Fuel gas radiant tube abnormity identification method based on combustion ion current
By analyzing the combustion ion current data and establishing a combustion abnormality identification model, the problem of difficulty in accurately monitoring the combustion stability of gas radiation tube status monitoring is solved, and the effect of early warning and avoiding systemic failures is achieved.
Patent Information
- Application Number
- CN202411863592.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-17
- Publication Date
- 2025-04-25
AI Technical Summary
The status monitoring of existing gas radiation tubes is difficult to accurately monitor combustion stability, which makes it difficult to detect combustion abnormalities in time, which may cause equipment damage and safety hazards.
By analyzing the combustion ion current data, two-mean clustering, data cleaning, abnormality rate setting, local outlier factor detection and other methods are used to establish a equipment combustion abnormality recognition model, and give an abnormality warning of gas radiation tubes in advance.
Accurate monitoring of the combustion status of gas radiation tubes is achieved, early warning is made and combustion abnormalities are avoided to systematic failures, ensuring safe and stable operation of the equipment.
Smart Images
Figure CN119885006A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method for identifying abnormalities of a gas radiant tube based on combustion ion current. Background Art
[0002] The heating zone of the continuous annealing furnace uses a gas radiant tube to heat the cold hard coil from room temperature to the recrystallization annealing temperature, reduce the strength of the strip and improve its shaping to obtain better mechanical processing properties. The gas radiant tube is an industrial heating equipment, usually used in industrial furnaces, heat treatment furnaces, drying equipment and other occasions that require thermal energy. It uses gas as fuel, generates heat through combustion, and transfers the heat to the object or space that needs to be heated by radiation. The combustion stability of the gas radiant tube is an important technical indicator to measure the performance of the continuous annealing furnace. In the long-term production process, due to changes in the ratio of the mixed gas, fluctuations in calorific value, blockage of the gas or combustion air solenoid valve of the gas radiant tube, loosening of the flow regulating nut and a series of other problems will have a huge impact on the combustion effect of the gas radiant tube. Due to the lack of monitoring of the gas combustion safety system, it is difficult to detect the abnormal combustion of the gas radiant tube burner in time. The long-term existence of abnormal conditions such as unstable combustion, flame not shutting down, and oxygen-poor combustion will cause damage to the gas radiant tube.
[0003] The stability of gas radiant tube combustion not only affects the economic and environmental performance of production, but also the personal safety of the crew members, so monitoring its combustion status is of great significance. An accurate and reliable combustion stability monitoring system can judge the combustion status of gas radiant tubes based on combustion data, thereby guiding the unit operators to extract the current operating conditions based on the combustion history data, providing a reference for real-time production operation.
[0004] The automatic control and optimization calculation of the furnace temperature of the existing continuous annealing furnace are carried out through a mathematical model. The main function of the mathematical model is to automatically calculate the gas demand according to the actual strip annealing temperature requirements to ensure the stability of the strip temperature. However, due to the increase in the number of steel types of existing products, the production schedule is changeable, and the specifications vary greatly. On the one hand, the operation is complicated, and on the other hand, due to too much manual intervention, it is not conducive to the strict implementation of standardized operations. The original design of the combustion controller of the annealing furnace is difficult to adapt to the characteristics of time-varying, nonlinear and uncertainty, and it is difficult to accurately monitor the combustion stability of the gas radiant tube.
[0005] The combustion ion current is the current generated by the separation of charges due to the chemical reaction of the fuel and the oxidant during the combustion process. In the combustion area, the high temperature causes the gas molecules to ionize, generating free electrons and positive and negative ions. When these charged particles move in a directional manner under the action of an external electric field, a combustion ion current is formed. This method proposes that by analyzing the combustion ion current data, an early warning of radiation tube abnormalities can be issued one day or more in advance, so as to achieve the goal of early maintenance and avoid the development of systemic failures. Summary of the invention
[0006] The technical problem to be solved by the present invention is to provide a method for identifying abnormalities of gas radiant tubes based on combustion ion current. The method overcomes the defects of traditional gas radiant tube status monitoring, and gives an early warning of abnormalities of the gas radiant tubes by analyzing and monitoring the combustion ion current data, thereby achieving the purpose of early maintenance and avoiding the development of systemic failures, thereby eliminating safety hazards.
