Static signal data balance ratio-based lean-burn flameout precursor detection method

Abstract

The invention relates to a lean burn flameout detection method based on an electrostatic signal data equalization ratio, a used detection device comprises an electrostatic sensor, a signal conditioning circuit and a data acquisition system, and the signal conditioning circuit realizes signal amplification and filtering; the electrostatic sensor electrode captures the movement of charged particles in the flame in a non-contact manner through an electrostatic induction way to obtain an induction current signal; through the signal conditioning circuit, the induction current signal is converted into a voltage signal to be collected by the data collection system, and then the lean burn flameout precursor is judged according to the signal data balance ratio.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of flame detection, and relates to a lean blowout precursor detection method based on electrostatic signal data equalization ratio. BACKGROUND

[0002] To reduce the emission of pollutants such as nitrogen oxides, aircraft engines and gas turbines are usually operated in a lean state. However, the lean state is unstable, and a flameout will cause serious accidents such as flight safety accidents or industrial shutdowns. Generally, there are precursor phenomena such as local flameout and flame front fluctuation enhancement before flameout. Therefore, timely detection of lean blowout precursors can realize blowout warning and provide protection for the safe operation of the combustor.

[0003] Currently, the detection of lean blowout precursors is mainly based on the flame light emission characteristics, and is detected by photodetectors or cameras. Photodetectors detect the timing signals of the light emission intensity of flame free radicals CH* or OH*; cameras capture the visible light emitted by the flame to form a flame image. However, the photodetector probe and the camera objective lens are easily affected by the high temperature of the flame, and need additional cooling equipment. For flame detection in a sealed combustion chamber, a window needs to be opened, which increases the complexity of the system. Therefore, photodetectors and cameras are usually used for detection of laboratory simulation combustion devices. In addition to light emission, flame combustion is also accompanied by ionization process, producing different kinds of charged particles. Ion probes are often used to detect charged particles in the flame, and by applying a bias, an ion current signal is generated. However, the ion probe belongs to contact measurement, which will cause certain disturbance to the flame. Electrostatic sensors can capture the movement of charged particles in the flame through electrostatic induction to obtain electrostatic signals without opening a window and without the need for cooling devices, and have strong environmental adaptability. Electrostatic sensors have been widely used in the field of gas-solid two-phase flow measurement, but have not been used for research on lean blowout precursor detection.

[0004] According to the reflection of lean blowout precursor in flame detection signal, researchers have proposed different kinds of lean blowout indicators. Li et al. used ion probe to detect the flame in a pulse combustor, and took the ratio of operating frequency energy to low frequency energy of ion current signal as lean blowout indicator [1]. However, the operating frequency energy in this indicator only exists in the flame generated by pulse combustor, and does not exist in the flame of other types of combustors. Yi et al. used photodetector to detect the swirling premixed flame, and took the ratio of standard deviation to mean of free radical OH*emission intensity signal as lean blowout indicator [2]. However, this indicator is not suitable for bluff body premixed flame, and there is a problem of misjudging lean blowout precursor as stable combustion state [3]. Chaudhari R R et al. used camera to shoot swirling premixed flame images, and took the ratio of red light intensity to blue light intensity of flame image as lean blowout indicator [4], but the applicability of this indicator to other types of flames is uncertain. Further research found that when the lean blowout precursor occurs, the amplitude probability distribution of detection signal changes from normal distribution to Rayleigh distribution, which is a common phenomenon in the study of lean blowout of bluff body premixed flame and swirling premixed flame [2][3][5]. De S et al. based on the symmetry of free radical CH*time series signal amplitude probability distribution, first took the skewness coefficient of free radical CH*time series signal as lean blowout indicator, but this indicator misjudges stable combustion state as lean blowout precursor [6].

[0005] References

[0006] [1]. Li F, Xu L, Du M, et al. Ion current sensing-based lean blowout detection for a pulse combustor[J]. Combustion and Flame, 2017, 176: 263-271.

[0007] [2]. Yi T, Gutmark E J. Real-time prediction of incipient lean blowout in gas turbine combustors[J]. AIAA Journal, 2007, 45(7): 1734-1739.

[0008] [3]. Chang L, Cao Z, Fu B, et al. Lean blowout detection for bluff-body stabilized flame[J]. Fuel, 2020, 266: 117008.

[0009] [4]. Chaudhari R R, Sahu R P, Ghosh S, et al. Flame color as a lean blowout predictor[J]. International Journal of Spray and Combustion Dynamics, 2013, 5(1): 49-65.

