complete set of flue gas element data information, thereby gradually constructing a flue gas emission monitoring data time series. This data time series completely records the dynamic changes in the flue gas emissions of the target thermal power plant during the entire preset monitoring period, providing an indispensable basic data resource for subsequent in-depth analysis of flue gas emission characteristics, evaluation of emission stability, and judgment of whether there are abnormal exceeding conditions.

[0021] Step S200: performing a fluctuation state analysis on each flue gas element in the flue gas based on the flue gas emission monitoring data time series to generate each fluctuation index.

[0022] Specifically, in the process of analyzing the fluctuation state of each flue gas element based on the flue gas emission monitoring data time series and generating fluctuation indicators, first, the data segmentation algorithm is used to accurately divide the overall flue gas emission monitoring data time series into each independent element monitoring time series according to the type difference of the flue gas elements, so as to ensure that the data of each element can be analyzed and processed independently. For the identification of the intensity of fluctuations, a fluctuation analysis algorithm based on statistical principles is adopted. When identifying the frequent fluctuation amplitude of the element, the absolute value of the deviation between the data point and the mean in the element monitoring time series is calculated, and the larger deviation value is selected to determine the frequent fluctuation amplitude. In terms of identifying the frequent fluctuation frequency of the element, the element monitoring time series is converted from the time domain to the frequency domain with the help of Fourier transform technology. In the frequency domain signal, the frequency component identification algorithm is used to accurately locate the part with a frequency component greater than the preset threshold, and then the mean of this part of the frequency is calculated to obtain the frequent fluctuation frequency of each element. When generating fluctuation indicators, the standardization algorithm is used to standardize the frequent fluctuation amplitude and frequency and convert them into dimensionless data. For example, for the fluctuation amplitude, the range standardization or standard deviation standardization method can be used to unify its numerical range. Afterwards, the standardized fluctuation amplitude and frequency are weighted and summed according to the weighting algorithm determined in advance through a large number of experiments and data analysis. The determination of the weighting coefficient comprehensively considers the relative importance of the fluctuation amplitude and frequency on the overall fluctuation state, thereby generating a fluctuation index that can accurately reflect the fluctuation state of each flue gas element, providing a scientific and reliable basis for the reasonable configuration of the subsequent dual-channel monitoring.

[0023] Step S300: configuring a dual-channel monitoring system based on the various fluctuation indicators.

[0024] Specifically, when configuring the dual-channel monitoring based on each fluctuation index, two different sampling frequencies are first set, namely the first sampling frequency and the second sampling frequency, wherein the second sampling frequency is higher than the first sampling frequency, to meet the monitoring requirements of flue gas emission data at different time resolutions. Then, relying on the slowfast network, a network architecture for monitoring dual channels is constructed, and the first sampling frequency and the second sampling frequency are used to widely collect data for each flue gas element, and each first element emission sample data and multiple second element emission sample data are obtained. At the same time, the corresponding long-term abnormal excess risk identification data are collected, so as to train and optimize multiple initial monitoring dual channels. Then, each fluctuation index is carefully compared with the preset fluctuation index. For the first flue gas element whose fluctuation index is greater than the preset fluctuation index, accurate matching and screening are performed from the trained initial monitoring dual channels to determine its corresponding monitoring channel, thereby completing the configuration of the monitoring dual channels, so that the monitoring system can flexibly adjust the monitoring strategy according to the fluctuation characteristics of the flue gas elements, more efficiently capture abnormal information in the flue gas emission data, and improve the accuracy and reliability of the entire monitoring process.

[0025] Step S400: extracting the flue gas emission data of the target thermal power plant through the dual monitoring channels and performing long-term abnormal exceeding standard judgment to generate a long-term abnormal exceeding standard judgment result.

