Real-time detection method of gasification burner

Abstract

The invention discloses a real-time detection method of a gasification burner, which realizes millisecond-level anomaly detection and high-precision real-time monitoring through voiceprint feature extraction and a machine learning model, effectively inhibits environmental noise through a guided wave technology and a noise separation algorithm, has strong anti-interference performance, comprehensively evaluates the state of the gasification burner by combining acoustic features and process parameters, and improves the detection accuracy. The false alarm rate is reduced, multi-dimensional analysis is achieved, jet flow matching abnormity and abrasion are early warned in advance, non-planned shutdown is reduced, and the service life of equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of real-time detection of a gasification burner of a gasifier, and in particular to the technical field of a real-time detection method of a gasification burner. Background Art

[0002] In the pressurized gasification process, the flame stability and structural health of the gasification burner directly affect the gasification efficiency and safety. In existing technologies, detection methods mostly rely on temperature, pressure sensors and optical monitoring methods, which have the following problems: 1. Insufficient real-time performance: Traditional sensors have slow response speeds and are unable to capture transient anomalies, such as explosion pulses, in a timely manner; 2. Poor anti-interference ability: The high temperature, high pressure and multiphase flow environment in the gasifier can easily interfere with optical and electrical signals; 3. Lack of multi-parameter coupling analysis: A single signal is difficult to fully characterize the burner's jet matching, atomization effect, and wear status; Therefore, in order to achieve comprehensive, real-time and high-precision monitoring of gasification burners in water-coal slurry pressurized gasification technology and pulverized coal pressurized gasification technology, it is necessary to study and design a real-time detection method for the flame stability and structural health status of gasification burners. Summary of the Invention

[0003] The purpose of the present invention is to solve the problems existing in the prior art. A real-time detection method for a gasification burner is proposed. Based on acoustic wave sensing and artificial intelligence models, the flame stability and structural health status of the gasification burner are detected in real time. It can solve the problems of detection lag, weak anti-interference ability and insufficient multi-parameter coupling analysis in the prior art, and realize comprehensive, real-time and high-precision monitoring of the operating status of the gasification burner.

[0004] To achieve the above object, the present invention proposes a real-time detection method for a gasification burner, comprising the following steps: S1. Acoustic signal acquisition: several acoustic wave sensors are closely attached to the outer surface of the pipe wall where the gasification burner does not enter the gasifier to collect acoustic signals in real time; S2. Acoustic signal processing and feature extraction: noise separation technology is used to process the collected acoustic signals to extract the following characteristic soundprint components: central oxygen jet characteristic frequency band, outer ring oxygen jet characteristic frequency band, coal slurry atomization characteristic frequency band, gas-liquid-solid premixing characteristic frequency band and explosion characteristic transient pulse; S3. Multi-dimensional feature fusion and model construction: constructing a pre-trained flame state classification model including multi-dimensional acoustic feature vectors and gasification operation characteristic parameters; S4. Status determination and early warning: Based on the stability score and abnormality type code output by the classification model, the flame stability status and structural health status of the gasification burner are determined.

[0005] Preferably, the acoustic signal acquisition in step S1 can be applied to gasification burners of water-coal slurry pressurized gasification technology and gasification burners of pulverized coal pressurized gasification technology.

[0006] Preferably, the acoustic wave sensor in step S1 is any one of a non-contact acoustic wave sensor and a contact acoustic wave sensor, and there is at least one acoustic wave sensor, which is installed at any one position or multiple positions simultaneously on the outer wall surface of the outer epoxy main pipe outside the mounting flange where the gasification burner does not enter the gasification furnace, the outer wall surface of the coal slurry main pipe, the outer wall surface of the central oxygen main pipe and the outer wall surface of the outer ring coal slurry pipe.