[0007] In order to solve the above technical problems, the gas radiant tube abnormality identification method based on combustion ion current of the present invention comprises the following steps: Step 1: Collect and count the historical data of combustion ion current of all gas-fired radiation tubes in the continuous annealing furnace, summarize the historical data of combustion ion current and solve the two-mean clustering to obtain v_l and v_h, where v_l is the mean value of combustion ion current in non-combustion state, and v_h is the mean value of combustion ion current in combustion state; Step 2: Data cleaning: remove values less than 0 and significant abnormal values greater than 1.5 v_h in the combustion ion current data. The significant abnormal values are abnormal values caused by the abnormality of the non-gas radiant tube itself. Step 3, set the abnormal rate C, take all the combustion ion current data closer to v_h than v_l as the combustion ion current value in the combustion state, and take all the combustion ion current data closer to v_l than v_h as the combustion ion current value in the non-combustion state, examine the two sample standard deviations σ_l and σ_h, and count the data volume in the intervals [3σ_l,v_h-3σ_h] and [v_h+3σ_h,1.5v_h] of all the combustion ion current data to the total sample ratio as the abnormal rate C; Step 4: Outlier detection, using outlier detection based on local outlier factor LOF. , Among them, LRD(X) is the local reachability density of data point X, which represents the average distance between data point X and its K nearest neighbors. i) represents the average distance between the nearest neighbor of data point X and its K nearest neighbors. LOF(X) is the local outlier factor of data point X. The LOF value represents a ratio, which represents the ratio of the LRD of data point X to the average LRD of the K nearest neighbors of data point X. The higher the ratio, the greater the probability of considering X as an outlier. Step 5: Arrange the LOF values in the entire data set in descending order, and consider the data points with LOF values within the abnormal rate C range as outliers; Step 6: Identify and determine the abnormality of the gas radiant tube. Determine the flame instability event. If more than 50% of the outliers of the combustion ion current of the gas radiant tube are in the interval [4,0.9v_h], it is determined that the gas radiant tube has a flame instability abnormality and a warning is issued. Determine the flame non-shutdown event. If more than 50% of the outliers of the combustion ion current of the gas radiant tube are in the interval [1,4], it is determined that the flame non-shutdown abnormality exists in the gas radiant tube and a warning is issued. Determine the long-term flameless event. If the proportion p of the outliers of the combustion ion current of the gas radiant tube in the interval [0,1] is higher than the proportion p of the normal state by more than 50%, it is determined that the gas radiant tube has a long-term flameless abnormality and a warning is issued.
[0008] Furthermore, in the step 1, all the gas-fired radiation tubes of the continuous annealing furnace are gas-fired radiation tube arrays of the same model, and the historical period of the combustion ion current historical data is one month.
[0009] Furthermore, the significant abnormal value in step 2 is caused by an abnormality in the data acquisition sensor or acquisition system.