[0026] Specifically, in the process of extracting the flue gas emission data of the target thermal power plant through monitoring dual channels and making long-term abnormal exceeding judgments, the long short-term memory network (LSTM) algorithm is used to input the first sampling data set and the second sampling data set into the LSTM network in sequence according to the time series. With its unique memory unit structure, the LSTM network can effectively learn the dependence and change law of flue gas emission data on long time series, and accurately model the long-term trend and short-term fluctuation in the data. In the network training process, a large amount of historical flue gas emission data and the corresponding long-term abnormal exceeding labeling results are used as training sets. The weight parameters of the network are continuously adjusted through the back propagation algorithm, so that the network can accurately identify the association between different data patterns and long-term abnormal exceeding conditions. When the combination of flue gas emission data of the target thermal power plant to be detected is input, the LSTM network analyzes the data according to the learned pattern, calculates the degree of data deviation from the normal range, and compares it with the preset abnormal threshold. If the calculation result exceeds the threshold, it is determined that there is a long-term abnormal exceeding condition, thereby generating the corresponding long-term abnormal exceeding judgment result, providing a reliable basis for subsequent environmental supervision and enterprise response measures.

[0027] Step S500: Generate a flue gas emission monitoring result based on the long-term abnormal exceeding standard judgment result.

[0028] Specifically, taking the long-term abnormal exceeding judgment result obtained through rigorous analysis as the core basis, the relevant data information is fully integrated and sorted out, and then detailed and standardized flue gas emission monitoring results are generated. This monitoring result not only clearly presents the final judgment conclusion on whether the flue gas emission of the target thermal power plant has long-term abnormal exceeding, but also covers a wealth of auxiliary information. For example, for the flue gas elements that are judged to be abnormally exceeding for a long time, the start and end time of the exceeding period, the specific numerical range of the exceeding data, the quantitative indicators of the exceeding degree, etc. are recorded in detail; at the same time, for the flue gas elements that have not exceeded the standard, the emission data statistical information during the monitoring period is also provided, such as the average concentration, the concentration change range, etc. In addition, some key parameter information in the data collection process can also be included, such as the operating status of the monitoring equipment, the time interval of data collection, etc. It realizes the conversion of the long-term abnormal exceeding judgment result into a comprehensive, intuitive and multi-dimensional flue gas emission monitoring result, improves the integrity, accuracy and practicality of the flue gas emission monitoring result, enables the monitoring result to more accurately reflect the flue gas emission status of the thermal power plant, and enhances the positioning and quantification capabilities of the long-term abnormal exceeding situation.

[0029] In a possible implementation, step S200 further includes:

[0030] Step S210: Segment the flue gas emission monitoring data time series according to flue gas elements to generate monitoring time series for each element.

[0031] Step S220: Identify the severity of fluctuations in the monitoring time series of each element and generate the fluctuation indicators.

[0032] Specifically, a clustering segmentation algorithm based on feature identifiers is used to segment the flue gas emission monitoring data series according to the flue gas elements, and a unique feature identifier is set for each known flue gas element. These identifiers are determined based on the physicochemical properties of the element or a specific data pattern. Then, the entire flue gas emission monitoring data series is traversed, and for each data point, its feature information related to the flue gas element is extracted and matched and compared with the pre-set feature identifiers. Once the match is successful, the data point is classified into the corresponding flue gas element data set. By repeating this process, the entire data series is gradually divided into independent element monitoring series, ensuring that each element monitoring series only contains data of specific flue gas elements, thereby providing a clear structure and non-interfering data basis for the subsequent precise analysis of each element, greatly improving the accuracy and efficiency of the analysis.

[0033] The process of identifying the severity of fluctuations in the monitoring time series of each element and generating fluctuation indicators is as follows: first, based on the monitoring time series of each element, determine the frequent fluctuation amplitude of each element in the form of root mean square fluctuation, calculate the mean of the element monitoring time series, sum the squares of the differences between each data point and the mean and divide by the total number of data points and take the square root to obtain the root mean square fluctuation amplitude, which comprehensively reflects the fluctuation amplitude and energy level of the element emission data relative to the mean. Next, use spectrum analysis technology to identify the frequent fluctuation frequency of each element based on the element monitoring time series, that is, perform Fourier transform on the time series to obtain the frequency domain signal, screen out the frequency interval with higher energy or amplitude, and determine its center frequency or main frequency component as the frequent fluctuation frequency. Afterwards, standardize the frequent fluctuation amplitude and frequent fluctuation frequency of each element respectively to eliminate dimensional differences. For example, the fluctuation amplitude is standardized by range standardization or standard deviation standardization, and the fluctuation frequency is also processed accordingly. Finally, based on the weighting coefficients determined in advance through a large number of experiments and data analysis, the standardized fluctuation amplitude and frequency are weighted and summed to generate a fluctuation index that can accurately characterize the severity of the monitoring time series fluctuations and the stability of the state of each element, providing a key basis for subsequent monitoring and analysis.

