An air preheater state parameter anomaly judgment method based on a dynamic benchmark library

Abstract

The application discloses a kind of based on dynamic benchmark library's air preheater state parameter exception judging method, using the key parameters including unit load, coal low calorific value, ambient temperature and boiler flue gas quantity that influence air preheater operation as boundary condition, to distinguish different air preheater operating conditions, monitor air preheater operating current under each operating condition, air preheater flue gas resistance, heat exchange efficiency and flue gas side air leakage rate, establish air preheater state parameter benchmark library, and the judging method of data deviation anomaly, fluctuation anomaly and deterioration anomaly is elaborated in detail.The application will real-time operation data into exception judging model and is calculated and analyzed, and operating personnel can find air preheater state parameter exception and handle in time, so as to avoid equipment problem or failure influence expansion, save production cost, improve the market competitiveness of power generation enterprise.

Description

Technical Field

[0001] This invention relates to air preheater technology, and more particularly to a method for judging abnormal air preheater state parameters based on a dynamic benchmark library. Background Technology

[0002] Air preheaters, or air generators for short, are crucial components of thermal power generating units. Abnormalities or accidents can frequently occur due to aging, deterioration, inadequate maintenance, and operational errors. Existing DCS alarm systems have broad setting values ​​and simple anomaly detection methods. Due to workload limitations, operators have limited time to monitor parameter trends, making it difficult to effectively detect long-term equipment deterioration trends. Often, by the time parameters exceed warning limits and thermal protection systems activate, significant equipment deterioration and malfunctions have already occurred, missing the optimal intervention window and potentially escalating the accident. Equipment deterioration or anomalies develop gradually; the earlier anomalies are detected and addressed, the smaller the consequences and impact. Therefore, timely and accurate anomaly detection is paramount.

[0003] By acquiring real-time and historical parameters of the air preheater within the DCS system, and deeply mining and analyzing massive amounts of operational data, alarm values ​​can be reasonably set through logical analysis and mathematical calculations. The optimal expectation of parameters under specific operating conditions can be given. When the parameters deviate from the optimal range and do not reach the upper and lower limit design values, a judgment can be made. Abnormal information can be issued in the early stage of equipment abnormality before serious consequences occur, which is of great significance to safe production. Summary of the Invention

[0004] This invention addresses the shortcomings of existing technologies by proposing a method for judging anomalies in air preheater state parameters based on a dynamic benchmark library. The aim is to acquire real-time and historical operating data of the air preheater, establish and update a benchmark library of air preheater state parameters under different operating conditions, and use this dynamic benchmark library to judge anomalies in state parameters such as air preheater operating current, flue gas resistance, heat exchange efficiency, and flue gas leakage rate, thereby monitoring and detecting deviations, fluctuations, and deterioration anomalies during operation.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for judging the anomalies of air preheater state parameters based on a dynamic benchmark library includes the following steps:

[0007] Step S1: Use key parameters affecting air preheater operation, including unit load, lower heating value of coal, ambient temperature and boiler flue gas volume, as boundary conditions to distinguish different air preheater operating conditions. Collect state parameters under each operating condition, including air preheater operating current, air preheater flue gas resistance, heat exchange efficiency and flue gas side leakage rate, and establish an air preheater state parameter benchmark library. One state parameter under one operating condition corresponds to one benchmark library cell.

[0008] Step S2: Perform data cleaning on the collected real-time data, i.e., determine whether there are missing values, sudden increases or decreases, or / and exceed the upper and lower limits. If the data cleaning is normal, proceed to the step of judging abnormal fluctuations in the air preheater status parameters.

[0009] If the fluctuation is normal, proceed to step S2-1; if the data cleaning is abnormal or the fluctuation is abnormal, repeat step S2.

[0010] Step S2-1: Analyze and process real-time operating data, find the corresponding benchmark cell based on boundary conditions, and determine whether the current air preheater meets the steady-state operating conditions; if it meets the steady-state operating conditions, proceed to step S2-2; otherwise, proceed to step S2-4.