[0007] Preferably, the acoustic wave sensor in step S1 is a contact type acoustic wave sensor. There are multiple acoustic wave sensors, and the multiple acoustic wave sensors form a ring array. The acoustic wave sensors are installed simultaneously at three locations: the outer wall surface of the outer epoxy main pipe outside the mounting flange where the gasification burner does not enter the gasification furnace, the outer wall surface of the coal slurry main pipe, and the outer wall surface of the central oxygen main pipe.

[0008] Preferably, the acoustic wave sensor in step S1 is any one of a non-contact acoustic wave sensor and a contact acoustic wave sensor, and there is at least one acoustic wave sensor. The acoustic wave sensor is installed at any one position or multiple positions simultaneously on the outer wall surface of the main oxygen pipe and the outer wall surface of the pulverized coal pipe outside the mounting flange where the gasification burner does not enter the gasification furnace.

[0009] Preferably, the acoustic wave sensor in step S1 is a contact type acoustic wave sensor. There are multiple acoustic wave sensors, and the multiple acoustic wave sensors form a ring array. The acoustic wave sensors are installed simultaneously at three positions: the outer wall surface of the main oxygen pipe outside the mounting flange and the outer wall surface of the pulverized coal pipe where the gasification burner does not enter the gasification furnace.

[0010] Preferably, the noise separation technology processing in step S2 includes: time-frequency domain filtering separation method, blind source separation method, variational mode decomposition separation method and sparse representation separation method based on signal processing; waveguide and waveguide technology separation method, acoustic filter separation method and noise cover and damping layer separation method based on physical structure; deep learning separation method, clustering algorithm separation method and generative adversarial network separation method based on machine learning.

[0011] Preferably, the noise separation technology in step S2 is preferably a waveguide coupled noise separation technology that combines waveguide and waveguide technology with variational mode decomposition separation method.

[0012] Preferably, the multidimensional acoustic feature vector in step S3 includes: energy distribution of each frequency band, time-frequency domain correlation coefficient and pulse time occurrence rate; the gasification operation characteristic parameters include: gasification synthesis gas system pressure, coal slurry load, oxygen pipeline total pressure, gasification burner external epoxy pipe pressure, gasification burner coal slurry pipe pressure, gasification burner coal slurry pressure difference, total oxygen flow rate, central oxygen flow rate ratio, oxygen-coal ratio, synthesis gas components and proportions, methane content and slag mouth pressure difference; the flame state classification model is an artificial intelligence model based on a neural network.

[0013] Preferably, the judgment logic of the stability score and abnormality type code in step S4 includes: ① when the occurrence rate of explosion characteristic transient pulses exceeds a preset threshold, an uneven atomization warning is generated; ② when the energy distribution ratio of the central oxygen jet and the outer ring oxygen jet deviates from the historical benchmark value to a certain extent, a jet matching abnormality alarm is generated; ③ when the time-frequency correlation coefficient of the outer ring oxygen characteristic frequency band continues to decrease and exceeds a preset threshold, a gasification burner wear warning is generated.

[0014] The beneficial effects of the present invention are as follows: The present invention realizes millisecond-level anomaly detection and high-precision real-time monitoring through voiceprint feature extraction and machine learning models. The waveguide technology and noise separation algorithm effectively suppress environmental noise and have strong anti-interference ability. It combines acoustic characteristics and process parameters to comprehensively evaluate the status of the gasification burner, reduce the false alarm rate, perform multi-dimensional analysis, and provide early warning of jet matching anomalies and wear, thereby reducing unplanned downtime and extending equipment life.

[0015] The features and advantages of the present invention will be described in detail through embodiments with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a structural schematic diagram of a real-time detection method for a gasification burner of the present invention applied to detection of a gasification burner in a water-coal slurry pressurized gasification technology; Figure 2 This is a structural schematic diagram of a real-time detection method for a gasification burner of the present invention applied to detection of a gasification burner in a pulverized coal pressurized gasification technology; Figure 3 It is a flow chart of a real-time detection method for a gasification burner according to the present invention.

