A Smart Control Method for Grate Coolers Based on Cement Kiln Operating Condition Identification

Abstract

This invention discloses an intelligent control method for grate coolers based on cement kiln operating condition identification. It improves existing grate cooler control schemes by utilizing kiln operating condition identification technology to collect kiln operating condition parameter data, achieving mutual matching between grate cooler operating parameters and cement kiln operating conditions. This enables automatic optimization of the grate pressure setpoint under typical operating conditions, improving the heat recovery efficiency of the cement kiln. It does not rely on operator skill or experience, avoiding the problems of frequent manual modifications to determine the grate pressure setpoint, which require lengthy adjustments and adjustments, and where operator skill and experience vary, making it impossible to guarantee that the grate cooler's operating control parameters are at their optimal values ​​under the current operating conditions.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for grate coolers in the cement industry, and in particular to an intelligent control method for grate coolers based on cement kiln operating condition identification. Background Technology

[0002] The grate cooler is the core equipment in a cement kiln head, its main function being to rapidly cool high-temperature clinker and recover its heat. Maintaining the grate cooler in good working order is crucial for improving clinker heat recovery efficiency and stabilizing kiln system conditions. It also helps prevent the formation of secondary free calcium, thus improving clinker quality. For cement plants, achieving intelligent control of the grate cooler, improving the heat recovery efficiency of the grate cooler's air-heat exchange, and maximizing the secondary air temperature can significantly reduce energy consumption and carbon emissions during cement firing. However, the grate cooler system is a strongly coupled, high-lag, and high-inertia system, and kiln conditions change frequently, making it difficult to establish an accurate mathematical model. Currently, the control of grate coolers usually relies on the operator's experience, requiring frequent modification of the grate pressure setpoint or direct adjustment of the grate speed. On the one hand, frequent modification of the grate pressure setpoint requires a long period of adjustment and experimentation, which can lead to poor clinker cooling, reduced quality, and wasted energy. On the other hand, the operators' techniques and experience vary, making it impossible to guarantee that the operating control parameters of the grate cooler are at their optimal values ​​under the current working conditions. Summary of the Invention

[0003] This invention aims to overcome the problems of existing technologies where frequent manual modifications to determine the grate pressure setting require long-term adjustments and exploration, and where differences in operator skills and experience make it impossible to guarantee that the operating control parameters of the grate cooler are at their optimal values ​​under the current working conditions. The invention provides an intelligent control method for grate coolers based on cement kiln operating condition identification.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A method for intelligent control of a grate cooler based on cement kiln operating condition identification includes the following steps: S1: Acquire kiln current signal data and preprocess the kiln current signal data; S2: Obtain a kiln current K-line graph using the kiln current K-line analysis method; S3: Input the kiln current K-line graph into an expert knowledge base, and derive the kiln operating condition strength index K11 and the kiln operating condition stability index K21 through an inference engine; S4: Collect the first real-time parameter data of the kiln operating condition, and derive the kiln operating condition strength correction coefficient K12 and the kiln operating condition stability correction coefficient K22 based on the first real-time parameter data of the kiln operating condition; S5: Based on the correction... The coefficients are used to correct the kiln operating condition intensity index K11 and the kiln operating condition stability index K21 to obtain the comprehensive kiln operating condition intensity index K13 and the comprehensive kiln operating condition stability index K23; S6: Collect the second real-time parameter data of the kiln operating condition, and obtain the grate pressure setpoint, grate speed control cycle and grate speed control parameters through the optimization algorithm library based on the second real-time parameter data of the kiln operating condition, the comprehensive kiln operating condition intensity index K13 and the comprehensive kiln operating condition stability index K23; S7: Use the grate pressure setpoint, grate speed control cycle and grate speed control parameters as inputs to the grate cooler control module to realize intelligent control of the grate cooler. This invention provides an intelligent control method for grate coolers based on cement kiln operating condition identification. It improves upon existing grate cooler control schemes by utilizing kiln operating condition identification technology to collect kiln operating condition parameter data. This enables the matching of grate cooler operating parameters with cement kiln operating conditions, achieving automatic optimization of the grate pressure setpoint under typical operating conditions. This improves the heat recovery efficiency of the cement kiln and avoids the problem of frequent manual modifications to the grate pressure setpoint, which require lengthy adjustments and adjustments, and where operator skills and experience vary, making it impossible to guarantee that the grate cooler's operating control parameters are at their optimal values ​​under the current operating conditions.

[0006] As a preferred embodiment of the present invention, S2 specifically involves: within the kiln current interception period, connecting the median kiln current M1 of the first half of the kiln current period with the median kiln current M2 of the second half of the kiln current period to form a rectangular prism, defining the area enclosed by this prism as the kiln current entity; intercepting the highest kiln current value H1 and the lowest kiln current value L1 within the period, connecting the highest and lowest kiln current values ​​with the kiln current entity using thin lines to obtain a kiln current K-line graph; the line connecting the highest kiln current value H1 and the kiln current entity is the upper shadow line of the kiln current, and the line connecting the lowest kiln current value L1 and the kiln current entity is the lower shadow line of the kiln current; if the kiln current median M1 is greater than the kiln current median M2, then the kiln current K-line is defined as a negative line; if the kiln current median M1 is less than the kiln current median M2, then the kiln current K-line is defined as a positive line. This invention proposes a kiln current K-line identification method. The kiln current K-line diagram serves as a link between the dynamic and static processes of the kiln system, efficiently extracting key information about the kiln current, effectively reducing data noise, and facilitating the summarization of experience by kiln operators and process experts.