[0011] Step S2-2: Determine whether the amount of data in the cell is greater than 10% of the upper limit of the data amount in the cell. If so, proceed to step S2-3. Otherwise, perform a judgment on the abnormal deviation of the air preheater state parameters under small data volume and steady-state conditions, and then proceed to step S3.

[0012] Step S2-3: Based on the grid content, perform abnormal judgment on the deviation of the air preheater state parameters under typical steady-state conditions. Determine whether the amount of data in the grid is greater than 30% of the set upper limit value of the amount of data in the grid. If so, perform a deterioration anomaly judgment and proceed to step S3; otherwise, proceed directly to step S3.

[0013] Step S2-4: Determine whether the amount of data in the cell is greater than 10% of the upper limit of the data amount in the cell. If so, proceed to step S2-5. Otherwise, perform a judgment on the abnormal deviation of the air preheater state parameters under small data volume and unsteady operating conditions, and then go to step S2.

[0014] Step S2-5: Based on the grid content, determine the abnormal deviation of the air preheater state parameters under typical unsteady conditions, and then proceed to step S2;

[0015] Step S3: If there are no deviation or deterioration abnormalities in the real-time status parameters of the air preheater, and the data collection interval is not less than the set grid update interval, then update the current value to the benchmark grid; otherwise, do not update. If the amount of data in the grid has reached the set upper limit, then delete the oldest historical data while updating; otherwise, add it directly. Finally, return to step S2.

[0016] In step S2, the method for judging abnormal fluctuations is as follows: take the historical data of each state parameter for the past 5 minutes, divide it into 5 time intervals, and count the number of times n, the deviation of the maximum and minimum values ​​in each interval exceeds the set fluctuation threshold. If n < 2, the state parameter is normal; otherwise, the fluctuation is considered abnormal.

[0017] In step S2-1, the method for judging steady-state operating conditions is as follows: after selecting a series of key parameters characterizing the unit operation, including the lower heating value of coal, unit load, flue gas oxygen content and furnace negative pressure, the average value of each parameter within a set time range is calculated. If the difference between the real-time parameter and each average value is within the fluctuation error range, the operating condition is considered to be steady state.

[0018] The aforementioned method for judging abnormal air preheater state parameters based on a dynamic benchmark library, wherein the method for judging abnormal state parameter deviations is as follows: Air preheater state parameter X under each operating condition... i The numbers i = 1, 2, 3, ... follow a Gaussian distribution X. i : Where μ i Let σ be the mathematical expectation. i 2 For variance;

[0019] a. In steps S2-3, the method for judging abnormal deviations under typical steady-state conditions is as follows: if each state parameter is within the interval (μ i -3σ i ,μ i +3σ i If i = 1, 2, 3, ..., then the state parameters are normal; otherwise, the deviation is abnormal.

[0020] b. In step S2-5, the method for judging abnormal deviations under typical unsteady-state conditions is as follows: if each state parameter is within the interval (μ i -6σ i ,μ i +6σ i If i = 1, 2, 3, ..., then the state parameters are normal; otherwise, the deviation is abnormal.

[0021] c. In step S2-2, the method for judging abnormal deviations of the air preheater state parameters under small data volume and steady-state conditions is as follows: if each state parameter is within the interval (μ i -5% μ i ,μ i +5%μ i If i = 1, 2, 3, ..., then the state parameters are normal; otherwise, the deviation is abnormal.

[0022] d. In steps S2-4, the method for judging abnormal deviations of the air preheater state parameters under small data volumes and unsteady-state conditions is as follows: if each state parameter is within the interval (μ i -10% μ i ,μ i +10%μ i If i = 1, 2, 3, ..., then the state parameters are normal; otherwise, the deviation is abnormal.

[0023] In step S2-3, the method for judging abnormal deterioration is as follows: read the contents of the benchmark library grid corresponding to the state parameters of each air preheater under a certain working condition, take the latest data of 5% of the upper limit value of the data volume in the grid and the early data of 5%, calculate the average value respectively, and then calculate the difference to obtain the deterioration deviation. If the deterioration deviation of the air preheater state parameter is less than the set deterioration threshold, the state parameter is normal; otherwise, it is considered abnormal deterioration.