[0017] In the figure: 1—central oxygen pipe of water-coal slurry gasification burner; 2—two-channel sound sensor; 3—water-coal slurry pipe of water-coal slurry gasification burner; 4—three-channel sound sensor array; 5—external epoxy pipe of water-coal slurry gasification burner; 6—four-channel sound sensor array; 7—main oxygen pipe of pulverized coal process burner; 8—pulverized coal pipe of pulverized coal process burner. DETAILED DESCRIPTION

[0018] Example 1 See Figure 1 and Figure 3 The present invention provides a real-time detection method for a gasification burner, which is applied to the detection of a gasification burner in a water-coal slurry pressurized gasification technology, and comprises the following steps: S1. Install contact sensors in a circular array on the outer wall surfaces of the outer epoxy main pipe, coal slurry main pipe and central oxygen main pipe; S2. Collect acoustic signals, use variational mode decomposition to separate background noise, and extract the energy distribution of each characteristic frequency band; S3, input the acoustic characteristics, real-time oxygen-coal ratio (0.8-1.2), and coal slurry pressure difference (2-5 MPa) into the neural network model; S4. The model output stability score is 75 points (out of 100), and the "decrease in time-frequency correlation of external epoxy" warning is triggered, prompting the inspection of the wear of the gasifier nozzle head.

[0019] Example 2 See Figure 1 and Figure 2 The present invention provides a real-time detection method for a gasification burner, which is applied to the detection of a gasification burner in a pulverized coal pressurized gasification technology, and comprises the following steps: S1. Install a non-contact sensor on the outer wall of the central oxygen main pipe; S2, collect acoustic signals, process the signals using blind source separation method, and extract explosion transient pulse characteristics; S3. When the pulse rate exceeds 5 times / minute, an abnormal atomization warning is triggered and the pulverized coal particle size distribution is adjusted to return to normal.

[0020] The present invention realizes millisecond-level anomaly detection and high-precision real-time monitoring through voiceprint feature extraction and machine learning models. The waveguide technology and noise separation algorithm effectively suppress environmental noise and have strong anti-interference ability. It combines acoustic characteristics and process parameters to comprehensively evaluate the status of the gasification burner, reduce the false alarm rate, perform multi-dimensional analysis, and provide early warning of jet matching anomalies and wear, thereby reducing unplanned downtime and extending equipment life.

[0021] The present invention can significantly improve the operational safety and economy of the gasifier and has broad market potential.

[0022] The above embodiments are intended to illustrate the present invention, not to limit the present invention. Any solution that is a simple transformation of the present invention falls within the protection scope of the present invention.

Claims

1. A real-time detection method for a gasification burner, characterized in that: The following steps are involved: S1. Acoustic signal acquisition: several acoustic wave sensors are closely attached to the outer surface of the pipe wall where the gasification burner does not enter the gasifier to collect acoustic signals in real time; S2. Acoustic signal processing and feature extraction: noise separation technology is used to process the collected acoustic signals to extract the following characteristic soundprint components: central oxygen jet characteristic frequency band, outer ring oxygen jet characteristic frequency band, coal slurry atomization characteristic frequency band, gas-liquid-solid premixing characteristic frequency band and explosion characteristic transient pulse; S3. Multi-dimensional feature fusion and model construction: constructing a pre-trained flame state classification model including multi-dimensional acoustic feature vectors and gasification operation characteristic parameters; S4. Status determination and early warning: Based on the stability score and abnormality type code output by the classification model, the flame stability status and structural health status of the gasification burner are determined.

2. The real-time detection method for a gasification burner according to claim 1, characterized in that: The acoustic signal acquisition in step S1 can be applied to gasification burners of water-coal slurry pressurized gasification technology and gasification burners of pulverized coal pressurized gasification technology.