[0034] In a possible implementation, step S220 further includes:

[0035] Step S221: identifying the frequent fluctuation amplitude of each element based on the monitoring time sequence of each element.

[0036] Step S222: identifying the frequent fluctuation frequency of each element based on the monitoring timing of each element.

[0037] Step S223: The frequent fluctuation amplitudes of the various elements and the frequent fluctuation frequencies of the various elements are normalized and weighted to generate the various fluctuation indicators.

[0038] Specifically, for each element monitoring time series, the root mean square fluctuation method is used to identify the frequent fluctuation amplitude of each element. First, the mean of the element monitoring time series is calculated. Then, for each data point in the time series, the square of its difference from the mean is calculated. All these difference squares are summed and divided by the total number of data points to obtain the mean square value. Finally, the square root of the mean square value is taken, and the result is the root mean square fluctuation amplitude of the element monitoring time series. This root mean square fluctuation amplitude can comprehensively reflect the fluctuation amplitude of the element emission data relative to the mean in the entire monitoring time series, highlighting the overall energy level and average degree of change of the fluctuation. Compared with simple maximum deviation and other methods, it more comprehensively characterizes the violent characteristics of element fluctuations, and provides a reliable and representative frequent fluctuation amplitude indicator for the subsequent accurate evaluation of the stability of element emissions and comprehensive analysis with other fluctuation characteristics.

[0039] Based on the monitoring time series of each element, spectrum analysis technology is used to identify the frequent fluctuation frequency of each element, and fast Fourier transform is performed on the element monitoring time series to convert the time domain signal into the frequency domain signal, so as to obtain the distribution of the emission data of the element in the frequency domain. Next, a series of preset frequency intervals are determined in the frequency domain, and the energy or amplitude in each frequency interval is counted to reflect the relative importance of different frequency components in the entire monitoring time series. Then, the frequency intervals with relatively high energy or amplitude are screened out, and the central frequency or main frequency component in these intervals is determined as the significant fluctuation frequency of the element. By analyzing the number of occurrences, distribution patterns and persistence of these significant fluctuation frequencies throughout the monitoring process, the frequent fluctuation frequencies of each element can be accurately identified. For example, if the emission data of an element frequently fluctuates with high energy in a specific frequency interval during a certain period of time, the central frequency of the frequency interval is identified as one of the frequent fluctuation frequencies of the element during this period, which helps to gain a deeper understanding of the periodicity and regular changes of element emissions and their potential associations with other factors, and provides a key basis for a comprehensive assessment of the fluctuation characteristics of element emissions.

[0040] The frequent fluctuation amplitude and frequency of each element are standardized. For the fluctuation amplitude, the formula Processing, where is the original fluctuation amplitude, μ A and σ A are the mean and standard deviation of the fluctuation amplitude, respectively. In this way, the fluctuation amplitude is standardized to a specific range. Similarly, for the fluctuation frequency, the formula is used Standardization, where is the original oscillation frequency, μ C and σ C They are the mean and standard deviation of the fluctuation frequency. After the standardization process is completed, the standardized fluctuation amplitude and fluctuation frequency are weighted. This weighting process determines the weight value based on actual needs and experience, and adds the standardized fluctuation amplitude and fluctuation frequency according to their respective weights to finally generate a fluctuation index that can comprehensively reflect the fluctuation of each element. This fluctuation index can more accurately describe the fluctuation characteristics of the element in the monitoring time series, and provide a reliable basis for subsequent monitoring and analysis.

