An engine combustion diagnosis identification system and method based on optical fiber sensing monitoring

Abstract

The application provides an engine combustion diagnosis identification system and method based on optical fiber sensing monitoring, which comprises a signal acquisition module, a signal processing module, a feature extraction module, a parameter calculation module, a state matching module and a diagnosis output module; the signal acquisition module is used for acquiring continuous sensing signals collected by distributed optical fiber sensors arranged around an igniter of a combustion chamber; the signal processing module is used for filtering and photoelectric conversion of the continuous sensing signals to obtain spontaneous spectral intensity data; the feature extraction module is used for extracting space-time features from the spontaneous spectral intensity data to generate a feature vector set; the parameter calculation module is used for calculating combustion characteristic parameters based on the feature vector set; the state matching module is used for generating an operating state index; and the diagnosis output module is used for outputting a combustion diagnosis report according to a combustion health state evaluation strategy library. The application has reasonable concept and can significantly improve the combustion state identification precision and fault early warning capability.

Description

Technical Field

[0001] This invention relates to the technical field of combining fiber optics and data processing in the combustion process of aerospace engines, and specifically to an engine combustion diagnosis and identification system and method based on fiber optic sensing monitoring. Background Technology

[0002] In the field of aerospace propulsion, especially in advanced air-breathing propulsion systems such as scramjet engines, real-time and accurate diagnosis of the combustion process is a key technical challenge for optimizing performance, ensuring stable operation, and preventing malfunctions. Scramjet engines operate under hypersonic conditions, with extremely high airflow velocity within their combustion chambers and a complex and variable combustion environment. This environment exhibits strong turbulence, shock wave-combustion coupling, and unsteady combustion phenomena such as combustion oscillations, partial flameout, and thermo-acoustic instability. Traditional contact measurement methods, such as thermocouples and pressure probes, not only interfere with the flow field and combustion process but also struggle to operate stably for extended periods under extreme high-temperature and high-speed conditions. Furthermore, they cannot achieve simultaneous distributed measurement of multiple locations and parameters within the combustion chamber. While existing optical diagnostic methods (such as PLIF) can achieve non-contact measurement, these systems are complex, costly, and often limited to laboratory environments, making them unsuitable for direct application in long-term online monitoring of actual engines.

[0003] Therefore, there is an urgent need to develop an online diagnostic technology that can adapt to the actual working environment of engines, has high spatiotemporal resolution, and enables distributed sensing of key combustion parameters. Fiber optic sensing technology, with its advantages of resistance to electromagnetic interference, high temperature resistance, small size, and ability to achieve multi-point distributed measurement, provides a new technical approach for real-time monitoring of complex combustion processes in engine combustion chambers. However, existing fiber optic sensing solutions in combustion diagnosis mostly focus on the measurement of single physical quantities (such as temperature or pressure), lacking the ability to synchronously sense and comprehensively analyze the coupling effects of multiple physical fields during combustion, and failing to effectively establish a complete diagnostic system from raw signals to intelligent identification of combustion status and fault early warning. Summary of the Invention

[0004] To address the technical problems existing in the background art, this invention proposes an engine combustion diagnosis and identification system and method based on fiber optic sensing monitoring. Its concept is reasonable and realizes non-contact, multimodal, distributed real-time monitoring of key parameters of the combustion process. Through spatiotemporal feature extraction and historical path matching, it significantly improves the accuracy of combustion status identification and fault early warning capability. It can acquire the three-dimensional structure, thermochemical information and dynamic characteristics of the flame without disturbing the flame, providing reliable data support for engine combustion chamber design optimization, status assessment and health management.

[0005] To solve the above-mentioned technical problems, the present invention provides an engine combustion chamber combustion diagnosis and identification system based on multimodal fiber optic sensing and monitoring, which includes a signal acquisition module, a signal processing module, a feature extraction module, a parameter calculation module, a state matching module and a diagnostic output module;

[0006] The signal acquisition module is used to acquire continuous sensing signals collected by distributed fiber optic sensors deployed around the combustion chamber igniter.