[0024] In step S3, the benchmark cell update interval should not exceed 1 minute, and the maximum amount of data in a cell should not be less than 150,000.

[0025] Beneficial effects of this invention:

[0026] The method of this invention uses key parameters affecting the operation of the air preheater as boundary conditions, mines real-time and historical data, establishes a benchmark library of air preheater state parameters, and elaborates in detail the judgment methods for data deviation anomalies, fluctuation anomalies and deterioration anomalies.

[0027] This invention incorporates real-time operational data into an anomaly detection model for calculation and analysis. Operators can promptly detect and address anomalies in the air preheater's status parameters, thereby preventing equipment problems or malfunctions from escalating. It can also play a greater role in saving production costs and enhancing the market competitiveness of power generation companies. Attached Figure Description

[0028] Figure 1 The flowchart of the air preheater state parameter anomaly judgment method based on dynamic benchmark library of the present invention.

[0029] Figure 2 An example of the method for judging abnormal air preheater state parameters based on a dynamic benchmark library according to the present invention. Detailed Implementation

[0030] To make the technical concept of the present invention clearer, the technical solution of the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the following embodiments are only used to explain and illustrate the preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0031] Example 1: Anomaly Judgment Method for Air Preheater State Parameters Based on Dynamic Benchmark Library

[0032] Taking the air preheater operation status monitoring of a 600MW coal-fired unit as an example, see... Figure 1 and Figure 2 The judgment method includes the following steps:

[0033] Step 1: Use key parameters affecting air preheater operation, including unit load L and lower heating value of coal Q. net Ambient temperature T at and boiler flue gas volume q gAs boundary conditions, to distinguish different air preheater operating conditions, state parameters under each operating condition are collected, including the air preheater operating current I. AH ΔP, flue gas resistance of air preheater AH,g Air preheater heat exchange efficiency η and flue gas side leakage rate

[0034] Data mining was performed on historical data from June 1, 2021 to May 31, 2022 to establish a benchmark library of air preheater state parameters, as shown in Tables 1 and 2 below.

[0035] Table 1 Boundary conditions and requirements affecting air preheater operation

[0036]

[0037] Table 2. Status parameters affecting air preheater operation

[0038]

[0039] Database management systems such as MySQL, PostgreSQL, and SQL Server can be selected to build a benchmark library of air preheater status parameters for data storage and retrieval;

[0040] Step 2: Clean the collected real-time data. If the data has one or more of the following conditions: missing values, sudden increases or decreases, or exceeding upper or lower limits, the data cleaning is considered abnormal, and Step 2 is repeated; otherwise, proceed to Step 3.

[0041] Step 3: Analyze and process air preheater operating status parameters such as air preheater operating current, air preheater flue gas resistance, air preheater heat exchange efficiency, and flue gas side leakage rate to determine if there are any abnormal fluctuations, under the following conditions:

[0042] Historical data for each state parameter over the past 5 minutes were collected and divided into 5 time intervals. The number of times (n) the maximum and minimum deviations in each interval exceeded the set fluctuation thresholds was counted. If n < 2, the state parameter was considered normal; otherwise, the fluctuation was considered abnormal. Specifically, the set fluctuation thresholds for the air preheater operating current were 0.6 A, the air preheater flue gas resistance was 0.5 kPa, the air preheater heat exchange efficiency was 5%, and the flue gas side leakage rate was 3%.

[0043] Taking the judgment of abnormal air preheater current fluctuation as an example, the data of the past 5 minutes is taken and divided into 5 time periods: T1 = (12.4, 12.3, 12.1, ..., 12.5), T2 = (12.5, 12.3, 12.3, ..., 12.6), T3 = (12.6, 12.7, 12.5, ..., 12.5), T4 = (12.5, 12.3, 12.3, ..., 12.6), T5 = (12.7, 12.7, 12.9, ..., 12.5). The air preheater current fluctuation threshold is set to 0.6A. The maximum and minimum deviations within T1 to T5 are 0.5, 0.3, 0.3, 0.4, and 0.3, respectively. Therefore, n = 0, and it is considered that there is no abnormal fluctuation in the air preheater current at this time.