3. The real-time detection method for a gasification burner according to claim 1, characterized in that: The acoustic wave sensor in step S1 is any one of a non-contact acoustic wave sensor and a contact acoustic wave sensor, and there is at least one acoustic wave sensor. The acoustic wave sensor is installed at any one position or multiple positions simultaneously on the outer wall surface of the outer epoxy main pipe outside the mounting flange where the gasification burner does not enter the gasification furnace, the outer wall surface of the coal slurry main pipe, the outer wall surface of the central oxygen main pipe, and the outer wall surface of the outer ring coal slurry pipe.

4. A real-time detection method for a gasification burner according to claim 3, characterized in that: The acoustic wave sensor in step S1 is preferably a contact acoustic wave sensor. There are multiple acoustic wave sensors, and the multiple acoustic wave sensors form a ring array. The acoustic sensors are installed simultaneously at three locations: the outer wall surface of the outer epoxy main pipe outside the mounting flange where the gasification burner does not enter the gasification furnace, the outer wall surface of the coal slurry main pipe, and the outer wall surface of the central oxygen main pipe.

5. The real-time detection method for a burner in an entrained flow pressurized gasification process according to claim 1, characterized in that: The acoustic wave sensor in step S1 is any one of a non-contact acoustic wave sensor and a contact acoustic wave sensor. There is at least one acoustic wave sensor, and the acoustic wave sensor is installed at any one position or multiple positions simultaneously on the outer wall surface of the main oxygen pipe and the outer wall surface of the pulverized coal pipe outside the mounting flange where the gasification burner does not enter the gasification furnace.

6. The real-time detection method for a burner in an entrained flow pressurized gasification process according to claim 5, characterized in that: The acoustic wave sensor in step S1 is preferably a contact acoustic wave sensor. There are multiple acoustic wave sensors, and the multiple acoustic wave sensors form a ring array. The acoustic wave sensors are installed simultaneously at three locations on the outer wall surface of the main oxygen pipe outside the mounting flange and the outer wall surface of the pulverized coal pipe where the gasification burner does not enter the gasifier.

7. The real-time detection method for a gasification burner according to claim 1, characterized in that: The noise separation technology processing in step S2 includes: time-frequency domain filtering separation method, blind source separation method, variational mode decomposition separation method and sparse representation separation method based on signal processing; waveguide and waveguide technology separation method, acoustic filter separation method and noise cover and damping layer separation method based on physical structure; deep learning separation method, clustering algorithm separation method and generative adversarial network separation method based on machine learning.

8. The real-time detection method for a gasification burner according to claim 7, characterized in that: The noise separation technology in step S2 is preferably a waveguide coupled noise separation technology that combines waveguide and waveguide technology with variational mode decomposition separation method.

9. The real-time detection method for a gasification burner according to claim 1, characterized in that: The multidimensional acoustic feature vector in step S3 includes: energy distribution of each frequency band, time-frequency domain correlation coefficient and pulse time occurrence rate; the gasification operation characteristic parameters include: gasification synthesis gas system pressure, coal slurry load, oxygen pipeline total pressure, gasification burner external epoxy pipe pressure, gasification burner coal slurry pipe pressure, gasification burner coal slurry pressure difference, total oxygen flow rate, central oxygen flow rate ratio, oxygen-coal ratio, synthesis gas components and proportions, methane content and slag mouth pressure difference; the flame state classification model is an artificial intelligence model based on a neural network.

10. The real-time detection method for a gasification burner according to claim 1, characterized in that: The judgment logic of the stability score and abnormality type code in step S4 includes: ① when the occurrence rate of explosion characteristic transient pulses exceeds a preset threshold, an uneven atomization warning is generated; ② when the energy distribution ratio of the central oxygen jet and the outer ring oxygen jet deviates from the historical benchmark value to a certain extent, a jet matching abnormality alarm is generated; ③ when the time-frequency correlation coefficient of the outer ring oxygen characteristic frequency band continues to decrease and exceeds a preset threshold, a gasification burner wear warning is generated.

Images