WO2026007302A1 - 一种基于同步压缩变换的振荡燃烧故障检测方法 - Google Patents

一种基于同步压缩变换的振荡燃烧故障检测方法

Info

Publication number
WO2026007302A1
WO2026007302A1 PCT/CN2024/132262 CN2024132262W WO2026007302A1 WO 2026007302 A1 WO2026007302 A1 WO 2026007302A1 CN 2024132262 W CN2024132262 W CN 2024132262W WO 2026007302 A1 WO2026007302 A1 WO 2026007302A1
Authority
WO
WIPO (PCT)
Prior art keywords
flame
time
sub
combustion
frequency
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/CN2024/132262
Other languages
English (en)
French (fr)
Inventor
肖俊峰
于倩倩
王玮
高松
夏家兴
李丹
王峰
李晓丰
李乐
郭菡
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Thermal Power Research Institute Co Ltd
Original Assignee
Xian Thermal Power Research Institute Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xian Thermal Power Research Institute Co Ltd filed Critical Xian Thermal Power Research Institute Co Ltd
Publication of WO2026007302A1 publication Critical patent/WO2026007302A1/zh
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/10Pre-processing; Data cleansing
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10Complex mathematical operations
    • G06F17/11Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10Complex mathematical operations
    • G06F17/14Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/21Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/24Classification techniques
    • G06F18/243Classification techniques relating to the number of classes
    • G06F18/2433Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0004Industrial image inspection
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/90Determination of colour characteristics