[0041] In a possible implementation, step S300 further includes:

[0042] Step S310: configuring a first sampling frequency and a second sampling frequency, wherein the second sampling frequency is greater than the first sampling frequency.

[0043] Step S320: Based on the slowfast network, a network structure for monitoring dual channels is constructed, and each first element emission sample data, multiple second element emission sample data and corresponding long-term abnormal exceeding risk identification data are collected for each flue gas element according to the first sampling frequency and the second sampling frequency, and multiple initial monitoring dual channels are trained and constructed.

[0044] Step S330: determining whether each fluctuation index configuration is greater than a preset fluctuation index, matching the corresponding initial monitoring dual channels for the first flue gas element whose fluctuation index configuration is greater than the preset fluctuation index, and generating the monitoring dual channels.

[0045] Specifically, the sampling frequency is configured. The first sampling frequency and the second sampling frequency are determined, and the second sampling frequency is higher than the first sampling frequency. This configuration is designed based on the demand for flue gas emission monitoring. The first sampling frequency is relatively low, and its function is to sample the flue gas emission data at a wider time interval, so as to capture the general trend of changes in flue gas emissions over a longer period of time, such as collecting data every 10 minutes. The second sampling frequency is higher, and it can sample at shorter time intervals, such as collecting data every 1 minute, so that it can accurately capture the subtle fluctuations and rapid changes in the flue gas emission data in a short period of time. By setting these two different sampling frequencies, the flue gas emission data can be collected from different time scales, laying the foundation for subsequent more comprehensive and accurate analysis and monitoring of flue gas emissions.

[0046] The network structure of monitoring dual channels is constructed based on the slowfast network. The slowfast network has a unique dual-path architecture and can process information at different time scales well. Data is collected for each flue gas element according to the configured first sampling frequency and second sampling frequency. For each flue gas element, the emission sample data of each first element is collected at the first sampling frequency. These data reflect the emission of the flue gas element at a relatively long time interval. At the same time, multiple second element emission sample data are collected at the second sampling frequency. The data can capture the emission fluctuations of the flue gas element in a shorter time interval. In addition, it is necessary to collect the long-term abnormal excess risk identification data corresponding to these sample data. The identification data can indicate whether the sample data has a long-term abnormal excess. Finally, the collected first element emission sample data, second element emission sample data and long-term abnormal excess risk identification data are used to train and construct multiple initial monitoring dual channels. During the training process, by continuously adjusting the network parameters, these initial monitoring dual channels can accurately judge whether the flue gas element has a long-term abnormal excess according to the input sample data, thereby providing a basis for subsequent precise monitoring.

[0047] The fluctuation index of each flue gas element is compared and judged with the preset fluctuation index one by one. The preset fluctuation index is a benchmark value set based on a large amount of historical data and relevant environmental standards, which is used to measure the normal range of fluctuation of flue gas elements. When checking each fluctuation index, if it is found that the fluctuation index of a flue gas element (defined as the first flue gas element) is greater than the preset fluctuation index, this indicates that the emission fluctuation of the element is relatively violent, and there is a high risk of long-term abnormal exceeding the standard. For the first flue gas element whose fluctuation index is greater than the preset fluctuation index, it is necessary to match it from the multiple initial monitoring dual channels that have been trained before. This matching process is based on the monitoring ability of each initial monitoring dual channel for flue gas elements with different fluctuation characteristics learned during the training process. Through precise matching, the initial monitoring dual channel that is most suitable for monitoring the first flue gas element is found and determined, and then the monitoring dual channel that is finally used for flue gas emission monitoring of the target thermal power plant is generated. Such a dual-channel configuration can be targeted to monitor those flue gas elements with violent fluctuations and higher risks efficiently and accurately, ensuring that possible long-term abnormal exceeding of the standard can be discovered in time.

[0048] In a possible implementation, step S400 further includes:

[0049] Step S410: sampling a first flue gas element in the flue gas emission data of the target thermal power plant according to a first sampling frequency to generate a first sampling data set.

[0050] Step S420: sampling the first flue gas element in the flue gas emission data of the target thermal power plant according to a second sampling frequency to generate a plurality of second sampling data sets.