[0010] [5]. Mahesh S, Mishra D P. Dynamic sensing of blowout in turbulent CNG inversejet flame[J]. Combustion and Flame, 2015, 162(8): 3046-3052.

[0011] [6]. De S, Bhattacharya A, Mondal S, et al. Investigation of flame behavior and dynamics prior to lean blowout in a combustor with varying mixedness of reactants for the early detection of lean blowout[J]. International Journal of Spray and Combustion Dynamics, 2019, 11: 1756827718812519. SUMMARY

[0012] The present application is based on a non-contact electrostatic sensor, and proposes a lean blowout indicator based on the data equalization ratio of electrostatic signals, which realizes the precursor detection of lean blowout of the flame. The technical scheme of the present application is as follows:

[0013] A lean blowout detection method based on the data equalization ratio of electrostatic signals, the detection device used includes an electrostatic sensor, a signal conditioning circuit and a data acquisition system, wherein the signal conditioning circuit realizes signal amplification and filtering; the electrostatic sensor electrode captures the movement of charged particles in the flame through electrostatic induction to obtain an induced current signal; through the signal conditioning circuit, the induced current signal is converted into a voltage signal which is collected by the data acquisition system, and then the precursor of lean blowout is judged through the signal data equalization ratio; the steps are as follows:

[0014] (1) The induced current signal of the electrostatic sensor electrode is converted into an induced voltage signal after the signal conditioning circuit including amplification and filtering;

[0015] (2) Data acquisition system collects induced voltage signal {x1, x2, ···, xi, ···, xn}, wherein 1≤i≤n, n is the total data amount; calculate the average value i ,···,x n-1 ,x n} of xi, wherein 1≤i≤n, n is the total data amount; calculate the average value

[0016] (3) Compare the size of x i and , respectively calculate the number of data points n1 greater than and the number of data points n2 less than ;

[0017] (4) Take the absolute value after the difference between n1 and n2, and then compare it with n, defined as data balance ratio D;

[0018] (5) A predetermined stable combustion state and lean blowout precursor threshold; when D is less than the threshold, it is determined to be a stable combustion state; when D is greater than or equal to the threshold, it is determined to be a lean blowout precursor.

[0019] Further, the electrostatic sensor includes at least one electrode, which is arranged near the outlet of the burner, and the electrode faces the flame or is exposed to the outside, or is covered with a high-temperature-resistant insulating material as a separation layer; the other surface of the electrode is shielded by electrostatic shielding, and an insulating material is added between the metal shielding layer and the electrode.

[0020] Further, the threshold is 0.1.

[0021] Further, the thickness and axial length of the electrode are 2mm and 7mm respectively, and the electrode is placed 4mm away from the edge of the widest part of the flame.

[0022] The application has the following advantages:

[0023] (1) Compared with photoelectric detectors and cameras, the electrostatic sensor has better adaptability to harsh combustion environments.

[0024] (2) Compared with ion probes, the electrostatic sensor is a non-contact measurement and will not interfere with the flame.

[0025] (3) The lean blowout precursor detection method based on the data balance ratio of the electrostatic signal in the application is applicable to bluff body and swirl premixed flame, and has certain universality. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is the flow chart of the lean blowout detection of the application.

[0027] Figure 2 is the effect diagram of the lean blowout index of the application.

[0028] The reference signs are as follows:

[0029] 1, isolation layer; 2, electrode; 3, insulation layer; 4, shielding layer; 5, electrode lead-out wire; 6, electrostatic sensor. DETAILED DESCRIPTION

[0030] Reference Figure 1 Embodiments of the present application are introduced. The bluff body premixed flame generated by a bluff body burner is taken as the detection object. The electrostatic sensor takes a single electrode to measure the flame, and the thickness and axial length of the electrode are 2 mm and 7 mm respectively. The electrode 2 is made of copper, the insulation layer 3 is made of polyethylene, the shielding layer 4 is made of stainless steel, and the electrode surface is not added with the isolation layer 1. The distance between the electrostatic electrode and the flame is adjusted so that the electrode is not in contact with the flame, and it is ensured that the insulation layer of the electrostatic sensor is not burned by high temperature. At the same time, if the distance between the electrode and the flame is too large, the intensity of the electrostatic signal will be reduced, and the effective signal of the charged particle movement of the flame cannot be detected. Considering the above two aspects, the electrode is placed at a distance of 4 mm from the edge of the widest part of the flame for detection.