[0007] As a preferred embodiment of the present invention, step S3 specifically involves: inputting the kiln current K-line graph into the expert knowledge base; the inference engine starting from the first rule in the expert knowledge base and sequentially scanning all rules in the expert knowledge base, matching the rule premises with the current facts in the comprehensive database; when a conflict occurs in the matching rules, a depth-first strategy is adopted, and after a high-weight rule is triggered, the low-weight rule is ignored in this search; based on the kiln current K-line graph and the rules, the intensity and stability of the current kiln system operating conditions are inferred, resulting in the kiln operating condition intensity index K11 and the kiln operating condition stability index K21. The kiln operating condition intensity index K11 and the kiln operating condition stability index K21 can basically reflect changes in the kiln conditions, but to fully consider the impact of changes in other parameters on the kiln conditions, further corrections to the kiln operating condition intensity index K11 and the kiln operating condition stability index K21 are needed.

[0008] As a preferred embodiment of the present invention, the first real-time parameter data in S4 includes the kiln system feed rate, flue gas NOx, secondary air temperature, tertiary air temperature, decomposition furnace outlet temperature, free calcium hourly value, raw material silicon ratio, raw material aluminum ratio, raw material lime saturation coefficient, decomposition furnace coal feed rate, and kiln head coal feed rate.

[0009] In a preferred embodiment of the present invention, step S4 specifically involves: decreasing the weight of the first real-time parameter according to its timeliness; weighting the intensity index and stability index of each kiln condition real-time parameter to obtain the kiln condition intensity correction coefficient K12 and the kiln condition stability correction coefficient K22. Using the kiln system parameters under typical operating conditions as a benchmark, the intensity index and stability index of each parameter under the current operating condition are given by comparing the real-time kiln system parameters with historical parameters under typical kiln conditions. Based on the timeliness of each parameter and its impact on the kiln condition, the weight of the parameter on the kiln condition is dynamically adjusted.

[0010] As a preferred embodiment of the present invention, S5 specifically involves: dividing the kiln condition stability index K21 by the kiln condition stability correction coefficient K22 to obtain the comprehensive kiln condition stability index K23, and multiplying the kiln condition strength correction coefficient K12 by the kiln condition strength index K11 to obtain the comprehensive kiln condition strength index K13.

[0011] As a preferred embodiment of the present invention, the second real-time parameter data includes grate pressure, oil pump pressure, secondary air temperature, air temperature, clinker temperature, blower current, and kiln system feed rate.

[0012] As a preferred embodiment of the present invention, S6 specifically involves: collecting second real-time parameter data of kiln operation; correlating the kiln system operation, kiln system stability, grate cooler operation, and other operating condition judgment conditions with actual operating data based on the second real-time parameter data of kiln operation, the comprehensive kiln operation intensity index K13, and the comprehensive kiln operation stability index K23; inferring the trend of related parameters of the grate cooler system; and recommending grate speed control parameters, grate speed control cycle, and grate pressure setpoints based on historical data, parameter ranges, and trend changes using an optimization algorithm library. After obtaining the comprehensive kiln operation intensity index K13 and the comprehensive kiln operation stability index K23, parameters such as kiln system feed rate, grate pressure, oil pump pressure, secondary air temperature, air temperature, clinker temperature, and fan current are collected. Using the optimization algorithm library, online optimization of parameters such as grate speed control mode, grate speed control cycle, and grate pressure setpoints is achieved, which are then used as inputs to the grate cooler control module.

[0013] Therefore, the present invention has the following beneficial effects: The intelligent control method for grate coolers based on cement kiln operating condition identification of the present invention improves the existing grate cooler control scheme. By utilizing kiln operating condition identification technology, it collects kiln operating condition parameter data to achieve mutual matching between the grate cooler operating parameters and the cement kiln operating conditions. It realizes the automatic optimization setting of the grate pressure setpoint under typical operating conditions, without relying on the operator's skills and experience. This avoids the problem that frequently modifying and determining the grate pressure setpoint manually requires a long period of adjustment and exploration, and that the operator's skills and experience vary, making it impossible to guarantee that the grate cooler's operating control parameters are at the optimal value under the current operating conditions. Attached Figure Description

[0014] Figure 1 This is a flowchart of the method of the present invention;

[0015] Figure 2 This is a schematic diagram of the kiln K-line of the present invention;

[0016] Figure 3 This is a schematic diagram illustrating the reasoning behind the comprehensive kiln operating condition intensity index K13 and the comprehensive kiln operating condition stability index K23 of the present invention.