[0007] The signal processing module is used to filter and perform photoelectric conversion on the continuous sensing signal to obtain spontaneous spectral intensity data;

[0008] The feature extraction module is used to extract spatiotemporal features from the spontaneous spectral intensity data and generate a feature vector set.

[0009] The parameter calculation module is used to calculate combustion characteristic parameters based on the feature vector set;

[0010] The state matching module is used to perform similarity matching between the feature vector set and the historical combustion evolution path to generate operating state indicators;

[0011] The diagnostic output module is used to output a combustion diagnostic report based on the combustion health status assessment strategy library.

[0012] An engine combustion diagnosis and identification system and method based on fiber optic sensing monitoring, specifically including the following steps, based on the aforementioned engine combustion diagnosis and identification system based on multimodal fiber optic sensing monitoring:

[0013] 1) Distributed fiber optic sensors are installed at specific locations of the igniter in the combustion chamber, and the signal acquisition module synchronously acquires continuous sensing signals during the combustion process through multiple channels;

[0014] 2) The signal processing module performs filtering and photoelectric conversion processing on the collected continuous sensing signals to obtain spontaneous spectral intensity data within a specific wavelength range;

[0015] 3) The feature extraction module extracts spatiotemporal features from the acquired spontaneous spectral intensity data and generates a feature vector set characterizing the micro-variation features of the multi-physics field in the combustion chamber;

[0016] 4) Based on the generated feature vector set, the parameter calculation module calculates at least one of the following combustion characteristic parameters: flame centroid position, heat release rate, ignition delay time, flame propagation speed, and its probability density function;

[0017] 5) Call the set of combustion evolution paths in the historical database, and the state matching module performs similarity matching between the feature vector set and the historical paths to generate an operating status index containing flame position, flame intensity, temperature information and time dynamic characteristic index.

[0018] 6) The diagnostic output module determines the overall ignition and combustion status of the combustion chamber based on the preset combustion chamber combustion health status assessment strategy library, assesses the probability of potential fault risks, and outputs a combustion diagnostic report that includes the location coordinates of the fault area and the quantitative analysis results of combustion turbulence characteristics.

[0019] As a preferred embodiment of the present invention, step 1) specifically includes the following steps:

[0020] 1.1) Determine the deployment locations and channel configurations of the distributed fiber optic sensors.

[0021] The sensor array probes are deployed within a predefined monitoring area around the combustion chamber igniter to achieve continuous sensing of the spatial distribution of the combustion process in a multi-point, multi-channel manner. Each fiber optic sensor channel corresponds to an independent sensing unit, or a fiber optic coupler is used to split the signal collected by a single fiber optic sensor into multiple channels for collecting self-emitting signals within a specific wavelength or band range. The fiber optic coupler is used to split the signal collected by a single fiber optic sensor into multiple channels, and a synchronous triggering mechanism is used between the channels to ensure the spatiotemporal alignment of the signal acquisition, so as to facilitate subsequent multiphysics coupling analysis.

[0022] 1.2) Acquire multi-channel synchronous continuous sensor signal stream

[0023] During engine operation, the self-luminous signal of the flame generated in the combustion chamber is captured in real time by fiber optic sensors placed at specific locations, forming a raw sensing signal containing time-series information. The signal acquisition module synchronously and continuously acquires the self-luminous signal of each fiber optic sensor channel at a preset sampling frequency to obtain a multi-channel synchronous continuous sensing signal stream, ensuring that rapid dynamic changes that may occur during combustion can be captured.

[0024] 1.3) Signal preprocessing and quality verification

[0025] The obtained multi-channel synchronous continuous sensing signal stream is subjected to preliminary signal preprocessing and quality verification.

[0026] As a preferred embodiment of the present invention, the specific process of step 1.3) is as follows: check the validity of the self-emitting signals of each fiber optic sensor channel, identify and mark the signal abnormality segments that may be caused by fiber optic sensor failure or external interference; calibrate and correct the original sensing signal formed in step 1.2) to eliminate noise components such as zero drift introduced by ambient light or fiber optic sensor background noise; record the trigger and acquisition timestamps of each channel in the multi-channel signal bundled by a fiber optic coupler and the corresponding engine operating condition parameters, so as to lay a data foundation for subsequent information processing.