Definitions

  • This invention relates to the fields of signal processing and fault diagnosis technology, and in particular to an oscillating combustion fault detection method based on synchronous compression transformation.
  • Gas turbines as highly efficient energy conversion devices, are widely used in power generation, ship propulsion, and mechanical power.
  • the operational stability of their combustors has a decisive impact on the performance and safety of the entire system.
  • Gas turbine combustors typically employ a lean premixed swirl combustion method. While this method reduces combustion temperature and NOx emissions, it is highly susceptible to oscillating combustion faults. These faults severely affect the operating and combustion efficiency of the gas turbine and its combined cycle, and may also cause fluctuations in critical parameters such as combustor temperature and NOx concentration, posing a threat to equipment and personnel safety. Combustion instability phenomena, such as oscillating combustion, may occur during gas turbine operation, leading to decreased equipment performance and potentially even equipment damage. Therefore, rapid detection and diagnosis of oscillating combustion faults in gas turbines are particularly important.
  • this invention provides an oscillating combustion fault detection method based on synchronous compression transformation, which can reduce the time cost of fault detection, improve the accuracy of fault detection, and reduce the false alarm rate.
  • an oscillating combustion fault detection method based on synchronous compression transform comprising: acquiring flame chemiluminescence signals; performing adaptive filtering and noise reduction processing on the acquired flame chemiluminescence signals; extracting first-order mode time coefficients from flame images using an improved intrinsic orthogonal decomposition method; obtaining comprehensive combustion feature information using higher-order modes; developing a real-time synchronous compression transform algorithm to perform time-frequency analysis on the data and update energy distribution in real time to respond to minute changes in combustion state; combining sensor data to perform multi-parameter joint analysis to obtain combustion state assessment; acquiring flame images from different angles using multiple high-speed cameras; reconstructing the three-dimensional flame morphology using computer vision technology to provide fault detection information; constructing a fault prediction model to automatically complete the entire fault detection process and achieve early warning of combustion faults.
  • the adaptive filtering noise reduction process includes, when performing oscillating combustion fault detection, distinguishing between noise and actual combustion oscillation signals, selecting an adaptive filter based on the steepest descent method, and combining noise estimation and signal enhancement in the filter.
  • the adaptive filter update rule is as follows:
  • w(n) represents the filter weight vector
  • n represents the discrete time step.
  • represent a constant
  • x(n) represent the input flame signal at time n
  • e(n) represent the error signal at time n
  • represent the forgetting factor.
  • the improved intrinsic orthogonal decomposition method includes performing POD analysis on the flame image set after filter processing to obtain intrinsic modes and intrinsic values, and selecting the M intrinsic modes with the highest energy according to the magnitude of the intrinsic values to form a new mode set, representing the most significant spatial features in the data.
  • the time coefficient represents the change of the projection of the flame image onto each intrinsic mode over time.
  • the calculation formula is as follows:
  • the calculated time coefficients are stored in a set:
  • the development of the real-time synchronous compression transform algorithm includes: rearranging the coefficients of the time-frequency transform through a synchronous compression operator, moving the time-frequency coefficients of the signal at any point in the time-frequency plane to the center of energy position, enhancing the energy concentration of the instantaneous frequency; applying a window function on the basis of Fourier transform to divide the signal time domain; obtaining the frequency distribution under different time windows through window function sliding; and arranging these short-time spectra in time sequence to describe the time-varying law of the signal frequency components.
  • G(t,f) represents the result of the Fourier transform
  • s( ⁇ ) represents the original signal
  • represents the time variable of the original signal
  • g( ⁇ -t) represents the result of the Fourier transform
  • t represents the time variable
  • f represents the frequency variable
  • i represents the imaginary unit
  • ⁇ (t, ⁇ ) represents the instantaneous frequency
  • represents the frequency variable
  • arg(G(t,f)) represents the argument
  • Re represents taking the real part. This represents the partial derivative with respect to time t;
  • T s (t, ⁇ ) ⁇ R G (t, f) ⁇ ( ⁇ - ⁇ (t, ⁇ )) d ⁇
  • T ⁇ sub> s ⁇ /sub> (t, ⁇ ) represents the instantaneous frequency signal
  • ⁇ ( ⁇ - ⁇ (t, ⁇ )) represents the Dirac function
  • represents the instantaneous frequency variable
  • the time-frequency coefficients are redistributed and rearranged in the frequency direction
  • the synchronous compression transformation process at time t is the redistribution of frequency points and the process that goes through all times.
  • the multi-parameter joint analysis includes analyzing the data collected by the sensor to understand its distribution and characteristics, adjusting the parameters through optimization algorithms, and finding the optimal parameter combination after multiple iterations.
  • the combined parameters are substituted into the multi-parameter joint analysis formula to calculate the weighted contribution of each sensor.
  • the weights take into account the different sensitivities of the sensors to angle, velocity, and light intensity, and a normalization function is applied to ensure the comparability of data from different sensors.
  • the specific formula is as follows:
  • G( ⁇ ,v,I) represents the master function for multi-parameter joint analysis
  • represents the flame tilt angle
  • v represents the flame combustion speed
  • I represents the flame radiation intensity
  • n represents the number of sensors
  • ⁇ i represents the effective measurement domain of the i-th sensor
  • f ⁇ sub> i ⁇ /sub>( ⁇ ⁇ sub>i ⁇ /sub>, v ⁇ sub> i ⁇ /sub>, I ⁇ sub>i ⁇ /sub>) represents the weighted contribution
  • ⁇ sub> i ⁇ /sub> represents the angle parameter of the i-th sensor
  • ⁇ sub> i ⁇ /sub> represents the angle value measured by the i-th sensor
  • ⁇ ⁇ sub>0 ⁇ /sub> i represents the reference angle value of the i-th sensor
  • ⁇ sub>i ⁇ /sub> represents the adjustment coefficient for the velocity parameter of the i-th sensor
  • ⁇ ⁇ sub>i ⁇ /sub> represents the adjustment coefficient for the light intensity parameter of the i
  • the reconstruction of the three-dimensional flame morphology includes reconstructing the three-dimensional flame morphology using a computer vision algorithm by combining the results of multi-parameter joint analysis.
  • x, y, z represent the coordinate axes in three-dimensional space
  • ht represents the flame's change function over time
  • represents the spatial domain of the integration
  • represents the phase shift
  • represents the attenuation coefficient
  • the three-dimensional flame model gives the complete shape of the flame in three-dimensional space, including its changes over time.
  • the fault prediction model includes comparing the calculated flame state and flame shape with a normal behavior threshold. If the calculation result exceeds a preset threshold range, it is considered that there is abnormal behavior.
  • the solution is to adjust the oxygen/fuel ratio of the burner and check and clean the carbon deposits and sediment inside the burner.
  • the solutions are to check the fuel supply system to maintain a stable fuel flow; check whether the burner nozzles and combustion chamber are blocked; and check whether the internal structure is damaged.
  • the solution is to adjust the camera parameters to improve image quality, check whether the camera lens is contaminated or damaged, and check the brightness and color characteristics of the flame.
  • fault detection information The detected abnormal behaviors and their causes are summarized as fault detection information and recorded in the database for subsequent analysis and reporting. Fault detection reports are generated regularly to monitor the health and performance of the combustion process.
  • Another objective of this invention is to provide an oscillating combustion fault detection system based on synchronous compression transformation, which can improve the fault detection accuracy and diagnostic speed through an oscillating combustion fault detection algorithm.
  • the oscillating combustion fault detection system based on synchronous compression transformation according to the present invention, it includes: a data acquisition module, an intrinsic orthogonal decomposition method module, a synchronous compression transformation module, a multi-parameter joint analysis module, a three-dimensional flame morphology reconstruction module, and a fault detection module;
  • the data acquisition module uses a high-sensitivity flame chemiluminescence sensor to accurately capture the light signals generated by chemical reactions in the flame.
  • the sensor array is arranged at different positions of the burner to obtain the overall chemiluminescence characteristics of the flame.
  • the intrinsic orthogonal decomposition method module accurately predicts faults by analyzing and identifying change patterns during the combustion process.
  • the synchronous compression transformation module performs time-frequency analysis on sensor data through a real-time SCT algorithm, captures instantaneous changes during the combustion process, reflects subtle changes in the combustion state, and improves the sensitivity of fault detection.
  • the multi-parameter joint analysis module uses a multi-parameter joint analysis algorithm to evaluate the flame state
  • the three-dimensional flame shape reconstruction module uses computer vision technology to reconstruct the three-dimensional flame shape from flame images acquired from different angles.
  • the fault detection module analyzes the results of multi-parameter joint analysis and three-dimensional flame morphology reconstruction to detect abnormal flame behavior.
  • a computer device includes a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of an oscillating combustion fault detection method based on synchronous compression transformation.
  • a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of an oscillating combustion fault detection method based on synchronous compression transformation.
  • the invention utilizes high-speed photography technology to acquire flame images, which avoids the data lag problem caused by traditional pressure monitoring methods, enabling non-destructive and rapid detection of oscillating combustion faults; intrinsic orthogonal decomposition can reduce the dimensionality of flame images and extract features, retaining rich information of the flame images, reflecting the overall pulsation characteristics of the flame, and improving detection speed and accuracy; synchronous compression transformation performs energy rearrangement and instantaneous concentration through synchronous compression operators, making up for the shortcomings of low frequency resolution in traditional time-frequency analysis methods, and realizing accurate diagnosis and rapid detection of oscillating combustion faults.