[0044] If there are no abnormal fluctuations in the state parameters, proceed to step 4; otherwise, go to step 2.

[0045] Step 4: Select a series of key parameters characterizing the unit operation, including the lower heating value of coal, unit load, flue gas oxygen content and furnace negative pressure, and calculate the average value of each parameter within a set time range. If the difference between the real-time parameter and each average value is within the fluctuation error range, the operating condition is considered to be steady state.

[0046] The steady-state operating conditions are determined as follows:

[0047] a. Read the data of the lower heating value of coal over the past 10 minutes, calculate the average value, and the difference between the current value and the average value is within ±1000 kJ / kg, which is normal; otherwise, it is abnormal.

[0048] b. Read the unit load data for the past 10 minutes, calculate the average value, and the difference between the current value and the average value is within ±6MW if it is normal; otherwise, it is abnormal.

[0049] c. Read the oxygen content data of flue gas for the past 10 minutes, calculate the average value, and the difference between the current value and the average value is within ±0.5% if it is normal; otherwise, it is abnormal.

[0050] d. Calculate the average value of the current furnace negative pressure data over the past 3 seconds. The difference between the current value and the average value is considered normal if it is within ±100 Pa; otherwise, it is considered abnormal.

[0051] Table 3 below shows an example of a steady-state judgment. All parameters are within the set range, and the current air preheater operating condition is considered to be in a steady state.

[0052] Table 3 Real-time data of air preheater operation

[0053] Low calorific value of coal kJ / kg 18050 18322 -272 normal Unit load MW 465 463 2 normal Oxygen content in flue gas ％ 3.1 3.3 -0.2 normal Furnace negative pressure Pa -52 -30 -22 normal

[0054] If the steady-state condition is not met, proceed to step 4.1; if it is met, proceed to step 4.2.

[0055] Step 4.1: Determine if the data volume within a cell exceeds 10% of the set upper limit for data volume within the cell. If so, perform an anomaly check for air preheater state parameter deviations under typical unsteady-state conditions; otherwise, perform an anomaly check for air preheater state parameter deviations under small data volumes and unsteady-state conditions. The method is as follows:

[0056] Air preheater state parameters X under various operating conditions i The numbers i = 1, 2, 3, ... follow a Gaussian distribution X. i : Where μ i Let σ be the mathematical expectation. i 2 For variance;

[0057] a. The method for judging abnormal deviations under typical unsteady-state conditions is as follows: if each state parameter is within the interval (μ i -6σ i ,μ i +6σ i If i = 1, 2, 3, ..., then the state parameters are normal; otherwise, they are abnormal.

[0058] b. The method for judging abnormal deviations of air preheater state parameters under small data volume and unsteady operating conditions is as follows: If each state parameter is within the interval (μ i -10% μ i ,μ i +10%μ i If i = 1, 2, 3, ..., then the state parameters are normal; otherwise, they are abnormal.

[0059] Then, proceed to step 5.

[0060] Step 4.2: Determine whether the amount of data in the cell is greater than 10% of the upper limit of the data amount in the cell. If so, perform an anomaly judgment on the deviation of the air preheater state parameters under typical steady-state conditions, and then proceed to step 4.3. Otherwise, perform an anomaly judgment on the deviation of the air preheater state parameters under small data volume and steady-state conditions, and then proceed to step 5.

[0061] a. The method for judging abnormal deviations under typical steady-state conditions is as follows: if each state parameter is within the interval (μ i -3σ i ,μ i +3σ i If i = 1, 2, 3, ..., then the state parameters are normal; otherwise, they are abnormal.

[0062] b. The method for judging abnormal deviations of air preheater state parameters under small data volume and steady-state operating conditions is as follows: If each state parameter is within the interval (μ i -5% μ i ,μ i +5%μ iIf i = 1, 2, 3, ..., then the state parameters are normal; otherwise, they are abnormal.