[0027] As a preferred embodiment of the present invention, the specific process for calibrating and correcting the original sensing signal formed in step 1.2) is as follows:

[0028] 1.2.1) Calibration

[0029] Using a standard cold light source: Install filters to perform photoelectric conversion on the signals collected by all channels of the fiber optic sensor probe, and calibrate the gain and voltage response of the components so that each channel outputs the same voltage value under the same light intensity.

[0030] Check fiber optic status: If the fiber optic cable corresponding to the fiber optic sensor is found to be damaged during the calibration process, resulting in no signal, then the data of that channel is excluded. The signals of the remaining channels are valid and can be used in subsequent tests.

[0031] Gain adjustment and recalibration: After the first engine ignition test, the voltage is measured, and the gain of the photomultiplier tube is adjusted according to the voltage value. After the gain is adjusted, the voltage response is recalibrated using a standard cold light source to ensure signal quality.

[0032] 1.2.2) Signal background correction during the experiment

[0033] A cold flow test was performed before each ignition test, and the background signal was recorded. During data processing, the median value of the cold flow test data of each channel was taken as the background signal, and the self-ignition signal was zero-drift corrected to eliminate the inherent noise of the system.

[0034] 1.2.3) Signal normalization processing

[0035] The maximum value of the zero-drift corrected data of each channel is taken as the reference value, and the signal is dimensionless to normalize the signal range to [0, 1]. The normalized data is used for subsequent calculations of ignition delay, heat release rate, equivalence ratio, flame centroid and pulsation velocity.

[0036] As a preferred embodiment of the present invention, after obtaining spontaneous spectral intensity data within a specific wavelength range in step 2), it is also necessary to extract spontaneous spectral temporal features. The specific process is as follows: calculate the temporal statistical features of the spontaneous spectral intensity data within a specific wavelength range obtained by each channel, and extract temporal waveform features from the spontaneous spectral intensity data.

[0037] As a preferred embodiment of the present invention, step 3) specifically includes the following steps:

[0038] 3.1) Extraction of Multi-channel Spatial Distribution Features

[0039] Based on spontaneous spectral intensity data acquired synchronously through multiple channels, a two-dimensional or three-dimensional light intensity distribution map of the combustion chamber monitoring area is constructed, and spatial distribution features are extracted from the light intensity distribution map.

[0040] 3.2) Spatiotemporal joint feature fusion

[0041] The extracted temporal and spatial features are fused to generate a feature vector set. The specific process is as follows: the temporal features of each channel are concatenated in channel order and fused with the spatial distribution features. The fused feature vectors are associated with the corresponding timestamps and operating condition labels and arranged in time series to form a feature vector set representing the multi-physics field of the combustion chamber.

[0042] As a preferred embodiment of the present invention, step 4) specifically includes the following steps:

[0043] 4.1) Calculation of the flame centroid position

[0044] The fiber optic sensor positions are used to define the flame centroid and calculate flame pulsation velocity. In a Cartesian coordinate system, the igniter position is defined as the zero point. The flame centroid position is calculated by weighting the sensor placement positions with dimensionless light intensity; the weights are the normalized light intensity values ​​at each sensor location. The formula for calculating the flame centroid position is as follows:

[0045] ;

[0046] In the above formula, C represents the coordinates of the flame's center of mass; I i P is the normalized light intensity value at the i-th sensing point; i is the spatial coordinate of the point; N is the number of effective sensing points; the flame centroid position is used to characterize the overall position of the flame in the combustion chamber and its dynamic offset;

[0047] 4.2) Estimation of heat release rate

[0048] Based on the spontaneous spectral intensity data, combined with the pre-calibrated correspondence between wavelength and heat release rate, the local heat release rate is characterized by light signals of a specific wavelength, and the intensity of the 300nm light signal corresponding to the hydroxyl main peak OH*(0,0) is used as the heat release rate proxy to sense the heat release.