  • Figure 1 is a schematic flowchart of an oscillating combustion fault detection method based on synchronous compression transformation provided by an embodiment of the present invention.
  • Figure 2 shows the detection results of an oscillating combustion fault detection method based on synchronous compression transformation provided by an embodiment of the present invention.
  • one embodiment or “embodiment” as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention.
  • the phrase "in one embodiment” appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.
  • connection Unless otherwise explicitly specified and limited, the terms “installation,” “connection,” and “joining” in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
  • this embodiment provides a method for detecting oscillating combustion faults based on synchronous compression transformation, comprising:
  • S1 Acquire flame chemiluminescence signals and perform adaptive filtering and noise reduction processing on the acquired flame chemiluminescence signals.
  • the adaptive filtering noise reduction process includes, when performing oscillating combustion fault detection, distinguishing between noise and actual combustion oscillation signals, selecting an adaptive filter based on the steepest descent method, and combining noise estimation and signal enhancement with the filter.
  • w(n) represents the filter weight vector
  • n represents the discrete time step.
  • represent a constant
  • x(n) represent the input flame signal at time n
  • e(n) represent the error signal at time n
  • represent the forgetting factor.
  • S2 The improved intrinsic orthogonal decomposition method is used to extract the first-order modal time coefficients from the flame image, and the higher-order modes are used to obtain comprehensive combustion feature information.
  • the improved intrinsic orthogonal decomposition method includes performing POD analysis on the flame image set after filter processing to obtain intrinsic modes and eigenvalues, and selecting the M eigenmodes with the highest energy according to the magnitude of the eigenvalues to form a new mode set, representing the most significant spatial features in the data.
  • the time coefficient represents the change of the projection of the flame image onto each intrinsic mode over time.
  • the calculation formula is as follows:
  • the calculated time coefficients are stored in a set:
  • S3 Develop a real-time synchronous compression transformation algorithm to perform time-frequency analysis on the data and update the energy allocation in real time to respond to minute changes in the combustion state.
  • the development of the real-time synchronous compression transform algorithm includes rearranging the coefficients of the time-frequency transform using a synchronous compression operator, moving the time-frequency coefficients of the signal at any point in the time-frequency plane to the center of energy position, enhancing the energy concentration of the instantaneous frequency, applying a window function on the basis of the Fourier transform to divide the signal time domain, obtaining the frequency distribution under different time windows by sliding the window function, and then arranging these short-time spectra in time sequence to describe the time-varying law of the signal frequency components;
  • G(t,f) represents the result of the Fourier transform
  • s( ⁇ ) represents the original signal
  • represents the time variable of the original signal
  • g( ⁇ -t) represents the result of the Fourier transform
  • t represents the time variable
  • f represents the frequency variable
  • i represents the imaginary unit
  • ⁇ (t, ⁇ ) represents the instantaneous frequency
  • represents the frequency variable
  • arg(G(t,f)) represents the argument
  • Re represents taking the real part. This represents the partial derivative with respect to time t;
  • T ⁇ sub> s ⁇ /sub> (t, ⁇ ) represents the instantaneous frequency signal
  • ⁇ ( ⁇ - ⁇ (t, ⁇ )) represents the Dirac function
  • represents the instantaneous frequency variable
  • the time-frequency coefficients are redistributed and rearranged in the frequency direction
  • the synchronous compression transformation process at time t is the redistribution of frequency points and the process that goes through all times.
  • S4 Combine sensor data to perform multi-parameter joint analysis to obtain combustion status assessment. Use multiple high-speed cameras to acquire flame images from different angles, and reconstruct the three-dimensional flame morphology through computer vision technology to provide fault detection information.
  • the multi-parameter joint analysis includes analyzing the data collected by the sensors to understand the distribution and characteristics, adjusting the parameters through optimization algorithms, and finding the optimal parameter combination through multiple iterations.
  • the combined parameters are substituted into the multi-parameter joint analysis formula to calculate the weighted contribution of each sensor.
  • the weights take into account the different sensitivities of the sensors to angle, velocity, and light intensity, and a normalization function is applied to ensure the comparability of data from different sensors.
  • the specific formula is as follows:
  • G( ⁇ ,v,I) represents the master function for multi-parameter joint analysis
  • represents the flame tilt angle
  • v represents the flame combustion speed
  • I represents the flame radiation intensity
  • n represents the number of sensors
  • ⁇ i represents the effective measurement domain of the i-th sensor
  • f ⁇ sub> i ⁇ /sub>( ⁇ ⁇ sub>i ⁇ /sub>, v ⁇ sub> i ⁇ /sub>, I ⁇ sub>i ⁇ /sub>) represents the weighted contribution
  • ⁇ sub> i ⁇ /sub> represents the angle parameter of the i-th sensor
  • ⁇ sub> i ⁇ /sub> represents the angle value measured by the i-th sensor
  • ⁇ ⁇ sub>0 ⁇ /sub> i represents the reference angle value of the i-th sensor
  • ⁇ sub>i ⁇ /sub> represents the adjustment coefficient for the velocity parameter of the i-th sensor
  • ⁇ ⁇ sub>i ⁇ /sub> represents the adjustment coefficient for the light intensity parameter of the i
  • the reconstruction of the three-dimensional flame morphology includes reconstructing the three-dimensional flame morphology using computer vision algorithms by combining the results of multi-parameter joint analysis.
  • x, y, z represent the coordinate axes in three-dimensional space
  • ht represents the flame's change function over time
  • represents the spatial domain of the integration
  • represents the phase shift
  • represents the attenuation coefficient
  • the three-dimensional flame model gives the complete shape of the flame in three-dimensional space, including its changes over time.
  • the fault prediction model includes comparing the calculated flame state and flame shape with a normal behavior threshold. If the calculation result exceeds a preset threshold range, it is considered that there is abnormal behavior.
  • the solution is to adjust the oxygen/fuel ratio of the burner and check and clean the carbon deposits and sediment inside the burner.
  • the solutions are to check the fuel supply system to maintain a stable fuel flow; check whether the burner nozzles and combustion chamber are blocked; and check whether the internal structure is damaged.
  • the solution is to adjust the camera parameters to improve image quality, check whether the camera lens is contaminated or damaged, and check the brightness and color characteristics of the flame.
  • fault detection information The detected abnormal behaviors and their causes are summarized as fault detection information and recorded in the database for subsequent analysis and reporting. Fault detection reports are generated regularly to monitor the health and performance of the combustion process.
  • an embodiment of the present invention which provides an oscillating combustion fault detection method based on synchronous compression transformation.
  • scientific demonstration is carried out through experiments.
  • Figure 2 shows two verification conditions involved in the embodiment of the present invention.
  • the pressure signals of the experimental platform used for verification under oscillation and stable conditions were collected and subjected to STFT analysis.
  • Figure 3 shows the synchronous compression transformation analysis results of the pressure signals under the two conditions.
  • the synchronous compression transformation concentrated the energy of the signal under test, superimposing the spectrum within the pseudo-frequency range, concentrating the energy on the actual instantaneous frequency, improving time-frequency convergence, and greatly improving the frequency resolution.
  • the synchronous compression transformation analysis results of the flame CH* light intensity signal under the two conditions demonstrate that the frequency domain resolution of the time-frequency analysis results is greatly improved through synchronous compression transformation analysis.
  • the flame image was subjected to intrinsic orthogonal decomposition for data dimensionality reduction and feature extraction to obtain the first-order modal time coefficient. This coefficient was then used as the data input for STFT (Sequential Transform-Through) fault diagnosis.
  • the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium.
  • the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution can be embodied in the form of a software product.
  • This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.
  • the aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
  • Computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM).
  • electrical connections electronic devices
  • portable computer disk drives magnetic devices
  • RAM random access memory
  • ROM read-only memory
  • EPROM or flash memory erasable and editable read-only memory
  • CDROM portable optical disc read-only memory
  • computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
  • an embodiment of the present invention provides an oscillating combustion fault detection system based on synchronous compression transformation, characterized in that it includes a data acquisition module, an intrinsic orthogonal decomposition method module, a synchronous compression transformation module, a multi-parameter joint analysis module, a three-dimensional flame morphology reconstruction module, and a fault detection module.
  • the data acquisition module uses a high-sensitivity flame chemiluminescence sensor to accurately capture the light signals generated by chemical reactions in the flame.
  • the sensor array is arranged at different positions in the burner to obtain the overall chemiluminescence characteristics of the flame.
  • the intrinsic orthogonal decomposition method module accurately predicts faults by analyzing and identifying change patterns during the combustion process.
  • the synchronous compression transformation module performs time-frequency analysis on sensor data through the real-time SCT algorithm, captures instantaneous changes in the combustion process, reflects subtle changes in the combustion state, and improves the sensitivity of fault detection.
  • the multi-parameter joint analysis module uses a multi-parameter joint analysis algorithm to evaluate the flame state.
  • the 3D flame shape reconstruction module uses computer vision technology to reconstruct the 3D flame shape from flame images acquired from different angles.
  • the fault detection module analyzes the results of multi-parameter joint analysis and three-dimensional flame morphology reconstruction to detect abnormal flame behavior.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Algebra (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Photometry And Measurement Of Optical Pulse Characteristics (AREA)