[0031] The basic detection steps of the present application are as follows:

[0032] (1) The current signal sensed by the electrostatic sensor electrode is converted into a voltage signal after the signal conditioning circuit including amplification and filtering;

[0033] (2) The data acquisition system collects the sensed voltage signals {x1, x2, ···, xn}, where 1≤i≤n, and n is the total data amount; and calculates the average value i ,···,x n-1 ,x n} of the sensed voltage signals. The average value is as shown in the following formula:

[0034]

[0035] (3) The size of x i and is compared, and the number of data points n1 greater than x i and the number of data points n2 less than x are calculated respectively, as shown in the following formula:

[0036]

[0037] (4) The absolute value of the difference between n1 and n2 is taken, and then compared with n, and defined as the data balance ratio D, as shown in the following formula:

[0038]

[0039] (5) The preset stable combustion state and lean combustion flameout precursor dividing threshold value; the present application recommends that the threshold value is 0.1, and the threshold value can also be fine-tuned in actual application; when D is less than 0.1, it is determined that the stable combustion state; when D is greater than or equal to 0.1, it is determined that the lean combustion flameout precursor.

[0040] The following is the experimental process of the present embodiment, and the value of the stable combustion state and lean combustion flameout precursor dividing threshold value is determined through experiments.

[0041] Step 1: The induced current signal generated by the electrode of the electrostatic sensor is amplified and filtered by the conditioning circuit to convert the signal into a voltage signal, which is collected by the data acquisition system. The data sampling rate is 25 kHz, the sampling time is 5 s, and the output electrostatic signal contains a total of 125000 data points, denoted as {x1, x2, ···, x i ,···,x 124999 ,x 125000}.

[0042] Step 2: The present application detects the equivalence ratio of the flame setting as 0.75, 0.8, 0.85, 0.85, 0.9, 0.95, 1.0, 1.1, 1.2 and 1.3. The average value of the electrostatic signal x under different equivalence ratios is calculated.

[0043] Step 3: Compare the size of x i and in the electrostatic signal under different equivalence ratios, respectively calculate the number of data points n1 greater than and the number of data points n2 less than .

[0044] Step 4: Take the absolute value of the difference between n1 and n2, and then divide by 125000 to calculate D under different equivalence ratios, as shown in Figure 2 . From Figure 2 , it can be seen that the stable combustion state and lean combustion flameout precursor dividing threshold value set to 0.1 is more appropriate.

Claims

1. A lean misfire detection method based on the ratio of electrostatic signal data equalization, the detection device used includes electrostatic sensor, signal conditioning circuit and data acquisition system, wherein, The signal conditioning circuit realizes signal amplification and filtering; the static sensor electrode captures the charged particle movement in the flame through the static induction method to obtain an induced current signal; the induced current signal is converted into a voltage signal by the signal conditioning circuit and collected by the data acquisition system, and then the lean blowout precursor is determined through signal data balance ratio; the steps are as follows: (1) the induced current signal of the static sensor electrode is converted into an induced voltage signal after the signal conditioning circuit including amplification and filtering; (2) The data acquisition system acquires induced voltage signals {x1, x2, ..., x i ,···,x n-1 ,x n }, where 1≤i≤n, and n is the total number of data; calculate its average value x; (3) comparison x i The number of data points greater than x, n1, and the number of data points less than x, n2, are calculated, respectively, based on the size of x. (4) the absolute value of the difference between n1 and n2 is obtained, and then the ratio of the absolute value to n is obtained, which is defined as the data balance ratio D; (5) a preset stable combustion state and lean blowout precursor threshold is set; when D is less than the threshold, it is determined that the combustion state is stable; when D is greater than or equal to the threshold, it is determined that the lean blowout precursor is detected.

2. The lean misfire detection method based on electrostatic signal data balance ratio according to claim 1, characterized in that, The static sensor includes at least one electrode, which is arranged near the outlet of the burner, and the electrode faces the flame or is exposed to the outside, or is covered with a high-temperature-resistant insulating material as a separation layer; The other surface of the electrode is shielded by a metal shielding layer, and an insulating material is added between the metal shielding layer and the electrode.

3. The lean misfire detection method based on electrostatic signal data balance ratio of claim 1, wherein, The threshold is 0.

1.

4. The lean misfire detection method based on electrostatic signal data balance ratio of claim 1, wherein, The thickness and axial length of the electrode are 2 mm and 7 mm respectively, and the electrode is arranged at a distance of 4 mm from the edge of the widest part of the flame.

Images