[0017] Figure 4 This is a schematic diagram illustrating the reasoning behind the kiln condition strength correction coefficient K12 and kiln condition stability correction coefficient K22 of the present invention.

[0018] Figure 5 This is a schematic diagram of the grate cooler optimization module of the present invention;

[0019] Figure 6 This is a schematic diagram of the control module for the grate cooler of the present invention. Detailed Implementation

[0020] The present invention will now be further described with reference to the accompanying drawings and specific embodiments.

[0021] like Figure 1 As shown, an intelligent control method for a grate cooler based on cement kiln operating condition identification includes the following steps:

[0022] S1: Acquire kiln current signal data and preprocess the kiln current signal data;

[0023] S2: Obtain the kiln current K-line chart through the kiln current K-line analysis method; S2 specifically involves: within the kiln current interception period, connecting the median kiln current M1 of the first half of the period and the median kiln current M2 of the second half of the period to form a rectangular cylinder, defining the area enclosed by this cylinder as the kiln current entity. The highest kiln current value H1 and the lowest kiln current value L1 within the interception period are then connected to the kiln current entity with thin lines to obtain the kiln current K-line chart. The line connecting the highest kiln current value H1 to the kiln current entity is the upper shadow line, and the line connecting the lowest kiln current value L1 to the kiln current entity is the lower shadow line. If the kiln current median M1 is greater than the kiln current median M2, the kiln current K-line is defined as a bearish line; if the kiln current median M1 is less than the kiln current median M2, the kiln current K-line is defined as a bullish line. This invention proposes a kiln current K-line identification method. The kiln current K-line diagram serves as a link between the dynamic and static processes of the kiln system, efficiently extracting key information about the kiln current, effectively reducing data noise, and facilitating the summarization of experience by kiln operators and process experts.

[0024] S3: Input the kiln current K-line chart into the expert knowledge base, and use the inference engine to derive the kiln operating condition intensity index K11 and the kiln operating condition stability index K21. Specifically, S3 involves inputting the kiln current K-line chart into the expert knowledge base. The inference engine starts from the first rule in the expert knowledge base and sequentially scans all rules, matching the rule premises with the current facts in the comprehensive database. When a rule conflict occurs, a depth-first strategy is adopted; after a high-weight rule is triggered, low-weight rules are ignored in this search. Based on the kiln current K-line chart, the intensity and stability of the current kiln system operating condition are inferred according to the rules, resulting in the kiln operating condition intensity index K11 and the kiln operating condition stability index K21. The kiln operating condition intensity index K11 and the kiln operating condition stability index K21 can basically reflect changes in the kiln condition, but to fully consider the impact of changes in other parameters on the kiln condition, further corrections are needed for the kiln operating condition intensity index K11 and the kiln operating condition stability index K21.

[0025] S4: Collect the first real-time parameter data of kiln operation. Based on the first real-time parameter data of kiln operation, deduce the kiln operation strength correction coefficient K12 and the kiln operation stability correction coefficient K22. The first real-time parameter data in S4 includes the kiln system feed rate, flue gas NOx, secondary air temperature, tertiary air temperature, decomposer outlet temperature, free calcium hourly value, raw meal silica ratio, raw meal alumina ratio, raw meal lime saturation coefficient, decomposer coal feed rate, and kiln head coal feed rate. Specifically, S4 involves decreasing the weight of the first real-time parameters according to their timeliness, and weighting the strength index and stability index of each kiln condition real-time parameter to derive the kiln operation strength correction coefficient K12 and the kiln operation stability correction coefficient K22. Using the kiln system parameters under typical operating conditions as a benchmark, by comparing the real-time kiln system parameters with the historical parameters of typical kiln conditions, the strength index and stability index of each parameter under the current operating conditions are given. Based on the timeliness of each parameter and its impact on the kiln condition, the weight of the parameters on the kiln condition is dynamically adjusted.

[0026] S5: Correct the kiln condition strength index K11 and the kiln condition stability index K21 according to the correction coefficient to obtain the comprehensive kiln condition strength index K13 and the comprehensive kiln condition stability index K23; S5 specifically: divide the kiln condition stability index K21 by the kiln condition stability correction coefficient K22 to obtain the comprehensive kiln condition stability index K23, and multiply the kiln condition strength correction coefficient K12 by the kiln condition strength index K11 to obtain the comprehensive kiln condition strength index K13.

[0027] S6: Collect the second real-time parameter data of kiln operation. Based on the second real-time parameter data of kiln operation, the comprehensive kiln operation intensity index K13, and the comprehensive kiln operation stability index K23, the grate pressure setpoint, grate speed control cycle, and grate speed control parameters are obtained through the optimization algorithm library. The second real-time parameter data includes grate pressure, oil pump pressure, secondary air temperature, air temperature, clinker temperature, blower current, and kiln system feed rate. Specifically, S6 involves: collecting the second real-time parameter data of kiln operation; correlating the kiln system operation, kiln system stability, grate cooler operation, and other operating conditions judgment conditions with the actual operating data based on the second real-time parameter data of kiln operation, the comprehensive kiln operation intensity index K13, and the comprehensive kiln operation stability index K23; inferring the trend of relevant parameters of the grate cooler system; and recommending grate speed control parameters, grate speed control cycle, and grate pressure setpoint based on historical data, parameter range, and trend changes. After obtaining the comprehensive kiln operating condition strength index K13 and the comprehensive kiln operating condition stability index K23, parameters such as the kiln system feed rate, grate pressure, oil pump pressure, secondary air temperature, air temperature, clinker temperature, and fan current are collected. Using the optimization algorithm library, parameters such as grate speed control mode, grate speed control cycle, and grate pressure setpoint are optimized online and used as inputs for the grate cooler control module.