[0049] 4.3) Determination of ignition delay time

[0050] In the feature vector set of the fused time series, the starting time point corresponding to the ignition trigger signal is identified, as well as the time point from the starting time point to the first time when the flame intensity reaches the preset ratio of steady-state intensity, which is the ignition delay time; then the mean and variance of the ignition delay time of multiple consecutive ignition processes are statistically analyzed.

[0051] 4.4) Calculation of flame propagation speed

[0052] Based on the time-varying sequence of the flame centroid position, the centroid's moving velocity is calculated as the macroscopic flame propagation velocity; or, based on the time delay estimation of multi-channel light intensity signals, the propagation time of the flame front between adjacent sensors is calculated through cross-correlation analysis, and the local flame propagation velocity is calculated by combining the sensor spacing. The flame propagation velocity is defined as the derivative of the flame centroid position change with time.

[0053] ;

[0054] In the above formula, d ij Let Δt be the distance between sensors i and j. ij Let i be the time delay for the flame front to propagate from i to j;

[0055] 4.5) Probability density function estimation

[0056] For time series data of flame propagation speed, calculate its probability density function to characterize the statistical distribution characteristics, stability and abnormal fluctuations of the probability density.

[0057] 4.6) Equivalent ratio

[0058] Based on the characteristic wavelength radiation intensity ratio in the spontaneous spectral intensity data, and / or based on at least one of the combustion characteristic parameters as the equivalence ratio proxy, the local equivalence ratio is sensed; combined with a pre-established equivalence ratio and feature correlation model or calibration function, the local or overall equivalence ratio distribution and its changing trend over time within the combustion area are calculated in real time.

[0059] As a preferred embodiment of the present invention, step 5) specifically includes the following steps:

[0060] 5.1) Construction and index generation of the combustion evolution path database

[0061] A historical database of combustion evolution paths is pre-established, which stores multiple combustion state evolution paths recorded under different operating conditions. Each path consists of a sequence of feature vectors arranged in chronological order, and is labeled with its operating parameters, combustion state label, flame position, flame intensity, temperature information, and actual observed or high-precision measured values ​​of time dynamic characteristic index. A multi-dimensional index structure based on key features is established for the set of combustion evolution paths in the historical database. The key features include at least one of the following: average light intensity, mean flame centroid coordinates, heat release rate, and mean flame propagation speed.

[0062] 5.2) Similarity matching between real-time feature vectors and historical paths

[0063] The feature vector sequence within the current time window is used as the query sequence, and the similarity is calculated with each combustion evolution path in the historical database. The difference in query sequence length is considered during the calculation process, and the matching distance or similarity score between the query sequence and each historical path is output.

[0064] 5.3) Candidate Path Selection and Weight Allocation

[0065] Based on the similarity score, at least one historical path that best matches the query sequence is selected as a candidate path;

[0066] 5.4) Weighted generation of operational status indicators:

[0067] Based on the flame location index, flame intensity index, and temperature information index associated with the candidate path, and combined with their assigned weights, an estimated value of the corresponding operating status index within the current time window is generated by weighted averaging. If multiple candidate paths are selected, the estimated value is a weighted sum of the values ​​corresponding to each candidate path.

[0068] As a preferred embodiment of the present invention, step 6) specifically includes the following steps:

[0069] 6.1) Construction and Invocation of the Combustion Health Status Assessment Strategy Library

[0070] A combustion health status assessment strategy library is pre-established, which contains multiple assessment rules and logical models. The assessment strategy library is constructed based on at least one of expert knowledge base, historical fault data, and combustion simulation data, and covers the feature criteria and thresholds corresponding to various states such as normal combustion, unstable combustion, partial flameout, thermoacoustic oscillation, and ignition failure. Each assessment rule in the assessment strategy library is associated with one or more operating status indicators and includes corresponding risk assessment weights and warning levels.

[0071] 6.2) Determination of overall combustion ignition state and combustion state:

[0072] The operational status indicators are input into the combustion health status assessment strategy library and judged step by step according to preset rules.

[0073] 6.3) Potential Failure Risk Probability Assessment and Location

[0074] When the determination result indicates that there is an abnormal combustion, the following steps may be performed, including but not limited to:

[0075] 6.3.1) Risk probability calculation: Based on the degree of deviation between the abnormal indicator and the corresponding threshold, and combined with the contribution weight of the indicator in historical failures, assess the probability of occurrence of each potential failure mode.