[0010] Since the method for identifying abnormalities of gas radiant tubes based on combustion ion current of the present invention adopts the above technical scheme, that is, the method targets the continuous annealing furnace, selects the combustion ion current of the gas radiant tube as a parameter, generates characteristic values that can describe the combustion stability based on the analysis of the combustion ion current data and business experience, and analyzes the combustion ion current data describing the combustion stability of each gas radiant tube at the same production stage and the same heating temperature, establishes an equipment combustion anomaly identification model based on rules and outlier detection algorithms, and determines the gas radiant tubes that may have combustion anomalies and the corresponding abnormal event types. The method overcomes the defects of traditional gas radiant tube status monitoring, and gives an early warning of gas radiant tube anomalies by analyzing and monitoring the combustion ion current data, so as to achieve the purpose of early maintenance and avoid the development of systemic failures, and eliminate safety hazards. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments: Figure 1 The present invention is a flowchart of a method for identifying abnormalities in a gas radiant tube based on combustion ion current. DETAILED DESCRIPTION
[0012] Implementation example Figure 1 As shown, the method for identifying abnormalities of a gas radiant tube based on combustion ion current of the present invention comprises the following steps: Step 1: Collect and count the historical data of combustion ion current of all gas-fired radiation tubes in the continuous annealing furnace, summarize the historical data of combustion ion current and solve the two-mean clustering to obtain v_l and v_h, where v_l is the mean value of combustion ion current in non-combustion state, and v_h is the mean value of combustion ion current in combustion state; Step 2: Data cleaning: remove values less than 0 and significant abnormal values greater than 1.5 v_h in the combustion ion current data. The significant abnormal values are abnormal values caused by the abnormality of the non-gas radiant tube itself. Step 3, set the abnormal rate C, take all the combustion ion current data closer to v_h than v_l as the combustion ion current value in the combustion state, and take all the combustion ion current data closer to v_l than v_h as the combustion ion current value in the non-combustion state, examine the two sample standard deviations σ_l and σ_h, and count the data volume in the intervals [3σ_l,v_h-3σ_h] and [v_h+3σ_h,1.5v_h] of all the combustion ion current data to the total sample ratio as the abnormal rate C; Step 4: Outlier detection, using outlier detection based on local outlier factor LOF. Among them, LRD(X) is the local reachability density of data point X, which represents the average distance between data point X and its K nearest neighbors. i ) represents the average distance between the nearest neighbor of data point X and its K nearest neighbors. LOF(X) is the local outlier factor of data point X. The LOF value represents a ratio, which represents the ratio of the LRD of data point X to the average LRD of the K nearest neighbors of data point X. The higher the ratio, the greater the probability of considering X as an outlier. Step 5: Arrange the LOF values in the entire data set in descending order, and consider the data points with LOF values within the abnormal rate C range as outliers; In traditional outlier detection based on outlier factor, the threshold of LOF(X) is directly given to determine which data points in the data set are outliers. This method uses the proportion of given outliers in the total sample, that is, the aforementioned abnormality rate C, to arrange the LOF values in the entire data set in descending order, and the data points with LOF values within the range of abnormality rate C are regarded as outliers. Step 6: Identify and determine the abnormality of the gas radiant tube. Determine the flame instability event. If more than 50% of the outliers of the combustion ion current of the gas radiant tube are in the interval [4,0.9v_h], it is determined that the gas radiant tube has a flame instability abnormality and a warning is issued. Determine the flame non-shutdown event. If more than 50% of the outliers of the combustion ion current of the gas radiant tube are in the interval [1,4], it is determined that the flame non-shutdown abnormality exists in the gas radiant tube and a warning is issued. Determine the long-term flameless event. If the proportion p of the outliers of the combustion ion current of the gas radiant tube in the interval [0,1] is higher than the proportion p of the normal state by more than 50%, it is determined that the gas radiant tube has a long-term flameless abnormality and a warning is issued.