[0051] Step S430: Combining the first sampling data set and the second sampling data set respectively, and inputting them into the monitoring dual channels for analysis, and outputting the long-term abnormal exceeding judgment result.

[0052] Specifically, for the first flue gas element in the flue gas emission data of the target thermal power plant, a system sampling operation is carried out according to a pre-set first sampling frequency. This sampling frequency is relatively low, and aims to capture the overall change trend of the element emission data over a longer time span. By collecting data points at equal time intervals, these data points are aggregated and integrated to generate the first sampling data set, presenting a more macroscopic change outline.

[0053] The same first flue gas element is sampled using a second sampling frequency, which is significantly higher than the first sampling frequency. This frequency can obtain data in a shorter time interval, thereby accurately capturing the subtle fluctuations and rapid changes in emission data in a short period of time. By using this high-frequency sampling method multiple times, multiple second sampling data sets are obtained, showing the emission fluctuations in a short period of time in detail.

[0054] The first sampling data set obtained according to the first sampling frequency and the multiple second sampling data sets obtained according to the second sampling frequency are orderly combined. This combination method can integrate the flue gas emission data information at different time resolutions to form data input samples that contain both long-term trends and short-term fluctuation details. Then, these combined samples are input into the built dual monitoring channels in sequence. The algorithm module inside the dual monitoring channels begins to fully analyze the data. One channel focuses on processing the long-term trend information reflected by the first sampling data set. By extracting features of long time series data, analyzing trends, and comparing with historical normal emission patterns, possible long-term abnormal change trends are identified; the other channel focuses on the short-term fluctuation details presented by multiple second sampling data sets, and uses advanced fluctuation analysis technology and anomaly detection algorithms to accurately judge frequent fluctuations in a short period of time, and capture abnormal fluctuation signals within a short period. Finally, the dual channels will comprehensively summarize and deeply integrate their respective analysis results. Based on the preset long-term abnormal excess judgment logic and threshold system, after rigorous reasoning and judgment, they will output the long-term abnormal excess judgment results of the first flue gas element in the flue gas emission data of the target thermal power plant. This result provides a key and reliable basis for subsequent environmental supervision decisions and enterprise rectification measures.

[0055] In a possible implementation, step S330 further includes:

[0056] Step S331: Determine a second flue gas element whose fluctuation index is less than or equal to the preset fluctuation index.

[0057] Step S332: acquiring the real-time emission monitoring data of the second flue gas element, and performing a real-time exceeding standard warning together with the calibrated emission permit information of the second flue gas element.

[0058] Specifically, a rigorous comparative analysis is conducted on the fluctuation index of each flue gas element and the preset fluctuation index. After precise calculation and careful judgment, the flue gas elements whose fluctuation index is less than or equal to the preset fluctuation index are selected and defined as the second flue gas elements. The emission characteristics of these second flue gas elements are relatively stable, and their fluctuation degree is within the preset safe range. Compared with elements with larger fluctuations, their long-term abnormal exceeding risk is relatively low and more difficult to accurately predict.

[0059] Focusing on the real-time over-standard warning of the second flue gas element, the real-time emission monitoring data of the second flue gas element is quickly obtained through special monitoring equipment and data acquisition systems. These data can immediately reflect the actual emission status of the element at the current moment, including key indicators such as emission concentration and emission rate. Subsequently, these real-time data are strictly compared with the pre-set calibrated emission permit information of the second flue gas element. The calibrated emission permit information is based on environmental protection regulations and the legal emission limit determined by comprehensive consideration of multiple factors such as local environmental carrying capacity. Once the real-time emission monitoring data approaches or exceeds the threshold range specified by the calibrated emission permit information, the early warning mechanism is immediately activated. The early warning method is that the sound and light alarm device on site sends a strong signal, and at the same time, it will send detailed over-standard warning information to the relevant environmental protection supervision department staff and the personnel responsible for environmental protection operation and maintenance within the enterprise through the automated information transmission system, informing which specific second flue gas element has exceeded the standard in real time and the degree of over-standard and other key information. In this way, potential excessive emission problems can be detected at the first time, prompting relevant personnel to take effective response measures in a timely manner, such as checking the operating status of production equipment, adjusting environmental protection treatment process parameters, etc., to ensure that the emission of the second flue gas element is always within the legal and compliant range, minimize the possibility of environmental pollution, and ensure the stability and safety of the surrounding ecological environment.