[0076] 6.3.2) Fault area location: Based on the spatial distribution anomalies of multi-channel spontaneous spectral intensity data and the abrupt change points of the flame centroid trajectory, combined with the location coordinates of the fiber optic sensor deployment, the spatial coordinates or area identifiers of the abnormal area are obtained.

[0077] 6.4) Quantitative Analysis of Combustion Turbulence Characteristics

[0078] Based on the time dynamic characteristic indicators in the feature vector set, especially the fluctuation distribution of flame propagation speed and the pulsating spectrum characteristics of light intensity signal, the combustion turbulence characteristics are analyzed to characterize the strength and stability of the coupling effect between flow and combustion in the combustion chamber.

[0079] 6.5) Generation and Output of Combustion Diagnostic Reports

[0080] By integrating the aforementioned step-by-step judgment results, fault risk probability assessment, fault area location, and quantitative analysis of combustion turbulence characteristics, the diagnostic output module generates and outputs a structured combustion diagnostic report.

[0081] As a preferred embodiment of the present invention, the process of step 6.2) of making step-by-step judgments according to preset rules includes:

[0082] 6.2.1) Ignition status determination: Based on the ignition delay time, initial flame propagation speed and initial flame intensity, determine whether ignition is successful;

[0083] 6.2.2) Combustion state determination: Based on the stability of the flame centroid position, the fluctuation characteristics of the heat release rate, and the oscillation frequency and amplitude in the time dynamic characteristic index, the combustion is determined to be in a stable combustion, slow-vibration combustion, or violent oscillation state.

[0084] As a preferred embodiment of the present invention, the content of the structured combustion diagnostic report in step 6.5) includes, but is not limited to:

[0085] 6.5.1) Classification and confidence level of overall combustion status;

[0086] 6.5.2) Potential fault types, risk probabilities, and location information;

[0087] 6.5.3) Comparative analysis of key combustion parameters with standard operating conditions;

[0088] 6.5.4) Quantitative indicators of turbulence characteristics and their impact on combustion stability assessment;

[0089] 6.5.5) Maintenance suggestions or control adjustment prompts.

[0090] By adopting the above technical solution, the present invention has the following beneficial effects:

[0091] This invention enables non-contact, multimodal, distributed real-time monitoring of key parameters in the combustion process; it significantly improves the accuracy of combustion status identification and fault early warning capabilities through spatiotemporal feature extraction and historical path matching; it can acquire the three-dimensional structure, thermochemical information, and dynamic characteristics of the flame without disturbing the flame; and it provides reliable data support for the design optimization, condition assessment, and health management of engine combustion chambers.

[0092] This invention can solve the technical problem of synchronous monitoring of multiple parameters and accurate identification of status in high-speed power devices such as scramjet engines under complex combustion environments, and provide reliable technical support for realizing real-time optimized control and health management of engine combustion processes.

[0093] This invention enables non-contact, multimodal, and distributed real-time monitoring of key parameters in the combustion process. Through spatiotemporal feature extraction and historical path matching, it significantly improves the accuracy of combustion status identification and fault early warning capabilities. It can acquire the three-dimensional structure, thermochemical information, and dynamic characteristics of the flame without disturbing the flame, providing reliable data support for the design optimization, condition assessment, and health management of engine combustion chambers. This invention is applicable to the condition monitoring, fault diagnosis, and performance optimization of combustion systems such as aero-engines and gas turbines.

[0094] This invention can effectively improve the accuracy and reliability of engine combustion operation status identification. It adopts non-contact measurement, which is of great significance to basic combustion research, engine design and development and industrial burner optimization. Without disturbing the flame, it can obtain key thermochemical information, capture the difference between the actual flame and the ideal symmetrical model, and reveal the three-dimensional structure of the flame through multi-point measurement. It can achieve high-precision identification and status assessment of the dynamic characteristics of the combustion process, and provide key data support for combustion chamber performance optimization and fault early warning. Attached Figure Description

[0095] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0096] Figure 1 This is a schematic diagram illustrating the implementation process of an engine combustion diagnosis and identification system and method based on fiber optic sensing monitoring, provided in an embodiment of the present invention.