[0028] S7: The grate pressure setpoint, grate speed control cycle, and grate speed control parameters are used as inputs to the grate cooler control module to achieve intelligent control of the grate cooler.

[0029] This invention provides an intelligent control method for grate coolers based on cement kiln operating condition identification. This method improves existing grate cooler control schemes by utilizing kiln operating condition identification technology to collect kiln operating condition parameter data. It achieves mutual matching between grate cooler operating parameters and cement kiln operating conditions, enabling automatic optimization of the grate pressure setpoint under typical operating conditions. This method does not rely on operator skill or experience, avoiding the problems of frequent manual modifications to determine the grate pressure setpoint, which require lengthy adjustments and adjustments, and where operator skill and experience vary, making it impossible to guarantee that the grate cooler's operating control parameters are at their optimal values ​​under the current operating conditions.

[0030] Example 1: In this example, a more detailed description of the intelligent control method for a grate cooler based on cement kiln operating condition identification according to the present invention is provided:

[0031] This invention, based on cement kiln operating condition identification, aims to improve the heat recovery efficiency of cement kilns by utilizing kiln operating condition identification technology to infer changes in clinker granulation and material layer, predict changes in secondary air temperature and grate cooler heat recovery efficiency, and make adjustments in advance. This invention includes a kiln condition identification module, a grate pressure optimization module, and a grate cooler control module.

[0032] (1) Kiln Condition Identification Module

[0033] For cement production, the kiln main unit current (kiln current) is the key parameter that best reflects the operating conditions of the kiln system. This solution uses the kiln current as an important basis for identifying the kiln operating conditions. It also draws on the K-line analysis method commonly used in the futures market to eliminate noise from the cement kiln main unit current and extract key information.

[0034] First, the kiln current data is preprocessed using a data preprocessing module to remove abnormal data. Then, a median filtering algorithm is used to eliminate occasional pulse interference.

[0035] Then, the kiln current K-line chart is obtained through kiln current K-line analysis. Within the intercepted period, the median kiln current M1 of the first half of the period and the median kiln current M2 of the second half of the period are connected to form a narrow rectangular column. The area enclosed by this column is defined as the kiln current entity. The highest kiln current value H1 and the lowest kiln current value L1 within the intercepted period are connected by a thin line. The line connecting the highest kiln current value H1 to the kiln current entity is called the upper shadow line, and the line connecting the lowest kiln current value L1 to the kiln current entity is called the lower shadow line. Figure 2 As shown. If the median kiln current M1 is greater than the median kiln current M2, then the kiln current K-line is defined as a negative line; if the median kiln current M1 is less than the median kiln current M2, then the kiln current K-line is defined as a positive line.

[0036] Examples of kiln current candlestick charts include half-hour, hourly, 2-hour, and 4-hour charts. The expert knowledge base primarily comprises a set of rules based on the experience of operators and process engineers in judging kiln current. These rules include, but are not limited to, using the number of bullish and bearish candlesticks to represent the overall trend, the size of the candlestick body to represent the internal dynamics and trend strength, and the length of the shadows to predict changes in kiln current inflection points.

[0037] like Figure 3 As shown, the inference engine starts with the first rule in the expert knowledge base and sequentially scans all rules in the knowledge base, matching the rule premises with the current facts in the comprehensive database. When a conflict occurs between matching rules, a depth-first strategy is adopted; once a high-weight rule is triggered, low-weight rules are ignored in this search. Based on the rules, the strength and stability of the current kiln system operating condition are inferred, resulting in the kiln operating condition strength index K11 and the kiln operating condition stability index K21.

[0038] The kiln condition intensity index K11 and the kiln condition stability index K21 can basically reflect the changes in kiln conditions. However, in order to fully consider the impact of changes in other parameters on kiln conditions, the kiln condition intensity index K11 and the kiln condition stability index K21 are further modified.

[0039] The real-time database collects real-time data on kiln condition parameters such as kiln system feed, flue gas NOx, secondary air temperature, tertiary air temperature, decomposer outlet temperature, free calcium hourly value, raw meal silica ratio, raw meal alumina ratio, raw meal lime saturation coefficient, decomposer coal feed, and kiln head coal feed. Through a signal processing module, abnormal kiln condition parameters are removed. In this example, the NOx content in the kiln tail flue is processed by removing outliers using three times the standard deviation before linear interpolation. The NOx content in the kiln tail flue after outlier processing is then processed using a moving median average algorithm. Moving average filtering algorithms are used for kiln system feed rate, decomposer coal feed, and kiln head coal feed. Median filtering algorithms are used for secondary air temperature, tertiary air temperature, and decomposer outlet temperature.