[0060] In a possible implementation, step S222 further includes:

[0061] Step S2221: Convert each element monitoring time sequence into each frequency domain signal through Fourier transform.

[0062] Step S2222: identify the frequency components of each frequency domain signal, obtain the average value of the fluctuation frequency corresponding to the part whose frequency component is greater than a preset threshold, and obtain the frequent fluctuation frequency of each element.

[0063] Specifically, the Fourier transform plays a key role in the obtained monitoring time series of each element. Based on mathematical principles, the Fourier transform can decompose and reconstruct the monitoring time series data of each element presented in the time domain. It regards the flue gas element emission data that changes continuously over time as a superposition of countless sine waves and cosine waves with different frequencies, amplitudes and phases. Through the calculation process of Fourier transform, these element monitoring time series are accurately converted from the time domain expression with time as the independent variable to various frequency domain signals with frequency as the independent variable. In the frequency domain signal, the energy or amplitude distribution corresponding to different frequency components can be clearly seen. For example, the low-frequency part may reflect the long-term trend change of flue gas emissions, while the high-frequency part reflects short-term rapid fluctuations or periodic oscillations. This conversion provides a new perspective and powerful tools for in-depth analysis of the fluctuation characteristics of each element monitoring time series, so that the fluctuation law that was originally complex and difficult to distinguish in the time domain can be presented in a more intuitive and structured way in the frequency domain, thus laying a solid foundation for the subsequent accurate identification of the frequent fluctuation frequency of each element.

[0064] In-depth frequency component identification is carried out on each frequency domain signal obtained by conversion. First, a preset threshold is set. This threshold is determined based on the study of the normal fluctuation frequency range of flue gas emissions and the statistical analysis of a large amount of historical data. Then, the frequency components in each frequency domain signal are analyzed one by one, and those frequency components with amplitudes greater than the preset threshold are screened out. These components represent the fluctuation frequency parts with high energy or significant influence in the monitoring time series. Finally, the mean value of the fluctuation frequency corresponding to these screened frequency components is calculated. This mean value can comprehensively reflect the fluctuation frequency of each element that frequently appears and has a greater influence during the monitoring process, thereby obtaining the frequent fluctuation frequency of each element. This result helps to further understand the regular changes in flue gas emissions and potential abnormal fluctuation patterns, and provides key frequency characteristic data for the entire flue gas emission monitoring and analysis system.

[0065] Embodiment 2, based on the same inventive concept as the flue gas emission monitoring method of a thermal power plant in the above embodiment, Figure 2 As shown, the present application provides a flue gas emission monitoring system for a thermal power plant, and the system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0066] The monitoring data time series generation module 10 is used to collect flue gas emission data from a target thermal power plant during a preset monitoring period and generate a flue gas emission monitoring data time series.

[0067] The fluctuation index generating module 20 performs fluctuation state analysis on each flue gas element in the flue gas based on the flue gas emission monitoring data time series to generate each fluctuation index.

[0068] The dual-channel monitoring configuration module 30 is configured to monitor the dual channels based on the respective fluctuation indicators.

[0069] The abnormal exceeding standard judgment result generation module 40 is used to extract the flue gas emission data of the target thermal power plant through the monitoring dual channels and perform long-term abnormal exceeding standard judgment to generate a long-term abnormal exceeding standard judgment result.

[0070] The smoke emission monitoring result generating module 50 is used to generate the smoke emission monitoring result based on the long-term abnormal exceeding standard judgment result.

[0071] Furthermore, the volatility index generating module 20 further includes:

[0072] An element monitoring time sequence generating unit is used to segment the flue gas emission monitoring data time sequence according to flue gas elements to generate each element monitoring time sequence.

[0073] A fluctuation severity identification unit is used to identify the fluctuation severity of each element monitoring time series and generate each fluctuation index.