[0097] Figure 2 This is a schematic diagram of the composition structure of an engine combustion chamber combustion diagnosis and identification device based on multimodal fiber optic sensing monitoring, provided in an embodiment of the present invention. Detailed Implementation

[0098] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0099] The present invention will be further explained below with reference to specific embodiments.

[0100] like Figure 2 The present invention provides an engine combustion chamber combustion diagnosis and identification system based on multimodal fiber optic sensing monitoring, which includes a signal acquisition module 210, a signal processing module 220, a feature extraction module 230, a parameter calculation module 240, a state matching module 250, and a diagnostic output module 260.

[0101] The signal acquisition module 210 is used to acquire continuous sensing signals from distributed fiber optic sensors deployed around key components of the combustion chamber, such as the igniter.

[0102] The signal processing module 220 is used to filter and perform photoelectric conversion on the signal to obtain spontaneous spectral intensity data.

[0103] The feature extraction module 230 is used to extract spatiotemporal features from the spontaneous spectral intensity data and generate a feature vector set.

[0104] The parameter calculation module 240 is used to calculate combustion characteristic parameters based on the feature vector set.

[0105] The state matching module 250 is used to perform similarity matching between the feature vector set and the historical combustion evolution path to generate operating status indicators.

[0106] The diagnostic output module 260 is used to output a combustion diagnostic report based on the combustion health status assessment strategy library.

[0107] like Figure 1 As shown, this embodiment provides an engine combustion diagnosis and identification method based on fiber optic sensing monitoring. Based on the aforementioned engine combustion diagnosis and identification system based on multi-mode fiber optic sensing monitoring, it specifically includes the following steps:

[0108] S100: Distributed fiber optic sensors are installed at specific locations on the igniter in the combustion chamber, and the signal acquisition module 210 synchronously acquires continuous sensing signals during the combustion process via multiple channels. As one implementation method, step S100 specifically includes the following steps:

[0109] S101. Determine the deployment location and channel configuration of the distributed fiber optic sensors: In a preferred embodiment of the present invention, at least one or more fiber optic sensors are included. The fiber optic sensors are deployed within a predefined monitoring area around the combustion chamber igniter to achieve continuous sensing of the spatial distribution of the combustion process in a multi-point, multi-channel manner. Each fiber optic sensor channel corresponds to an independent sensing unit, or a fiber optic coupler is used to split the signal collected by a fiber optic sensor into multiple channels for collecting self-emitting signals within a specific wavelength or band range. A fiber optic coupler is used to split the signal collected by a fiber optic sensor into signals of each channel in the multiple channels. A synchronous triggering mechanism is used between the channels to ensure the spatiotemporal alignment of the signal acquisition for subsequent multiphysics coupling analysis.

[0110] S102. Acquiring Multi-Channel Synchronous Continuous Sensing Signal Stream: During engine operation, the self-luminous signal of the flame generated in the combustion chamber is captured in real time by fiber optic sensors placed at specific locations, forming a raw sensing signal containing time-series information. The signal acquisition module 210 synchronously and continuously acquires the self-luminous signal of each fiber optic sensor channel at a preset sampling frequency (e.g., not less than 1MHz) to obtain a multi-channel synchronous continuous sensing signal stream, ensuring that rapid dynamic changes that may occur during combustion, such as partial flameout, successful ignition, and turbulent pulsation, can be captured.

[0111] S103. Signal Preprocessing and Quality Verification: Before sending the sensor signal stream into the subsequent processing module, preliminary signal preprocessing and quality verification are performed, including: checking the validity of the self-emitting signals of each fiber optic sensor channel; identifying and marking signal abnormalities that may be caused by fiber optic sensor failure or external interference (such as abnormal sensor signals, signal loss, saturation noise, etc., thus requiring readjustment of sensor gain); calibrating and correcting the original sensor signal formed in step S102 to eliminate noise components such as zero drift introduced by ambient light or fiber optic sensor background noise; recording the signal collected by a fiber optic sensor into multiple channels using a fiber optic coupler, the trigger and acquisition timestamps of each channel signal, and the corresponding engine operating parameters (such as intake pressure, fuel flow, etc.), laying the data foundation for subsequent information processing.