[0063] like Figure 2 For example, the air preheater is judged to be in steady state, and the data volume in the benchmark cell corresponding to the real-time data boundary conditions is 51352, which is 10% greater than the set upper limit of 150000 for the data volume in the cell, and the air preheater current follows a Gaussian distribution. N(12.37, 0.0667) is used to determine the abnormal deviation under typical steady-state operating conditions. The real-time value of the air preheater current is 12.40A, which is within the range [11.59, 13.14], therefore there is no abnormal deviation in the air preheater current.

[0064] Step 4.3: Determine whether the amount of data in the cell is greater than 30% of the upper limit of the data amount in the cell. If so, perform a degradation anomaly judgment and then proceed to step 5; otherwise, proceed directly to step 5.

[0065] The method for judging abnormal deterioration is as follows: Read the contents of the benchmark library grid corresponding to the state parameters of each air preheater under a certain operating condition, take the latest data of 5% of the upper limit of the data volume in the grid and 5% of the early data, calculate the average value and compare them. If the deterioration deviation of the air preheater state parameters is less than the set deterioration threshold, the state parameters are normal; otherwise, it is considered abnormal deterioration.

[0066] Among them, the deterioration threshold for the air preheater operating current is set at 1.2A, the deterioration threshold for the air preheater flue gas resistance is set at 1kPa, the deterioration threshold for the air preheater heat exchange efficiency is set at 5%, and the deterioration threshold for the flue gas side leakage rate is set at 3%.

[0067] like Figure 2 For example, the real-time data boundary condition corresponds to a benchmark cell with 51,352 data points, which is 30% greater than the set upper limit of 150,000 data points in the cell. Taking the air preheater flue gas side leakage rate as an example, the average value is calculated by taking the latest 7,500 data points (including the current real-time value) and the previous 7,500 data points in the cell, respectively. The results are 6.89% and 6.03%, respectively. The difference between the two is 0.86%, which is less than the set deterioration threshold of 3%. Therefore, the current air preheater flue gas side leakage rate is considered normal, and there is no abnormal deterioration.

[0068] Step 5: If the real-time status parameters of the air preheater conform to steady-state conditions, and there are no abnormal parameter deviations or degradations, and the data acquisition interval is not less than the set grid update interval, then update the current value to the benchmark database grid; otherwise, do not update. If the data volume in the grid has reached the set upper limit, then delete the oldest historical data while updating; otherwise, add it directly. Finally, return to Step 2.

[0069] The above are merely preferred embodiments of the present invention. It should be noted that for those skilled in the art, various improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also within the protection scope of the present invention.