[0013] This step is based on the principle of chemical reaction kinetics of gas combustion and the actual working condition of the gas radiant tube. The duration of the process from ignition to full stable combustion in the gas radiant tube is very short, which is an unstable combustion state. In this state, the ion current value in the gas radiant tube is in the interval of [4,0.9v_h]. The v_h value is obtained from the historical data of normal working of the gas radiant tube by clustering method, which represents the average value of the ion current value in the normal stable combustion state. The duration of the process from ignition to full stable combustion is related to the flame propagation speed and gas concentration diffusion in the field of chemical reaction kinetics. In the gas radiant tube studied by this method, the duration of the unstable combustion state under normal working conditions is extremely short. Therefore, the judgment rule for determining whether the gas radiant tube has flame instability anomaly is that if more than 50% of the outliers of the combustion ion current of the gas radiant tube are in the interval of [4,0.9v_h], it is determined that the gas radiant tube has flame instability anomaly and a warning is issued. Determine the flame non-shutdown event: According to the principle of chemical reaction kinetics of gas combustion and the actual working condition of the gas radiant tube, the gas radiant tube The flameout process is the blowout process of partial premixed combustion in the field of combustion principle. After the flame is blown out, some combustion intermediate products still move in the combustion reaction zone to generate ion current values in a very short time. The ion current value is in the range of [1,4]. For example, if there is an incompletely closed equipment failure such as gas valve leakage, there will be a weak combustion chemical reaction in the reaction zone of the gas radiation tube. Therefore, the judgment rule for determining that the gas radiation tube has a flame that does not shut down abnormally is that if more than 50% of the outliers of the combustion ion current of the gas radiation tube are in the range of [1,4], it is determined that the gas radiation tube has a flame that does not shut down abnormally, and a warning is issued; Determine the long-term flameless event: According to the principle of gas combustion chemical reaction kinetics and the actual working conditions of the gas radiation tube, the ion current value of the gas radiation tube in the flameless state is in the range of [0,1]. Therefore, the judgment rule for determining that the gas radiation tube has a long-term flameless abnormality is that if the proportion p of the outliers of the combustion ion current of the gas radiation tube in the range of [0,1] is higher than the proportion p of the normal state by more than 50%, it is determined that the gas radiation tube has a long-term flameless abnormality and a warning is issued. Since the abnormal rate C is set, a certain proportion of outliers will be detected in the combustion ion current data of the normal gas radiant tube. The combustion ion current of the normal gas radiant tube is concentrated near the peak value and 0 in the combustion state and non-combustion state. Therefore, most of the detected outliers are distributed in [0,1]U[0.9v_h,v_h] according to the proportion of the combustion state duration. The combustion ion current outliers detected in the three abnormal gas radiant tubes show differences in distribution. The abnormal state of the gas radiant tube can be determined based on the distribution of the combustion ion current outliers. The judgment rules of this method are based on statistical historical fault information, and flame instability events are determined first according to the probability of abnormal occurrence.
[0014] Preferably, in the step 1, all the gas-fired radiant tubes of the continuous annealing furnace are gas-fired radiant tube arrays of the same model, and the historical period of the combustion ion current historical data is one month.
[0015] Preferably, the significant abnormal value in step 2 is caused by an abnormality in the data acquisition sensor or acquisition system.
[0016] Usually, the abnormal status of gas radiant tubes is shown in the following table. Abnormal event name Abnormal event description Flame instability During combustion, the chemical reaction rate that characterizes the combustion intensity changes dramatically, showing obvious non-steady-state combustion. Long-term flame-free During the normal working cycle of the radiation tube array, a radiation tube does not work for a long time. Flame does not shut off After the radiation tube array ends a working state and blows out, a radiation tube still has a weak combustion phenomenon, and the flame is not shut off Under normal circumstances, the combustion ion current of the gas radiant tube is maintained near the peak value when it is in a combustion state, and remains at 0 in a non-combustion state. The gas radiant tube units in the gas radiant tube array are usually of the same model.
[0017] This method analyzes the combustion ion current data to identify the individual devices whose combustion conditions have changed due to the degradation of the state of the device itself, that is, the combustion ion current has obvious abnormalities, and prompts possible abnormal event alarms.
[0018] This method selects the combustion ion current of the gas radiant tube as a parameter for the continuous annealing furnace, generates a characteristic value that can describe the combustion stability based on the combustion ion current data analysis and business experience, and analyzes the current data describing the combustion stability of each equipment at the same production stage and the same heating temperature. An equipment combustion anomaly recognition model based on rules and outlier detection algorithms is established to determine the equipment that may have combustion anomalies and the corresponding abnormal event types.