Abstract

本发明涉及信号处理和故障诊断技术领域,本发明所述方法包括,采集火焰化学发光信号,并进行自适应滤波降噪处理;利用改进的本征正交分解方法和高阶模态,获得全面的燃烧特征信息;开发实时同步压缩变换算法对数据进行时频分析,并实时更新能量分配;结合传感器数据,进行多参数联合分析,获得燃烧状态评估,利用多个高速相机从不同角度获取火焰图像,通过计算机视觉技术重建三维火焰形态,提供故障检测信息;构建故障预测模型,实现燃烧故障的早期预警。本方法实现对振荡燃烧故障的无损、快速检测,保留火焰图像的丰富信息,反映火焰的整体脉动特点,提高检测速度和准确率,弥补了传统时频分析方法频率分辨率较低的缺点。

Description

一种基于同步压缩变换的振荡燃烧故障检测方法 技术领域
本发明涉及信号处理和故障诊断技术领域,尤其涉及一种基于同步压缩变换的振荡燃烧故障检测方法。
背景技术
燃气轮机作为一种高效的能源转换设备,被广泛应用于发电、船舶推进和机械动力等领域。然而,其燃烧室的运行稳定性对整个系统的性能和安全具有决定性影响。燃气轮机燃烧室通常采用贫预混旋流燃烧方式,这种方式虽能降低燃烧温度及NOx排放,但极易诱发振荡燃烧故障。这种故障会严重影响燃气轮机及其联合循环的运行效率和燃烧效率,同时可能导致燃烧室温度、NOx浓度等重要参数的波动,对设备和人员安全构成威胁。燃气轮机在运行过程中可能会出现燃烧不稳定现象,如振荡燃烧,这会导致设备性能下降,甚至可能造成设备损坏。因此,对燃气轮机振荡燃烧故障的快速检测和诊断显得尤为重要。
传统的振荡燃烧故障监测方法主要依赖于压力传感器对燃烧室时域压力信号进行频域分析。然而,这种方法存在一些明显的不足。首先,由于传感器的安装位置通常远离火焰,实际燃烧室压力与传感器测量压力之间存在误差,导致数据监测存在偏差和滞后现象。其次,由于传感器响应速度的限制,故障现象的诊断识别不及时,往往在发生严重损失之后才能发现故障,这对于实际生产是极为不利的。
现有的燃气轮机振荡燃烧故障检测技术通常采用声学检测、压力波动分析等方法。这些方法在一定程度上能够检测到燃烧振荡,但存在以下不足:
灵敏度不高:由于声学检测和压力波动分析等方法受到环境噪声和机械振动的影响,对于微弱的燃烧振荡信号检测效果不佳,容易出现误报或漏报。
定位困难:现有的检测技术难以精确定位燃烧振荡的具体位置,无法为故障处理提供准确的信息。
时效性差:传统的检测方法需要较长时间的信号采集和处理,无法实现实时监测和快速诊断。
发明内容
鉴于上述现有存在的问题,提出了本发明。
因此,本发明提供了一种基于同步压缩变换的振荡燃烧故障检测方法,能够降低了故障检测的时间成本,提高故障检测的准确率,降低虚警率。
为解决上述技术问题,本发明提供如下技术方案,一种基于同步压缩变换的振荡燃烧故障检测方法,包括:采集火焰化学发光信号,对采集到的火焰化学发光信号进行自适应滤波降噪处理;利用改进的本征正交分解方法对火焰图像提取1阶模态时间系数,利用高阶模态,获得全面的燃烧特征信息;开发实时同步压缩变换算法对数据进行时频分析,并实时更新能量分配,响应燃烧状态的微小变化;结合传感器数据,进行多参数联合分析,获得燃烧状态评估,利用多个高速相机从不同角度获取火焰图像,通过计算机视觉技术重建三维火焰形态,提供故障检测信息;构建故障预测模型,自动完成整个故障检测流程,实现燃烧故障的早期预警。
作为本发明所述的一种基于同步压缩变换的振荡燃烧故障检测方法的一种优选方案,其中:所述自适应滤波降噪处理包括,进行振荡燃烧故障检测时,区分噪声和实际的燃烧振荡信号,选择基于最陡下降法的自适应滤波器,滤波器结合噪声估计和信号增强;
自适应滤波器更新规则如下:
其中,w(n)表示滤波器权重向量,n表示离散时间步长,表示第n时刻的噪声功率估计,∈表示常数,x(n)表示第n时刻的输入的火焰信号,e(n)表示第n时刻的误差信号,α表示遗忘因子。
作为本发明所述的一种基于同步压缩变换的振荡燃烧故障检测方法的一种优选方案,其中:所述改进的本征正交分解方法包括,对滤波器处理后的火焰图像集合进行POD分析,得到本征模态和本征值,根据本征值的大小,选择能量最高的M个本征模态,形成一个新的模态集合,代表数据中最显著的空间特征;
对于每个选定的模态计算时间系数,时间系数表示火焰图像在每个本征模态上的投影随时间的变化,计算公式如下:
其中,ci(m)表示时间系数,I(m)表示第m个火焰图像帧,表示图像集合的均值,φi(x,y)表示火焰图像的本征模态,i是模态的索引;
将计算得到的时间系数存储在集合中:
其中,表示包含高阶模态时间系数的集合,集合揭示燃烧过程中的细微波动以及早期故障迹象。
作为本发明所述的一种基于同步压缩变换的振荡燃烧故障检测方法的一种优选方案,其中:所述开发实时同步压缩变换算法包括,通过同步压缩算子对时频变换的系数进行重排,将信号在时频平面任一点处的时频系数移到能量的重心位置,增强瞬时频率的能量集中程度,在傅里叶变换的基础上套用一个窗函数,将信号时域进行分割,通过窗函数滑动得到不同时刻窗口下的频率分布情况,再将这些短时频谱依时序进行排列,描述信号频率成分的时变规律;
对于确定的时变信号短时傅里叶变换窗函数为g(t),经STFT变换后有:
G(t,f)=∫Rs(τ)g(τ-t)e-i2πf(τ-t)
其中,G(t,f)表示傅里叶变换的结果,s(τ)表示原始信号,τ表示原始信号的时间变量,g(τ-t)表示,t表示时间变量,f表示频率变量,i表示虚数单位;
针对时变谐波信号以及窗函数信号做傅里叶变换,得到STFT时频表示中瞬时频率的估计结果,表示为:
其中,θ(t,ω)表示瞬时频率,ω表示频率变量,arg(G(t,f))表示辐角,Re表示取实部,表示对时间t的偏导数;
利用上述瞬时频率估计结果收集具有相同频率的STFT系数,根据狄拉克函数性质,给出同步压缩算子函数∫Rδ(η-ω0(t,ω)),得到同步压缩变换,表示为:
Ts(t,η)=∫RG(t,f)δ(η-θ(t,ω))dω
其中,Ts(t,η)表示瞬时频率信号,δ(η-θ(t,ω))表示狄拉克函数,η表示瞬时频率变量,在频率方向上对时频系数进行重新分配排列,在时间t时刻的同步压缩变换过程,即频率点重新分配并历经所有时刻。
作为本发明所述的一种基于同步压缩变换的振荡燃烧故障检测方法的一种优选方案,其中:所述多参数联合分析包括,对传感器收集的数据进行分析,了解分布和特性,通过优化算法调整参数,经过多次迭代,找到最佳的参数组合;
将参数组合代入多参数联合分析公式计算每个传感器的加权贡献,权重考虑传感器对角度、速度和光强的不同敏感度,并应用归一化函数确保不同传感器数据之间的可比性,具体公式为:
其中,G(θ,v,I)表示多参数联合分析主函数,θ表示火焰倾斜角度,v表示火焰燃烧速度,I表示火焰辐射强度,n表示传感器数量,Ωi表示第i个传感器的有效测量域,fii,vi,Ii)表示加权贡献,αi表示第i个传感器角度参数,θi表示第i个传感器测量的角度值,θ0i表示第i个传感器的参考角度值,βi表示第i个传感器速度参数的调整系数,γi表示第i个传感器光强参数的调整系数,分别表示第i个传感器角度测量值的均值和标准差,分别表示第i个传感器速度测量值的均值和标准差。
作为本发明所述的一种基于同步压缩变换的振荡燃烧故障检测方法的一种优选方案,其中:所述重建三维火焰形态包括,结合多参数联合分析的结果,利用计算机视觉算法重建火焰的三维形态;
通过时间相关函数考虑火焰随时间的变化,得到动态的三维火焰模型:
Fx,y,z,t)=∫ΩGθ,v,I)·htdt
h(t)=sin(ωt+φ)·e-λt
其中,x,y,z表示三维空间的坐标轴,ht表示火焰随时间的变化函数,Ω表示积分的空间域,φ表示相位偏移量,λ表示衰减系数;通过三维火焰模型给出火焰在三维空间中的完整形态,包括随时间的变化。
作为本发明所述的一种基于同步压缩变换的振荡燃烧故障检测方法的一种优选方案,其中:所述故障预测模型包括,比较计算出的火焰状态和火焰形态与正常行为阈值,计算结果超出预设的阈值范围,则认为存在异常行为;
当火焰温度低于1000K或高于2500K,则认定异常行为是燃烧不充分、燃烧器堵塞,解决措施为调整燃烧器的氧气/燃料比,检查并清理燃烧器内部的积碳和沉积物;