Claims

1. A method for judging abnormal state parameters of an air preheater based on a dynamic benchmark library, characterized in that: Includes the following steps: Step S1: Use key parameters affecting air preheater operation as boundary conditions to distinguish different air preheater operating conditions. The key parameters affecting air preheater operation include unit load, lower heating value of coal, ambient temperature and boiler flue gas volume. Collect state parameters under each operating condition and establish an air preheater state parameter benchmark library. The state parameters under each operating condition include air preheater operating current, air preheater flue gas resistance, heat exchange efficiency and flue gas side leakage rate. One state parameter under one operating condition corresponds to one benchmark library cell. Step S2: Perform data cleaning on the collected real-time data, i.e., determine whether there are missing values, sudden increases or decreases, or / and exceed the upper and lower limits. If the data cleaning is normal, proceed to the step of judging abnormal fluctuations in the air preheater status parameters. If the fluctuation is normal, proceed to step S2-1; if the data cleaning is abnormal or the fluctuation is abnormal, repeat step S2. Step S2-1: Analyze and process real-time operating data, find the corresponding benchmark cell based on boundary conditions, and determine whether the current air preheater meets the steady-state operating conditions; if it meets the steady-state operating conditions, proceed to step S2-2; otherwise, proceed to step S2-4. Step S2-2: Determine whether the amount of data in the cell is greater than 10% of the upper limit of the data amount in the cell. If so, proceed to step S2-3. Otherwise, perform a judgment on the abnormal deviation of the air preheater state parameters under small data volume and steady-state conditions, and then proceed to step S3. Step S2-3: Based on the grid content, perform abnormal judgment on the deviation of the air preheater state parameters under typical steady-state conditions. Determine whether the amount of data in the grid is greater than 30% of the set upper limit value of the amount of data in the grid. If so, perform a deterioration anomaly judgment and proceed to step S3; otherwise, proceed directly to step S3. Step S2-4: Determine whether the amount of data in the cell is greater than 10% of the upper limit of the data amount in the cell. If so, proceed to step S2-5. Otherwise, perform a judgment on the abnormal deviation of the air preheater state parameters under small data volume and unsteady operating conditions, and then go to step S2. Step S2-5: Based on the grid content, determine the abnormal deviation of the air preheater state parameters under typical unsteady conditions, and then proceed to step S2; The method for judging abnormal deviations in the state parameters is as follows: Air preheater state parameters under various operating conditions. X i , i The values ​​1, 2, 3, ... follow a Gaussian distribution. ,in μ i For mathematical expectation, σ i 2 For variance; a. In steps S2-3, the method for judging abnormal deviations under typical steady-state conditions is as follows: if each state parameter is within the range If the condition is within the acceptable range, the status parameters are normal; otherwise, the deviation is abnormal. b. In step S2-5, the method for judging abnormal deviations under typical unsteady-state conditions is as follows: if each state parameter is within the range If the condition is within the acceptable range, the status parameters are normal; otherwise, the deviation is abnormal. c. In step S2-2, the method for judging abnormal deviations of the air preheater state parameters under small data volume and steady-state conditions is as follows: if each state parameter is within the range If the condition is within the acceptable range, the status parameters are normal; otherwise, the deviation is abnormal. d. In steps S2-4, the method for judging abnormal deviations of air preheater state parameters under small data volumes and unsteady-state conditions is as follows: if each state parameter is within the range If the condition is within the acceptable range, the status parameters are normal; otherwise, the deviation is abnormal. Step S3: If there are no deviation or deterioration abnormalities in the real-time status parameters of the air preheater, and the data collection interval is not less than the set grid update interval, then update the current value to the benchmark grid; otherwise, do not update. If the amount of data in the grid has reached the set upper limit, then delete the oldest historical data while updating; otherwise, add it directly. Finally, return to step S2.

2. The method for judging abnormal air preheater state parameters based on a dynamic benchmark library according to claim 1, characterized in that: In step S2, the method for judging abnormal fluctuations is as follows: take the historical data of each state parameter for the past 5 minutes, divide it into 5 time intervals, and count the number of times n, the deviation of the maximum and minimum values ​​in each interval exceeds the set fluctuation threshold. If n < 2, the state parameter is normal; otherwise, the fluctuation is considered abnormal.

3. The method for judging abnormal air preheater state parameters based on a dynamic benchmark library according to claim 1, characterized in that: In step S2-1, the method for determining steady-state operating conditions is as follows: after selecting a series of key parameters characterizing the unit's operation, the average value of each parameter is calculated within a set time range. The series of key parameters characterizing the unit's operation include the lower heating value of coal, unit load, flue gas oxygen content, and furnace negative pressure. If the difference between the real-time parameters and each average value is within the fluctuation error range, the operating condition is considered to be steady state.

4. The method for judging abnormal air preheater state parameters based on a dynamic benchmark library according to claim 1, characterized in that: In step S2-3, the method for judging abnormal deterioration is as follows: read the contents of the benchmark library grid corresponding to the state parameters of each air preheater under a certain working condition, take the latest data of 5% of the upper limit value of the data volume in the grid and the early data of 5%, calculate the average value respectively, and then calculate the difference to obtain the deterioration deviation. If the deterioration deviation of the air preheater state parameter is less than the set deterioration threshold, the state parameter is normal; otherwise, it is considered abnormal deterioration.

5. The method for judging abnormal air preheater state parameters based on a dynamic benchmark library according to claim 1, characterized in that: In step S3, the benchmark cell update interval should not exceed 1 minute, and the maximum amount of data in a cell should not be less than 150,000.

Images