[0019] The method has simple input data acquisition, low cost, easy deployment, low requirements on computing software resources during operation, can also be used for real-time monitoring, and can be adaptive to the status monitoring of different combinations of continuous annealing furnace gas radiant tube arrays.
[0020] A combustion ion current sensor is installed in the gas radiant tube of the continuous annealing furnace, and the sensor data collected is output and stored in a medium. Data cleaning is performed, and the abnormal rate of the combustion ion current data is given according to the historical sample data. For each combustion ion current data set of a single gas radiant tube within a fixed period, the outlier value is counted based on the local outlier factor. According to the given rules, it is determined whether each single gas radiant tube detected is in an abnormal state. If the detected gas radiant tube meets the alarm condition, an abnormal warning is output to the intelligent operation and maintenance platform to remind the staff to arrange equipment maintenance.
Claims
1. A method for identifying abnormalities in a gas radiant tube based on combustion ion current, characterized in that The steps include: Step 1: Collect and count the historical data of combustion ion current of all gas-fired radiation tubes in the continuous annealing furnace, summarize the historical data of combustion ion current and solve the two-mean clustering to obtain v_l and v_h, where v_l is the mean value of combustion ion current in non-combustion state, and v_h is the mean value of combustion ion current in combustion state; Step 2: Data cleaning: remove values less than 0 and significant abnormal values greater than 1.5 v_h in the combustion ion current data. The significant abnormal values are abnormal values caused by the abnormality of the non-gas radiant tube itself. Step 3, set the abnormal rate C, take all the combustion ion current data closer to v_h than v_l as the combustion ion current value in the combustion state, and take all the combustion ion current data closer to v_l than v_h as the combustion ion current value in the non-combustion state, examine the two sample standard deviations σ_l and σ_h, and count the data volume in the intervals [3σ_l,v_h-3σ_h] and [v_h+3σ_h,1.5v_h] of all the combustion ion current data to the total sample ratio as the abnormal rate C; Step 4: Outlier detection, using outlier detection based on local outlier factor LOF. ， in, LRD(X) is the local reachability density of data point X, which represents the average distance between data point X and its K nearest neighbors. LRD (X i ) Represents the average distance between the nearest neighbor of data point X and its K nearest neighbors. LOF(X) is the local outlier factor of data point X. The LOF value represents a ratio, which represents the ratio of the LRD of data point X to the average LRD of the K nearest neighbors of data point X. The higher the ratio, the greater the probability of considering X as an outlier. Step 5: Arrange the LOF values in the entire data set in descending order, and consider the data points with LOF values within the abnormal rate C range as outliers; Step 6: Identify and determine the abnormality of the gas radiant tube. Determine the flame instability event. If more than 50% of the outliers of the combustion ion current of the gas radiant tube are in the interval [4,0.9v_h], it is determined that the gas radiant tube has a flame instability abnormality and a warning is issued. Determine the flame non-shutdown event. If more than 50% of the outliers of the combustion ion current of the gas radiant tube are in the interval [1,4], it is determined that the flame non-shutdown abnormality exists in the gas radiant tube and a warning is issued. Determine the long-term flameless event. If the proportion p of the outliers of the combustion ion current of the gas radiant tube in the interval [0,1] is higher than the proportion p of the normal state by more than 50%, it is determined that the gas radiant tube has a long-term flameless abnormality and a warning is issued.
2. The method for identifying abnormality of a gas radiant tube based on combustion ion current according to claim 1, characterized in that: In the step 1, all the gas-fired radiation tubes of the continuous annealing furnace are gas-fired radiation tube arrays of the same model, and the historical period of the combustion ion current historical data is one month.
3. The method for identifying abnormality of a gas radiant tube based on combustion ion current according to claim 1, characterized in that: The significant abnormal value in step 2 is caused by abnormality of the data acquisition sensor or the acquisition system.