当火焰振荡频率低于0.5Hz或高于2Hz,以及火焰形状低于0.5或高于1.5,都认定异常行为是燃烧器内部结构损坏、燃料供应不稳定,解决措施为检查燃料供应系统,保持燃料流量稳定;检查燃烧器的喷嘴和燃烧室是否堵塞、检查内部结构是否损坏;
当火焰图像的对比度低于0.6或高于1.5,则认定异常行为是相机设置不当或火焰本身特性引起,解决措施为调整相机的参数改善图像质量,检查相机的镜头是否污染损坏;检查火焰亮度、颜色特性;
将检测到的异常行为和原因汇总为故障检测信息记录在数据库中,用于后续分析和报告,定期生成故障检测报告,监控燃烧过程的健康状况和性能。
本发明的另一个目的是提供一种基于同步压缩变换的振荡燃烧故障检测系统,其能通过振荡燃烧故障检测算法提高的故障检测准确率和诊断速度。
作为本发明所述的一种基于同步压缩变换的振荡燃烧故障检测系统的一种优选方案,其中:包括数据采集模块、本征正交分解方法模块、同步压缩变换模块、多参数联合分析模块、三维火焰形态重建模块、故障检测模块;
所述数据采集模块,采用高灵敏度的火焰化学发光传感器,精确捕捉火焰中化学反应产生的光信号,传感器阵列布置在燃烧器的不同位置,获得火焰的整体化学发光特性;
所述本征正交分解方法模块,通过分析识别燃烧过程中的变化模式,准确地预测故障;
所述同步压缩变换模块,通过实时SCT算法对传感器数据进行时频分析,捕捉燃烧过程中的瞬时变化,反映燃烧状态的细微变化,提高故障检测的灵敏度;
所述多参数联合分析模块,结合同步压缩变换的结果,使用多参数联合分析算法,评估火焰状态;
所述三维火焰形态重建模块,利用计算机视觉技术,从不同角度获取的火焰图像,重建三维火焰形态;
所述故障检测模块,分析多参数联合分析和三维火焰形态重建的结果,检测火焰的异常行为。
一种计算机设备,包括存储器和处理器,所述存储器存储有计算机程序,其特征在于,所述处理器执行所述计算机程序时实现一种基于同步压缩变换的振荡燃烧故障检测方法的步骤。
一种计算机可读存储介质,其上存储有计算机程序,其特征在于,所述计算机程序被处理器执行时实现一种基于同步压缩变换的振荡燃烧故障检测方法的步骤。
本发明的有益效果:本发明在数据采集方面,利用高速摄影技术采集火焰图像能够避免传统压力监测方式带来的数据滞后问题,可实现对振荡燃烧故障的无损、快速检测;本征正交分解能够对火焰图像进行数据降维及特征提取,保留火焰图像的丰富信息,反映火焰的整体脉动特点,提高检测速度和准确率;同步压缩变换通过同步压缩算子进行能量重排与瞬时集中,弥补了传统时频分析方法频率分辨率较低的缺点,实现了振荡燃烧故障的精确诊断和快速检测。
附图说明
为了更清楚地说明本发明实施例的技术方案,下面将对实施例描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其它的附图。其中:
图1为本发明一个实施例提供的一种基于同步压缩变换的振荡燃烧故障检测方法流程示意图。
图2为本发明一个实施例提供的一种基于同步压缩变换的振荡燃烧故障检测方法的检测结果。
图3为本发明一个实施例提供的一种基于同步压缩变换的振荡燃烧故障检测方法的火焰振荡及稳定两个工况下的压力信号SST结果图。
图4为本发明一个实施例提供的一种基于同步压缩变换的振荡燃烧故障检测系统的工作模块示意图。
具体实施方式
为使本发明的上述目的、特征和优点能够更加明显易懂,下面结合说明书附图对本发明的具体实施方式做详细的说明,显然所描述的实施例是本发明的一部分实施例,而不是全部实施例。基于本发明中的实施例,本领域普通人员在没有做出创造性劳动前提下所获得的所有其他实施例,都应当属于本发明的保护的范围。
在下面的描述中阐述了很多具体细节以便于充分理解本发明,但是本发明还可以采用其他不同于在此描述的其它方式来实施,本领域技术人员可以在不违背本发明内涵的情况下做类似推广,因此本发明不受下面公开的具体实施例的限制。
其次,此处所称的“一个实施例”或“实施例”是指可包含于本发明至少一个实现方式中的特定特征、结构或特性。在本说明书中不同地方出现的“在一个实施例中”并非均指同一个实施例,也不是单独的或选择性的与其他实施例互相排斥的实施例。
本发明结合示意图进行详细描述,在详述本发明实施例时,为便于说明,表示器件结构的剖面图会不依一般比例作局部放大,而且所述示意图只是示例,其在此不应限制本发明保护的范围。此外,在实际制作中应包含长度、宽度及深度的三维空间尺寸。
同时在本发明的描述中,需要说明的是,术语中的“上、下、内和外”等指示的方位或位置关系为基于附图所示的方位或位置关系,仅是为了便于描述本发明和简化描述,而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作,因此不能理解为对本发明的限制。此外,术语“第一、第二或第三”仅用于描述目的,而不能理解为指示或暗示相对重要性。
本发明中除非另有明确的规定和限定,术语“安装、相连、连接”应做广义理解,例如:可以是固定连接、可拆卸连接或一体式连接;同样可以是机械连接、电连接或直接连接,也可以通过中间媒介间接相连,也可以是两个元件内部的连通。对于本领域的普通技术人员而言,可以具体情况理解上述术语在本发明中的具体含义。
实施例1
参照图1,为本发明的第一个实施例,该实施例提供了一种基于同步压缩变换的振荡燃烧故障检测方法,包括:
S1:采集火焰化学发光信号,对采集到的火焰化学发光信号进行自适应滤波降噪处理。
更进一步的,所述自适应滤波降噪处理包括,进行振荡燃烧故障检测时,区分噪声和实际的燃烧振荡信号,选择基于最陡下降法的自适应滤波器,滤波器结合噪声估计和信号增强;
自适应滤波器更新规则如下:
其中,w(n)表示滤波器权重向量,n表示离散时间步长,表示第n时刻的噪声功率估计,ε表示常数,x(n)表示第n时刻的输入的火焰信号,e(n)表示第n时刻的误差信号,α表示遗忘因子。
S2:利用改进的本征正交分解方法对火焰图像提取1阶模态时间系数,利用高阶模态,获得全面的燃烧特征信息。
更进一步的,所述改进的本征正交分解方法包括,对滤波器处理后的火焰图像集合进行POD分析,得到本征模态和本征值,根据本征值的大小,选择能量最高的M个本征模态,形成一个新的模态集合,代表数据中最显著的空间特征;
对于每个选定的模态计算时间系数,时间系数表示火焰图像在每个本征模态上的投影随时间的变化,计算公式如下:
其中,ci(m)表示时间系数,I(m)表示第m个火焰图像帧,表示图像集合的均值,φi(x,y)表示火焰图像的本征模态,i是模态的索引;
将计算得到的时间系数存储在集合中:
其中,表示包含高阶模态时间系数的集合,集合揭示燃烧过程中的细微波动以及早期故障迹象。
S3:开发实时同步压缩变换算法对数据进行时频分析,并实时更新能量分配,响应燃烧状态的微小变化。
更进一步的,所述开发实时同步压缩变换算法包括,通过同步压缩算子对时频变换的系数进行重排,将信号在时频平面任一点处的时频系数移到能量的重心位置,增强瞬时频率的能量集中程度,在傅里叶变换的基础上套用一个窗函数,将信号时域进行分割,通过窗函数滑动得到不同时刻窗口下的频率分布情况,再将这些短时频谱依时序进行排列,描述信号频率成分的时变规律;
对于确定的时变信号短时傅里叶变换窗函数为g(t),经STFT变换后有:
G(t,f)=∫Rs(τ)g(τ-t)e-i2πf(τ-t)
其中,G(t,f)表示傅里叶变换的结果,s(τ)表示原始信号,τ表示原始信号的时间变量,g(τ-t)表示,t表示时间变量,f表示频率变量,i表示虚数单位;
针对时变谐波信号以及窗函数信号做傅里叶变换,得到STFT时频表示中瞬时频率的估计结果,表示为:
其中,θ(t,ω)表示瞬时频率,ω表示频率变量,,arg(G(t,f))表示辐角,Re表示取实部,表示对时间t的偏导数;
利用上述瞬时频率估计结果收集具有相同频率的STFT系数,根据狄拉克函数性质,给出同步压缩算子函数∫Rδ(η-ω0(t,ω)),得到同步压缩变换,表示为:
Ts(t,η)=∫RG(t,f)δ(η-θ(t,ω))dω
其中,Ts(t,η)表示瞬时频率信号,δ(η-θ(t,ω))表示狄拉克函数,η表示瞬时频率变量,在频率方向上对时频系数进行重新分配排列,在时间t时刻的同步压缩变换过程,即频率点重新分配并历经所有时刻。