[0040] The expert historical database records the range of kiln system parameters under typical operating conditions. The typical kiln condition parameter range is as follows: kiln system output range of 580-630 t / h, decomposition furnace outlet temperature of 860-900℃, secondary air temperature of 1150-1280℃, head coal of 12-16.5 t / h, and tail coal of 29-36.5 t / h.

[0041] (2) Parameter optimization module for grate cooler

[0042] Based on kiln system parameters under typical operating conditions, and by comparing real-time kiln system parameters with historical parameters under typical kiln conditions, the strength index and stability index of each parameter under the current operating condition are given. According to the timeliness of each parameter and its impact on the kiln condition, the weight of the parameters on the kiln condition is dynamically adjusted. In the example, the weights of free calcium hourly values, raw material silica ratio, raw material aluminum ratio, and raw material lime saturation coefficient are decreased according to their timeliness. The strength index and stability index of each kiln condition parameter are weighted to derive the kiln condition strength correction coefficient K12 and the kiln condition stability correction coefficient K22. Figure 4 As shown.

[0043] Divide the kiln condition stability index K21 by the kiln condition stability correction coefficient K22 to obtain the comprehensive kiln condition stability index K23. Multiply the kiln condition strength correction coefficient K12 by the kiln condition strength index K11 to obtain the comprehensive kiln condition strength index K13.

[0044] like Figure 5 As shown, after obtaining the comprehensive kiln operating condition intensity index K13 and the comprehensive kiln operating condition stability index K23, parameters such as the kiln system feed rate, grate pressure, oil pump pressure, secondary air temperature, air temperature, clinker temperature, and fan current are collected. Using the optimization algorithm library, parameters such as grate speed control mode, grate speed control cycle, and grate pressure setpoint are optimized online and used as inputs for the grate cooler control module.

[0045] The data preprocessing module includes judging the confidence level of the data and removing abnormal data. In the example, Fourier transform is used to decompose the oil pressure of each column to obtain the cycle of the grate bed. Then, the maximum value within the three cycles is processed by the moving average filtering algorithm before output.

[0046] The optimization algorithm library is composed of grate cooler air-material matching rules, kiln operator experience, and local optimization search algorithms. The historical database records the parameter ranges of the grate cooler system under typical operating conditions. The parameter ranges for typical operating conditions are: kiln system feed rate 580-630 t / h, grate speed adjustment range 4.0-5.6 times / minute, tertiary air temperature 1100-1280℃, upper limit of oil pressure for each column 140 Bar, and grate underpressure range 8000-11000 Pa.

[0047] The system collects data in a real-time database, including the comprehensive kiln operating condition intensity index K13, comprehensive kiln operating condition stability index K23, oil pump pressure, secondary air temperature, air temperature, clinker temperature, kiln stability index K23, and kiln system feed rate. It then correlates the operating conditions of the kiln system, kiln system stability, and grate cooler with actual operating data to infer trends in relevant parameters of the grate cooler system. Based on historical data, parameter ranges, and trend changes, the optimization algorithm library recommends grate speed control modes, grate speed control cycles, and grate underpressure setpoints, and constrains these optimization parameters through a safety protection module.

[0048] The safety protection module is mainly used for interaction with the operator to realize the safety protection of the grate cooler, including one-key cut-off function, bumpless switching function, upper limit protection of grate pressure setting value, lower limit protection of grate pressure setting value, upper limit of grate pressure, lower limit of grate pressure, upper limit protection of grate cooler speed, lower limit protection of grate cooler speed, grate speed control cycle limit, and grate speed single cycle limit protection.

[0049] (3) Grate cooler control module

[0050] like Figure 6 As shown, the grate cooler control module adjusts its control strategy based on the grate speed control mode, grate speed control cycle, and grate pressure setpoint recommended by the grate cooler optimization module. According to the optimization module settings, the grate cooler controller can adjust the grate speed control mode.

[0051] In cases of significant kiln condition fluctuations, to prevent the operation of the grate cooler from exacerbating these fluctuations, expert-mandated control methods are employed to optimize the mechanical characteristics of the grate cooler and reduce its operating frequency while ensuring the oil pump pressure remains within limits.

[0052] When the output of the kiln system changes, an expert rule mode is adopted to quickly respond to changes in materials and avoid excessive pressure under the grate.

[0053] Under stable kiln conditions, the local optimization search algorithm in the optimization algorithm library is used to recommend the grate pressure setpoint with the goal of increasing the secondary air temperature, and the model prediction algorithm is used to achieve stable control of the grate pressure.