[0074] Furthermore, the fluctuation severity identification unit further includes:

[0075] An element frequent fluctuation amplitude identification unit is configured to identify the frequent fluctuation amplitude of each element based on the monitoring time sequence of each element.

[0076] An element frequent fluctuation frequency identification unit is configured to identify the frequent fluctuation frequency of each element based on the monitoring time sequence of each element.

[0077] A standardization processing unit is used to perform standardization processing on the frequent fluctuation amplitudes of the various elements and the frequent fluctuation frequencies of the various elements and then perform weighting to generate the various fluctuation indicators.

[0078] Furthermore, the monitoring dual-channel configuration module 30 also includes:

[0079] A sampling frequency configuration unit, the sampling frequency configuration unit is used to configure a first sampling frequency and a second sampling frequency, wherein the second sampling frequency is greater than the first sampling frequency.

[0080] A monitoring dual-channel construction unit, which is based on a slowfast network, constructs a network structure of a monitoring dual-channel, and collects each first element emission sample data, multiple second element emission sample data and corresponding long-term abnormal exceeding risk identification data for each flue gas element according to the first sampling frequency and the second sampling frequency, and trains and constructs multiple initial monitoring dual channels.

[0081] A monitoring dual-channel generation unit is used to determine whether the configuration of each fluctuation index is greater than a preset fluctuation index, and to match the corresponding initial monitoring dual channel for the first flue gas element whose fluctuation index configuration is greater than the preset fluctuation index to generate the monitoring dual channel.

[0082] Furthermore, the abnormal exceeding standard judgment result generating module 40 further includes:

[0083] The first sampling data set generating unit is used to sample the first flue gas element in the flue gas emission data of the target thermal power plant according to a first sampling frequency to generate a first sampling data set.

[0084] The second sampling data set generating unit is used to sample the first flue gas element in the flue gas emission data of the target thermal power plant according to a second sampling frequency to generate a plurality of second sampling data sets.

[0085] The long-term abnormality exceeding standard judgment result output unit is used to respectively combine the first sampling data set and the second sampling data set and input them into the monitoring dual channel for analysis, and output the long-term abnormality exceeding standard judgment result.

[0086] Furthermore, the monitoring dual-channel generation unit also includes:

[0087] A fluctuation index determination unit is used to determine a second flue gas element whose fluctuation index is less than or equal to the preset fluctuation index.

[0088] A real-time exceeding standard warning unit is used to obtain the real-time emission monitoring data of the second flue gas element and to provide a real-time exceeding standard warning with the calibrated emission permit information of the second flue gas element.

[0089] Furthermore, the element frequent fluctuation frequency identification unit further includes:

[0090] A frequency domain signal conversion unit is used to convert the individual element monitoring time series into individual frequency domain signals through Fourier transform.

[0091] A fluctuation frequency mean value acquisition unit is used to identify the frequency components of each frequency domain signal, obtain the fluctuation frequency mean value corresponding to the part whose frequency component is greater than a preset threshold, and obtain the frequent fluctuation frequency of each element.

[0092] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0093] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0094] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for monitoring flue gas emissions from a thermal power plant, characterized in that: include: Collect flue gas emission data from the target thermal power plant during the preset monitoring period and generate a time series of flue gas emission monitoring data; Performing fluctuation state analysis on each flue gas element in the flue gas based on the flue gas emission monitoring data time series to generate each fluctuation index; Based on the above mentioned volatility indicators, dual monitoring channels are configured; Extracting the flue gas emission data of the target thermal power plant through the dual monitoring channels and conducting long-term abnormal exceeding standard judgment, thereby generating a long-term abnormal exceeding standard judgment result; The flue gas emission monitoring result is generated based on the long-term abnormal exceeding standard judgment result.

2. A method for monitoring flue gas emissions from a thermal power plant as claimed in claim 1, characterized in that: Based on the flue gas emission monitoring data time series, the fluctuation state of each flue gas element in the flue gas is analyzed to generate various fluctuation indicators, including: Segmenting the flue gas emission monitoring data time series according to flue gas elements to generate monitoring time series for each element; The severity of fluctuations in the monitoring time series of each element is identified to generate the fluctuation indicators.