S4:结合传感器数据,进行多参数联合分析,获得燃烧状态评估,利用多个高速相机从不同角度获取火焰图像,通过计算机视觉技术重建三维火焰形态,提供故障检测信息。
更进一步的,所述多参数联合分析包括,对传感器收集的数据进行分析,了解分布和特性,通过优化算法调整参数,经过多次迭代,找到最佳的参数组合;
将参数组合代入多参数联合分析公式计算每个传感器的加权贡献,权重考虑传感器对角度、速度和光强的不同敏感度,并应用归一化函数确保不同传感器数据之间的可比性,具体公式为:
其中,G(θ,v,I)表示多参数联合分析主函数,θ表示火焰倾斜角度,v表示火焰燃烧速度,I表示火焰辐射强度,n表示传感器数量,Ωi表示第i个传感器的有效测量域,fii,vi,Ii)表示加权贡献,αi表示第i个传感器角度参数,θi表示第i个传感器测量的角度值,θ0i表示第i个传感器的参考角度值,βi表示第i个传感器速度参数的调整系数,γi表示第i个传感器光强参数的调整系数,分别表示第i个传感器角度测量值的均值和标准差,分别表示第i个传感器速度测量值的均值和标准差。
更进一步的,所述重建三维火焰形态包括,结合多参数联合分析的结果,利用计算机视觉算法重建火焰的三维形态;
通过时间相关函数考虑火焰随时间的变化,得到动态的三维火焰模型:
Fx,y,z,t)=∫ΩGθ,v,I)·htdt
h(t)=sin(ωt+φ)·e-λt
其中,x,y,z表示三维空间的坐标轴,ht表示火焰随时间的变化函数,Ω表示积分的空间域,φ表示相位偏移量,λ表示衰减系数;通过三维火焰模型给出火焰在三维空间中的完整形态,包括随时间的变化。
S5:构建故障预测模型,自动完成整个故障检测流程,实现燃烧故障的早期预警。
更进一步的,所述故障预测模型包括,比较计算出的火焰状态和火焰形态与正常行为阈值,计算结果超出预设的阈值范围,则认为存在异常行为;
当火焰温度低于1000K或高于2500K,则认定异常行为是燃烧不充分、燃烧器堵塞,解决措施为调整燃烧器的氧气/燃料比,检查并清理燃烧器内部的积碳和沉积物;
当火焰振荡频率低于0.5Hz或高于2Hz,以及火焰形状低于0.5或高于1.5,都认定异常行为是燃烧器内部结构损坏、燃料供应不稳定,解决措施为检查燃料供应系统,保持燃料流量稳定;检查燃烧器的喷嘴和燃烧室是否堵塞、检查内部结构是否损坏;
当火焰图像的对比度低于0.6或高于1.5,则认定异常行为是相机设置不当或火焰本身特性引起,解决措施为调整相机的参数改善图像质量,检查相机的镜头是否污染损坏;检查火焰亮度、颜色特性;
将检测到的异常行为和原因汇总为故障检测信息记录在数据库中,用于后续分析和报告,定期生成故障检测报告,监控燃烧过程的健康状况和性能。
实施例2
参照图2和图3,为本发明的一个实施例,提供了一种基于同步压缩变换的振荡燃烧故障检测方法,为了验证本发明的有益效果,通过实验进行科学论证。
图2为本发明实施例所涉及的两个验证工况,首先采集了验证所用实验台的振荡及稳定工况的压力信号,对其进行STFT分析,振荡工况在t=0.75s附近识别到明显的振荡主频,但频率分辨率较差;而对于稳定工况而言,无明显振荡主频出现。图3为两个工况下压力信号的同步压缩变换分析结果,对于振荡工况而言,同步压缩变换对待测信号进行了能量集中,将伪频率区间内的频谱叠加,使能量集中在实际瞬时频率上,提高时频聚集性,使频率分辨率大大提高;两个工况下火焰CH*光强信号的STFT分析结果,振荡工况在t=0.2s附近出现明显的主频响应,稳定工况无振荡主频出现。且火焰CH*光强信号在诊断速度上较压力信号表现出了明显的时间优势。两个工况下火焰CH*光强信号的同步压缩变换分析结果,通过同步压缩变换分析,时频分析结果的频域分辨率大大提高。对火焰图像采用本征正交分解的方法进行数据降维和特征提取,得到1阶模态时间系数,以此作为STFT的数据输入再次进行振荡燃烧故障诊断,以本征正交分解1阶模态时间系数作为STFT的输入,在t=0.01s即出现明显的主频响应,诊断速度大大提高;与压力信号和CH*光强信号一样,也存在频率分辨率低的问题。通过同步压缩变换分析,频谱图的频率分辨率得到很大提高,频率定位更精确。
应说明的是,以上实施例仅用于说明本发明的技术方案而非限制,尽管参照较佳实施例对本发明进行了详细说明,本领域的普通技术人员应当理解,可以对本发明的技术方案进行修改或者等同替换,而不脱离本发明技术方案的精神和范围,其均应涵盖在本发明的权利要求范围当中。
实施例3
本发明第三个实施例,其不同于前两个实施例的是:
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时,可以存储在一个计算机可读取存储介质中。基于这样的理解,本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行本发明各个实施例所述方法的全部或部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM,Read-OnlyMemory)、随机存取存储器(RAM,Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
在流程图中表示或在此以其他方式描述的逻辑和/或步骤,例如,可以被认为是用于实现逻辑功能的可执行指令的定序列表,可以具体实现在任何计算机可读介质中,以供指令执行系统、装置或设备(如基于计算机的系统、包括处理器的系统或其他可以从指令执行系统、装置或设备取指令并执行指令的系统)使用,或结合这些指令执行系统、装置或设备而使用。就本说明书而言,“计算机可读介质”可以是任何可以包含、存储、通信、传播或传输程序以供指令执行系统、装置或设备或结合这些指令执行系统、装置或设备而使用的装置。
计算机可读介质的更具体的示例(非穷尽性列表)包括以下:具有一个或多个布线的电连接部(电子装置)、便携式计算机盘盒(磁装置)、随机存取存储器(RAM)、只读存储器(ROM)、可擦除可编辑只读存储器(EPROM或闪速存储器)、光纤装置以及便携式光盘只读存储器(CDROM)。另外,计算机可读介质甚至可以是可在其上打印所述程序的纸或其他合适的介质,因为可以例如通过对纸或其他介质进行光学扫描,接着进行编辑、解译或必要时以其他合适方式进行处理来以电子方式获得所述程序,然后将其存储在计算机存储器中。
应当理解,本发明的各部分可以用硬件、软件、固件或它们的组合来实现。在上述实施方式中,多个步骤或方法可以用存储在存储器中且由合适的指令执行系统执行的软件或固件来实现。例如,如果用硬件来实现,和在另一实施方式中一样,可用本领域公知的下列技术中的任一项或他们的组合来实现:具有用于对数据信号实现逻辑功能的逻辑门电路的离散逻辑电路,具有合适的组合逻辑门电路的专用集成电路,可编程门阵列(PGA),现场可编程门阵列(FPGA)等。
实施例4
参照图4,为本发明的一个实施例,提供了一种基于同步压缩变换的振荡燃烧故障检测系统,其特征在于:包括数据采集模块、本征正交分解方法模块、同步压缩变换模块、多参数联合分析模块、三维火焰形态重建模块、故障检测模块。
数据采集模块,采用高灵敏度的火焰化学发光传感器,精确捕捉火焰中化学反应产生的光信号,传感器阵列布置在燃烧器的不同位置,获得火焰的整体化学发光特性。
本征正交分解方法模块,通过分析识别燃烧过程中的变化模式,准确地预测故障。
同步压缩变换模块,通过实时SCT算法对传感器数据进行时频分析,捕捉燃烧过程中的瞬时变化,反映燃烧状态的细微变化,提高故障检测的灵敏度;
多参数联合分析模块,结合同步压缩变换的结果,使用多参数联合分析算法,评估火焰状态。
三维火焰形态重建模块,利用计算机视觉技术,从不同角度获取的火焰图像,重建三维火焰形态。
故障检测模块,分析多参数联合分析和三维火焰形态重建的结果,检测火焰的异常行为。
应说明的是,以上实施例仅用于说明本发明的技术方案而非限制,尽管参照较佳实施例对本发明进行了详细说明,本领域的普通技术人员应当理解,可以对本发明的技术方案进行修改或者等同替换,而不脱离本发明技术方案的精神和范围,其均应涵盖在本发明的权利要求范围当中。