[0054] Example 2: In this example, the intelligent control method for grate coolers based on cement kiln operating condition identification of the present invention is described in detail in a specific application scenario:

[0055] First, the kiln current is processed. In this case study, the kiln current ranges from 800-1300A, with a sampling period of 1 second. Median filtering is applied to the kiln current over a period of 3 minutes, followed by maximum value filtering over a period of 1 minute. Using kiln current data from 7 PM to 11 PM on October 31, 2022, as an example, the filtered kiln current is calculated for 4-hour, 2-hour, and 1-hour candlestick charts. The results show that the 4-hour candlestick body range is 1128.5-1126.0A, with an upper shadow of 64.1 and a lower shadow of 32.8. Since the kiln current shows a slight downward trend, it is considered a weak bearish candlestick. The upper shadow is calculated by subtracting the average kiln current from 7 PM to 9 PM from the maximum value of the kiln current from 7 PM to 9 PM, and the lower shadow is calculated by taking the minimum kiln current from 9 PM to 11 PM. The 2-hour candlestick's body range is 1123.4-1128A, with the kiln current showing a slight upward trend, indicating a weak bullish candlestick. The lower shadow is calculated using the minimum kiln current value between 9 PM and 10 PM, and the upper shadow is calculated using the maximum kiln current value between 10 PM and 11 PM, resulting in an upper shadow of 64.6 and a lower shadow of 30.1. Similarly, the 1-hour candlestick's body range is calculated to be 1127.9-1146.5, with an upper shadow of 46.1 and a lower shadow of 40.6.

[0056] The calculated 4-hour, 2-hour, and 1-hour candlestick parameters are input into the expert knowledge base for inference. In the case study, the absolute value of the kiln current is considered weak if it is between 800-1000A, moderate if it is between 1000-1100A, strong if it is between 1100-1200A, and excessively strong if it is between 1200-1350A. The stability of the kiln condition is judged using the absolute values ​​of the upper and lower shadows. In the case study, it is agreed that the stability coefficient is 1.0-0.8 if the absolute value of the upper and lower shadows is between 0-50, 0.8-0.5 if it is between 50-100, 0.5-0.2 if it is between 100-150, and 0-0.2 if it is between 150-200. If the absolute value of the shadows is >200, the kiln condition is considered to be in a fluctuating state, and the stability coefficient is 0.

[0057] In this case study, the weight of the hourly candlestick chart is set at 0.5, the weight of the 2-hour candlestick chart at 0.3, and the weight of the 4-hour candlestick chart at 0.2. The calculated comprehensive kiln current is 1131.1A, the kiln operating condition intensity index K11 is 1.03, and the kiln condition shows an increasing trend. When processing the candlestick chart, the candlestick chart with the largest absolute value between the upper and lower shadows is selected. In this case study, the comprehensive candlestick index is 55.3, and the kiln operating condition stability index K21 is 0.73.

[0058] Based on historical data analysis under steady-state operating conditions, the tertiary air temperature range is 850-1000℃, the secondary air temperature range is 1050-1280℃, the kiln system output is 580-630t / h, the NOx concentration in the flue gas chamber is 600-1400ppm, the decomposition furnace outlet temperature is 860-900t, and the target range for free calcium is 0.5-1.2. The numerical update cycle for the kiln system on site is 8 hours, and the lag is large, so it is not adopted for the time being.

[0059] The expert rule base stipulates that the benchmark value for tertiary air temperature is 930℃, the benchmark value for secondary air temperature is 1100℃, the benchmark value for kiln system output is 630.0t / h, the benchmark value for NOx is 900ppm, the benchmark value for decomposition furnace outlet temperature is 880℃, and the benchmark value for free calcium is 1.0.

[0060] The expert rule base dynamically adjusts variable weights, which vary depending on the specific site. In this case, for the kiln operating condition intensity correction coefficient K12, during the free calcium renewal cycle, the weights are: free calcium 0.4, NOx 0.2, kiln system output 0.2, secondary blast temperature 0.1, tertiary blast temperature 0.05, and decomposer temperature 0.05. During the non-renewal cycle of free calcium, the weights are: NOx 0.4, free calcium 0.2, kiln system output 0.2, secondary blast temperature 0.1, tertiary blast temperature 0.05, and decomposer temperature 0.05.

[0061] The kiln operating condition stability correction coefficient K22, the weight of NOx is 0.6, the weight of secondary air temperature is 0.2, the weight of tertiary air temperature is 0.1, and the weight of decomposer temperature is 0.1. The standard deviation of each parameter is calculated. The kiln system output under steady-state conditions is a constant value and is not included in the calculation.

[0062] In this case, when processing NOx data, a 1-minute differential calculation is performed. If 0.1 < differential change rate < 20.0, the signal is considered feasible. If the differential is < 0.1, the signal is considered distorted. If the differential value is > 20, NOx backflushing is considered to be in progress, keeping the NOx signal unchanged from the previous cycle.

[0063] Taking 23:00 on October 31, 2022 as an example, the average temperature of the decomposition furnace was calculated to be 894.8℃ with a standard deviation of 1.88, the average temperature of the tertiary air was 998.4℃ with a standard deviation of 7.51, the NOx in the smoke chamber was 971.5℃ with a standard deviation of 63.1, the secondary air temperature was 1138.8℃ with a standard deviation of 51.3, the free calcium was 0.84 (the values ​​at that time had not been updated), and the kiln system output was 630t / h.