3. A method for monitoring flue gas emissions in a thermal power plant as claimed in claim 2, characterized in that: The intensity of fluctuations of the monitoring time series of each element is identified to generate each fluctuation index, including: Identify the frequent fluctuation amplitude of each element based on the monitoring time series of each element; Identify the frequent fluctuation frequency of each element based on the monitoring time sequence of each element; The frequent fluctuation amplitudes of the various elements and the frequent fluctuation frequencies of the various elements are normalized and weighted to generate the various fluctuation indices.

4. A method for monitoring flue gas emissions from a thermal power plant as claimed in claim 1, characterized in that: Based on the above mentioned volatility indicators, dual channels are configured for monitoring, including: Configure a first sampling frequency and a second sampling frequency, wherein the second sampling frequency is greater than the first sampling frequency; Based on the slowfast network, a network structure for monitoring dual channels is constructed, and each first element emission sample data, a plurality of second element emission sample data and corresponding long-term abnormal exceeding risk identification data are collected for each flue gas element according to the first sampling frequency and the second sampling frequency, and a plurality of initial monitoring dual channels are trained and constructed; It is determined whether each fluctuation index configuration is greater than a preset fluctuation index, and a corresponding initial monitoring dual channel is matched for a first flue gas element whose fluctuation index configuration is greater than the preset fluctuation index to generate the monitoring dual channel.

5. A method for monitoring flue gas emissions in a thermal power plant as claimed in claim 4, characterized in that: The flue gas emission data of the target thermal power plant is extracted through the dual monitoring channels and a long-term abnormal exceeding standard judgment is performed to generate a long-term abnormal exceeding standard judgment result, including: Sampling a first flue gas element in the flue gas emission data of the target thermal power plant according to a first sampling frequency to generate a first sampling data set; Sampling the first flue gas element in the flue gas emission data of the target thermal power plant according to a second sampling frequency to generate a plurality of second sampling data sets; The first sampling data set and the second sampling data set are respectively combined and input into the monitoring dual channels for analysis, and the long-term abnormal exceeding judgment result is output.

6. A method for monitoring flue gas emissions in a thermal power plant as claimed in claim 4, characterized in that: Determining whether each volatility index configuration is greater than a preset volatility index further includes: Determining a second flue gas element whose fluctuation index is less than or equal to the preset fluctuation index; The real-time emission monitoring data of the second flue gas element is obtained, and a real-time exceeding standard warning is issued together with the calibrated emission permit information of the second flue gas element.

7. A method for monitoring flue gas emissions from a thermal power plant as claimed in claim 3, characterized in that: Identifying the frequent fluctuation frequency of each element based on the monitoring time sequence of each element includes: Convert the monitoring time series of each element into frequency domain signals through Fourier transform; Frequency components of each frequency domain signal are identified, and the average value of the fluctuation frequency corresponding to the part whose frequency component is greater than a preset threshold is obtained to obtain the frequent fluctuation frequency of each element.

8. A flue gas emission monitoring system for a thermal power plant, characterized in that: include: A flue gas emission monitoring data time series generation module, the flue gas emission monitoring data time series generation module is used to collect flue gas emission data from a target thermal power plant during a preset monitoring period and generate a flue gas emission monitoring data time series; A fluctuation index generation module, wherein the fluctuation index generation module performs fluctuation state analysis on each flue gas element in the flue gas based on the flue gas emission monitoring data time series to generate each fluctuation index; A monitoring dual-channel configuration module, wherein the monitoring dual-channel configuration module configures the monitoring dual-channel based on the respective fluctuation indicators; An abnormal exceeding standard judgment result generation module, the abnormal exceeding standard judgment result generation module is used to extract the flue gas emission data of the target thermal power plant through the monitoring dual channels and perform long-term abnormal exceeding standard judgment to generate a long-term abnormal exceeding standard judgment result; A flue gas emission monitoring result generation module is used to generate a flue gas emission monitoring result based on the long-term abnormal exceeding standard judgment result.

Images