Claims (10)

  1. 一种基于同步压缩变换的振荡燃烧故障检测方法,其特征在于:包括,
    采集火焰化学发光信号,对采集到的火焰化学发光信号进行自适应滤波降噪处理;
    利用改进的本征正交分解方法对火焰图像提取1阶模态时间系数,利用高阶模态,获得全面的燃烧特征信息;
    开发实时同步压缩变换算法对数据进行时频分析,并实时更新能量分配,响应燃烧状态的微小变化;
    结合传感器数据,进行多参数联合分析,获得燃烧状态评估,利用多个高速相机从不同角度获取火焰图像,通过计算机视觉技术重建三维火焰形态,提供故障检测信息;
    构建故障预测模型,自动完成整个故障检测流程,实现燃烧故障的早期预警。
  2. 如权利要求1所述的一种基于同步压缩变换的振荡燃烧故障检测方法,其特征在于:所述自适应滤波降噪处理包括,进行振荡燃烧故障检测时,区分噪声和实际的燃烧振荡信号,选择基于最陡下降法的自适应滤波器,滤波器结合噪声估计和信号增强;
    自适应滤波器更新规则如下:
    其中,w(n)表示滤波器权重向量,n表示离散时间步长,表示第n时刻的噪声功率估计,ε表示常数,x(n)表示第n时刻的输入的火焰信号,e(n)表示第n时刻的误差信号,α表示遗忘因子。
  3. 如权利要求2所述的一种基于同步压缩变换的振荡燃烧故障检测方法,其特征在于:所述改进的本征正交分解方法包括,对滤波器处理后的火焰图像集合进行POD分析,得到本征模态和本征值,根据本征值的大小,选择能量最高的M个本征模态,形成一个新的模态集合,代表数据中最显著的空间特征;
    对于每个选定的模态计算时间系数,时间系数表示火焰图像在每个本征模态上的投影随时间的变化,计算公式如下:
    其中,ci(m)表示时间系数,I(m)表示第m个火焰图像帧,表示图像集合的均值,φi(x,y)表示火焰图像的本征模态,i是模态的索引;
    将计算得到的时间系数存储在集合中:
    其中,表示包含高阶模态时间系数的集合,集合揭示燃烧过程中的细微波动以及早期故障迹象。
  4. 如权利要求3所述的一种基于同步压缩变换的振荡燃烧故障检测方法,其特征在于:所述开发实时同步压缩变换算法包括,通过同步压缩算子对时频变换的系数进行重排,将信号在时频平面任一点处的时频系数移到能量的重心位置,增强瞬时频率的能量集中程度,在傅里叶变换的基础上套用一个窗函数,将信号时域进行分割,通过窗函数滑动得到不同时刻窗口下的频率分布情况,再将这些短时频谱依时序进行排列,描述信号频率成分的时变规律;
    对于确定的时变信号短时傅里叶变换窗函数为g(t),经STFT变换后有:
    G(t,f)=∫Rs(τ)g(τ-t)e-i2πf(τ-t)
    其中,G(t,f)表示傅里叶变换的结果,s(τ)表示原始信号,τ表示原始信号的时间变量,g(τ-t)表示,t表示时间变量,f表示频率变量,i表示虚数单位;
    针对时变谐波信号以及窗函数信号做傅里叶变换,得到STFT时频表示中瞬时频率的估计结果,表示为:
    其中,θ(t,ω)表示瞬时频率,ω表示频率变量,,arg(G(t,f))表示辐角,Re表示取实部,表示对时间t的偏导数;
    利用上述瞬时频率估计结果收集具有相同频率的STFT系数,根据狄拉克函数性质,给出同步压缩算子函数∫Rδ(η-ω0(t,ω)),得到同步压缩变换,表示为:
    Ts(t,η)=∫RG(t,f)δ(η-θ(t,ω))dω
    其中,Ts(t,η)表示瞬时频率信号,δ(η-θ(t,ω))表示狄拉克函数,η表示瞬时频率变量,在频率方向上对时频系数进行重新分配排列,在时间t时刻的同步压缩变换过程,即频率点重新分配并历经所有时刻。
  5. 如权利要求4所述的一种基于同步压缩变换的振荡燃烧故障检测方法,其特征在于:所述多参数联合分析包括,对传感器收集的数据进行分析,了解分布和特性,通过优化算法调整参数,经过多次迭代,找到最佳的参数组合;
    将参数组合代入多参数联合分析公式计算每个传感器的加权贡献,权重考虑传感器对角度、速度和光强的不同敏感度,并应用归一化函数确保不同传感器数据之间的可比性,具体公式为:
    其中,G(θ,v,I)表示多参数联合分析主函数,θ表示火焰倾斜角度,v表示火焰燃烧速度,I表示火焰辐射强度,n表示传感器数量,Ωi表示第i个传感器的有效测量域,fii,vi,Ii)表示加权贡献,αi表示第i个传感器角度参数,θi表示第i个传感器测量的角度值,θ0i表示第i个传感器的参考角度值,βi表示第i个传感器速度参数的调整系数,γi表示第i个传感器光强参数的调整系数,分别表示第i个传感器角度测量值的均值和标准差,分别表示第i个传感器速度测量值的均值和标准差。
  6. 如权利要求5所述的一种基于同步压缩变换的振荡燃烧故障检测方法,其特征在于:所述重建三维火焰形态包括,结合多参数联合分析的结果,利用计算机视觉算法重建火焰的三维形态;
    通过时间相关函数考虑火焰随时间的变化,得到动态的三维火焰模型:
    F(x,y,z,t)=∫ΩG(θ,v,I)·h(t)dt
    h(t)=sin(ωt+φ)·e-λt
    其中,x,y,z表示三维空间的坐标轴,h(t)表示火焰随时间的变化函数,Ω表示积分的空间域,φ表示相位偏移量,λ表示衰减系数;通过三维火焰模型给出火焰在三维空间中的完整形态,包括随时间的变化。
  7. 如权利要求6所述的一种基于同步压缩变换的振荡燃烧故障检测方法,其特征在于:所述故障预测模型包括,比较计算出的火焰状态和火焰形态与正常行为阈值,计算结果超出预设的阈值范围,则认为存在异常行为;
    当火焰温度低于1000K或高于2500K,则认定异常行为是燃烧不充分、燃烧器堵塞,解决措施为调整燃烧器的氧气/燃料比,检查并清理燃烧器内部的积碳和沉积物;
    当火焰振荡频率低于0.5Hz或高于2Hz,以及火焰形状低于0.5或高于1.5,都认定异常行为是燃烧器内部结构损坏、燃料供应不稳定,解决措施为检查燃料供应系统,保持燃料流量稳定;检查燃烧器的喷嘴和燃烧室是否堵塞、检查内部结构是否损坏;
    当火焰图像的对比度低于0.6或高于1.5,则认定异常行为是相机设置不当或火焰本身特性引起,解决措施为调整相机的参数改善图像质量,检查相机的镜头是否污染损坏;检查火焰亮度、颜色特性;
    将检测到的异常行为和原因汇总为故障检测信息记录在数据库中,用于后续分析和报告,定期生成故障检测报告,监控燃烧过程的健康状况和性能。
  8. 一种采用如权利要求1~7任一所述的一种基于同步压缩变换的振荡燃烧故障检测方法的系统,其特征在于:包括数据采集模块、本征正交分解方法模块、同步压缩变换模块、多参数联合分析模块、三维火焰形态重建模块、故障检测模块;
    所述数据采集模块,采用高灵敏度的火焰化学发光传感器,精确捕捉火焰中化学反应产生的光信号,传感器阵列布置在燃烧器的不同位置,获得火焰的整体化学发光特性;
    所述本征正交分解方法模块,通过分析识别燃烧过程中的变化模式,准确地预测故障;
    所述同步压缩变换模块,通过实时SCT算法对传感器数据进行时频分析,捕捉燃烧过程中的瞬时变化,反映燃烧状态的细微变化,提高故障检测的灵敏度;
    所述多参数联合分析模块,结合同步压缩变换的结果,使用多参数联合分析算法,评估火焰状态;
    所述三维火焰形态重建模块,利用计算机视觉技术,从不同角度获取的火焰图像,重建三维火焰形态;
    所述故障检测模块,分析多参数联合分析和三维火焰形态重建的结果,检测火焰的异常行为。
  9. 一种计算机设备,包括存储器和处理器,所述存储器存储有计算机程序,其特征在于,所述处理器执行所述计算机程序时实现权利要求1至7中任一项所述的方法的步骤。
  10. 一种计算机可读存储介质,其上存储有计算机程序,其特征在于,所述计算机程序被处理器执行时实现权利要求1至7中任一项所述的方法的步骤。
PCT/CN2024/132262 2024-07-01 2024-11-15 一种基于同步压缩变换的振荡燃烧故障检测方法 Pending WO2026007302A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202410872215.6A CN119106234A (zh) 2024-07-01 2024-07-01 一种基于同步压缩变换的振荡燃烧故障检测方法
CN202410872215.6 2024-07-01