[0064] The calculated kiln operating condition strength correction factor K12 is 1.44 (K12 lower limit is 0.0, upper limit is 2.0).

[0065] The overall standard deviation of the kiln condition is 1.88*0.1+63.1*0.6+51.3*0.2+7.51*0.1=49.1.

[0066] The expert rules stipulate that for a kiln condition deviation of 0-40, K22 is 1.0; for a kiln condition deviation of 40-60, K22 is 1.0-1.2; for a kiln condition deviation of 60-120, K22 is 1.2-1.5; for a kiln condition deviation of 120-200, K22 is 1.5-2.0; and for a kiln condition deviation greater than 200, K22 is 2.0.

[0067] In this example, the kiln operating condition stability correction coefficient K22 is calculated to be 1.1.

[0068] The comprehensive kiln operating condition intensity index K13 = 1.03 * 1.44 = 1.48.

[0069] The overall kiln operating stability index K23 = 0.73 / 1.1 = 0.66.

[0070] The grate cooler optimization algorithm library is mainly based on the kiln condition comprehensive strength index, kiln condition stability index, hourly average pressure under the grate, hourly variance of pressure under the grate, average secondary air temperature, and the upper limit of oil pressure for rows 1-6 of the grate cooler.

[0071] For the grate cooler, the maximum value of the oil pressure in columns 1-6 is filtered over 2 minutes, followed by a 1-minute first-order inertial filter. At the same time, the operator sets the target value of the secondary air temperature and the set value of the grate pressure. In this case, the second chamber grate of the grate cooler is recommended.

[0072] Taking the grate pressure setpoint at 23:00 on October 31, 2022 as an example, since the kiln system output is close to the upper limit of 630t / h under steady-state conditions, the impact of the kiln system output on the grate pressure can be disregarded.

[0073] The upper limit setting for oil pressure in columns 1-6 is 140 Bar, and the secondary air temperature setting is 1150℃. The average oil pressure values ​​for columns 1-6 during the time period were 126.4 Bar, 127.2 Bar, 133.5 Bar, 135.2 Bar, 116.2 Bar, and 110.6 Bar, respectively. Since none of the oil pressures exceeded the limits, D (the estimated value of oil pressure for airflow) = 0.

[0074] If the oil pressure exceeds the limit, D (the estimated value of oil pressure and air volume) = (a certain oil pressure - 140) * Ka, where the coefficient Ka is obtained based on the model relationship between oil pressure and grate speed.

[0075] The current optimized value of the cooler under-grate pressure is 8400 Pa. The average value of the under-grate pressure in the previous week was 8273 Pa, with a standard deviation of 68.4. The optimization period of the under-grate pressure is 20 minutes. If it is the first cycle of optimization, the reference value is the average value of the under-grate pressure in the previous cycle. If the variance of the under-grate pressure is greater than 120, it is considered that the current operating condition of the cooler fluctuates and the set value optimization is not carried out. When the comprehensive stability index K23 of the kiln condition < 0.2, the under-grate pressure in this cycle is not optimized.

[0076] The optimized value of the under-grate pressure = (8400 * 0.7 + 8273 * 0.3) + D (the optimized value component of the oil pressure) + Kb * (K13 - 1) * K23 = 8361.9 + 200 * (1.41 - 1) * 0.61 = 8412 Pa

[0077] The coefficient Kb is obtained from the long-term model relationship between the secondary air temperature and the under-grate pressure.

[0078] If the optimized value is higher than the upper limit of 8600 Pa, the upper limit is taken. If the optimized value is lower than 7600 Pa, the lower limit is taken. Obviously, the latest optimized value of the under-grate pressure in this round is 8412 Pa. That is, under the current kiln condition, due to the rise of the kiln condition, there is room for the secondary air temperature to rise. Maintaining the current under-grate pressure or appropriately increasing the under-grate pressure is conducive to increasing the secondary air temperature.

[0079] At the same time, the relationship between the comprehensive stability index K23 of the kiln condition and the adjustment period of the cooler grate speed is as follows: 0.7 < K23 < 1, the kiln condition is stable, the adjustment period of the cooler is 10 seconds, and according to the kiln condition, the appropriate under-grate pressure and grate speed are matched to optimize the secondary air temperature; 0.4 < K23 < 0.7, the adjustment period of the cooler is 30 seconds; 0.2 < K23 < 0.4, the adjustment period of the cooler is 60 seconds; when K23 < 0.2, the under-grate pressure circuit is cut off at this time, and the controller mainly controls that the oil pressure does not exceed the upper limit.

[0080] The present invention proposes a method for identifying the K line of the kiln current. The K line diagram of the kiln current is the link between the dynamic process and the static process of the kiln system, which efficiently extracts the key information of the kiln current, effectively reduces data noise, and is convenient for kiln operators and process experts to summarize experience.