Publications (1)

Publication Number Publication Date
WO2026007302A1 true WO2026007302A1 (zh) 2026-01-08

Family

ID=93712511

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2024/132262 Pending WO2026007302A1 (zh) 2024-07-01 2024-11-15 一种基于同步压缩变换的振荡燃烧故障检测方法

Country Status (2)

Country Link
CN (1) CN119106234A (zh)
WO (1) WO2026007302A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN121658907A (zh) * 2026-02-06 2026-03-13 青岛中微创芯电子有限公司 一种电力电子器件短时过载退化预测方法及系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN121009350B (zh) * 2025-08-18 2026-04-14 深圳妈湾电力有限公司 一种燃气发电数据分析系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103716013A (zh) * 2014-01-14 2014-04-09 苏州大学 变参数比例自适应滤波器
US20230305551A1 (en) * 2022-03-25 2023-09-28 Yokogawa Electric Corporation Method and system for automated fault detection
CN116861320A (zh) * 2023-05-31 2023-10-10 西北工业大学 基于短时傅里叶同步压缩变换的转子故障诊断方法
CN116990055A (zh) * 2023-08-04 2023-11-03 西北工业大学 一种振荡燃烧故障快速检测技术
CN117725526A (zh) * 2023-12-06 2024-03-19 中国船舶集团有限公司第七〇四研究所 一种基于FSST和AlexNet的船用离合器滚动轴承故障诊断方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112668518A (zh) * 2020-12-31 2021-04-16 中国地质大学(武汉) 一种对振动故障信号的vmsst时频分析方法
CN114199819B (zh) * 2021-11-12 2025-01-24 西安热工研究院有限公司 一种适用于燃气轮机的燃烧诊断装置及方法
CN117741505A (zh) * 2023-12-20 2024-03-22 国网陕西省电力有限公司电力科学研究院 基于峰度和Teager能量算子的配电网相继故障检测方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103716013A (zh) * 2014-01-14 2014-04-09 苏州大学 变参数比例自适应滤波器
US20230305551A1 (en) * 2022-03-25 2023-09-28 Yokogawa Electric Corporation Method and system for automated fault detection
CN116861320A (zh) * 2023-05-31 2023-10-10 西北工业大学 基于短时傅里叶同步压缩变换的转子故障诊断方法
CN116990055A (zh) * 2023-08-04 2023-11-03 西北工业大学 一种振荡燃烧故障快速检测技术
CN117725526A (zh) * 2023-12-06 2024-03-19 中国船舶集团有限公司第七〇四研究所 一种基于FSST和AlexNet的船用离合器滚动轴承故障诊断方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN121658907A (zh) * 2026-02-06 2026-03-13 青岛中微创芯电子有限公司 一种电力电子器件短时过载退化预测方法及系统

Also Published As

Publication number Publication date
CN119106234A (zh) 2024-12-10

Similar Documents

Publication Publication Date Title
WO2026007302A1 (zh) 一种基于同步压缩变换的振荡燃烧故障检测方法
JP7840676B2 (ja) 発電システム用のクラウドベースの音響監視、分析、および診断
CN109209783A (zh) 一种基于噪声检测叶片的雷击损伤的方法及装置
CN114565006A (zh) 一种基于深度学习的风力发电机叶片损伤检测方法及系统
CN114199819B (zh) 一种适用于燃气轮机的燃烧诊断装置及方法
US10451268B2 (en) Method for evaluating the clogging of a heat exchanger
CN114255784A (zh) 一种基于声纹识别的变电站设备故障诊断方法及相关装置
US20200409322A1 (en) Control of power generation system by visually monitoring gauge during operation
CN106969828B (zh) 一种汽轮机瓦振传感器故障诊断及通道校验系统及方法
CN106030273A (zh) 爆震强度测量
EP3757696A1 (en) Control of power generation system by visually monitoring valve during operation
CN104596775A (zh) 发动机失火诊断系统及其方法
CN119624093A (zh) 一种基于大数据分析的煤矿粉尘浓度实时监测方法
Sheard et al. Stall warning in a low-speed axial fan by visualization of sound signals
KR102256980B1 (ko) 기계학습을 기반으로 한 가스터빈 연소불안정 진단 시스템 및 이를 이용한 가스터빈 연소불안정 진단 방법
CN121165911B (zh) 一种散热单元控制方法及电子设备
CN109945968A (zh) 一种检测硬盘多部位受噪音冲击大小的装置、方法及系统
CN120561827B (zh) 一种基于图像融合分析的多源压力设备安全监测系统
CN121055315A (zh) 光伏微电网节能管理与智能调度方法及系统
Singla et al. Combustion dynamics validation of an annular reheat combustor
CN120522294A (zh) 变电站连接金具损伤识别方法、装置及计算机设备
KR102247305B1 (ko) 제로크로싱레이트를 기반으로 한 가스터빈 연소불안정 진단 시스템 및 이를 이용한 가스터빈 연소불안정 진단 방법
KR102256981B1 (ko) 필터 뱅크를 이용한 가스터빈 연소불안정 진단 시스템 및 이를 이용한 가스터빈 연소불안정 진단 방법
D’Elia et al. Step-by-step algorithm for the simulation of faulted bearings in non-stationary conditions
CN116990055A (zh) 一种振荡燃烧故障快速检测技术

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 24946168

Country of ref document: EP

Kind code of ref document: A1