[0081] The control algorithm of the cooler in the present invention is optimized based on typical operating conditions. The algorithm library combines the experience of operators and process personnel. Based on the typical operating conditions of the cement kiln and the cooler, the optimization of parameters such as the under-grate pressure of the cooler and the control period of the cooler is carried out, which greatly reduces the complexity of the algorithm and improves the interpretability of the algorithm.

[0082] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any change or replacement that can be thought of without creative labor shall be covered within the protection scope of the present invention.

Claims

1. A grate cooler intelligent control method based on cement kiln working condition identification, characterized in that, Includes the following steps: S1: Acquire kiln current signal data and preprocess the kiln current signal data; S2: Obtain the kiln current K-line chart through the kiln current K-line analysis method; within the kiln current interception period, connect the median kiln current M1 of the first half of the period and the median kiln current M2 of the second half of the period to form a rectangular cylinder, and define the area enclosed by it as the kiln current entity. Select the highest kiln current value H1 and the lowest kiln current value L1 within the interception period, and connect H1 and L1 to the kiln current entity with thin lines to obtain the kiln current K-line chart. The line connecting H1 and the kiln current entity is the upper shadow line of the kiln current, and the line connecting L1 and the kiln current entity is the lower shadow line of the kiln current. If M1 is greater than M2, the K-line is defined as a negative line; if M1 is less than M2, the K-line is defined as a positive line. S3: Input the kiln current K-line graph into the expert knowledge base, and use the inference engine to deduce the kiln operating condition intensity index K11 and the kiln operating condition stability index K21. S4: Collect the first real-time parameter data of the kiln operating condition, and deduce the kiln operating condition strength correction coefficient K12 and kiln operating condition stability correction coefficient K22 based on the first real-time parameter data of the kiln operating condition. S5: The kiln condition strength index K11 and the kiln condition stability index K21 are corrected according to the correction coefficient to obtain the comprehensive kiln condition strength index K13 and the comprehensive kiln condition stability index K23. S6: Collect the second real-time parameter data of kiln operation. Based on the second real-time parameter data of kiln operation, the comprehensive kiln operation strength index K13 and the comprehensive kiln operation stability index K23, obtain the grate pressure setpoint, grate speed control cycle and grate speed control parameters through the optimization algorithm library. S7: The grate pressure setpoint, grate speed control cycle, and grate speed control parameters are used as inputs to the grate cooler control module to achieve intelligent control of the grate cooler.

2. The grate cooler intelligent control method based on cement kiln working condition identification according to claim 1, characterized in that, S3 specifically involves: inputting the kiln current K-line chart into the expert knowledge base; the inference engine starting from the first rule in the expert knowledge base and sequentially scanning all rules in the expert knowledge base; matching the rule premises with the current facts in the comprehensive database; when a conflict occurs in the matching rules, a depth-first strategy is adopted; after a high-weight rule is triggered, the low-weight rule is ignored in this search; and the intensity and stability of the current kiln system operating condition are inferred according to the rules based on the kiln current K-line chart, namely the kiln operating condition intensity index K11 and the kiln operating condition stability index K21.

3. The grate cooler intelligent control method based on cement kiln working condition identification according to claim 1, characterized in that, The first real-time parameter data in S4 includes the kiln system feed rate, flue gas NOx, secondary air temperature, tertiary air temperature, decomposer outlet temperature, free calcium hourly value, raw meal silica ratio, raw meal aluminum ratio, raw meal lime saturation coefficient, decomposer coal feed rate, and kiln head coal feed rate.

4. The grate cooler intelligent control method based on cement kiln working condition identification according to claim 1 or 3, characterized in that, Specifically, S4 involves: decreasing the weight of the first real-time parameter according to its timeliness, and weighting the intensity index and stability index of each kiln condition real-time parameter to obtain the kiln condition intensity correction coefficient K12 and the kiln condition stability correction coefficient K22.

5. The grate cooler intelligent control method based on cement kiln working condition identification according to claim 1, characterized in that, Specifically, S5 involves dividing the kiln condition stability index K21 by the kiln condition stability correction coefficient K22 to obtain the comprehensive kiln condition stability index K23, and multiplying the kiln condition strength correction coefficient K12 by the kiln condition strength index K11 to obtain the comprehensive kiln condition strength index K13.

6. The grate cooler intelligent control method based on cement kiln working condition identification according to claim 1, characterized in that, The second real-time parameter data includes grate pressure, oil pump pressure, secondary air temperature, air temperature, clinker temperature, blower current, and kiln system feed rate.

7. A method for intelligent control of a grate cooler based on cement kiln operating condition identification according to claim 1 or 6, characterized in that, S6 specifically involves: collecting the second real-time parameter data of the kiln operating condition; correlating the kiln system operating condition, kiln system stability, and grate cooler operating condition judgment conditions with the actual operating condition data based on the second real-time parameter data of the kiln operating condition, the comprehensive kiln operating condition intensity index K13, and the comprehensive kiln operating condition stability index K23; inferring the trend of the occurrence of relevant parameters of the grate cooler system; and recommending grate speed control parameters, grate speed control cycle, and grate pressure setpoints based on historical data, parameter ranges, and trend changes using the optimization algorithm library.

Images