[0112] The specific process for calibrating and correcting the original sensing signal formed in step S102 above is as follows:

[0113] S1021, Calibration

[0114] Using a standard cold light source: With filters installed on the sensor probe, the gain and voltage response of the photoelectric conversion elements of all channels are calibrated; the purpose is to ensure that each channel outputs the same voltage value under the same light intensity.

[0115] Check fiber optic status: If the fiber optic cable corresponding to the sensor channel is found to be damaged during calibration, resulting in no signal, then the data of that channel is excluded. The signals of the remaining channels are valid and can be used in subsequent tests.

[0116] Gain adjustment and recalibration: After the first engine ignition test, the voltage is measured, and the gain of the photomultiplier tube is adjusted according to the voltage value; after the gain adjustment, the voltage response is recalibrated using a standard cold light source to ensure signal quality.

[0117] S1022, Signal background correction during the experiment

[0118] A cold flow test was performed before each ignition test, and the background signal was recorded. During data processing, the median value of the cold flow test data of each channel was taken as the background signal, and the self-ignition signal was zero-drift corrected to eliminate the inherent noise of the system.

[0119] S1023, Signal Normalization Processing

[0120] The maximum value of the zero-drift corrected data of each channel is taken as the reference value, and the signal is dimensionless to normalize the signal range to [0, 1]. The normalized data is used for subsequent calculations of ignition delay, heat release rate, equivalence ratio, flame centroid and pulsation velocity.

[0121] Through the above steps, a raw sensor signal stream with spatiotemporal synchronization and multi-channel continuous observation is obtained, providing a reliable data input basis for subsequent filtering, feature extraction and combustion state diagnosis.

[0122] S200, The signal processing module 220 performs filtering and photoelectric conversion processing on the acquired continuous sensing signal (the filtering and photoelectric conversion use a specific wavelength filter and a photoelectric conversion module, both of which are conventional modules) to obtain spontaneous spectral intensity data within a specific wavelength range. It is also necessary to extract the spontaneous spectral temporal features. The specific process is as follows: For the spontaneous spectral intensity data within a specific wavelength range acquired by each channel, calculate its temporal statistical features, including but not limited to: at least one of the following: mean, variance, skewness, kurtosis, and root mean square value.

[0123] For example, if n data points v are collected... i The following methods are conventional statistical methods for calculating the mean and variance:

[0124] For the collected signal V i, The mean value is calculated as follows, where i takes values ​​from 1 to n:

[0125] ;

[0126] The sample variance is calculated as follows:

[0127] ;

[0128] Furthermore, time-domain waveform features are extracted from spontaneous spectral intensity data.

[0129] S300: The feature extraction module 230 extracts spatiotemporal features from the acquired spontaneous spectral intensity data and generates a feature vector set characterizing the micro-variation features of the multi-physics field in the combustion chamber. As one implementation, step S300 specifically includes the following steps:

[0130] S301. Multi-channel spatial distribution feature extraction: Based on the spontaneous spectral intensity data acquired synchronously from multiple channels, construct a light intensity distribution map within the combustion chamber monitoring area; extract spatial distribution features from the light intensity distribution map, including but not limited to the asymmetry of light intensity distribution.

[0131] S302, Spatiotemporal joint feature fusion: The extracted temporal features and spatial features are fused to generate a feature vector set; specifically, the temporal features of each channel are concatenated in channel order and fused with the spatial distribution features. The fused feature vectors are then associated with the corresponding timestamps and operating condition (different fuels, oxidants, or different injection pressures, etc.) labels and arranged in time sequence to form the feature vector set characterizing the multiphysics field of the combustion chamber.

[0132] Through the above steps, the original multi-channel spontaneous spectral intensity data can be converted into a structured feature vector set with clear physical meaning, providing a data foundation for subsequent calculation of combustion characteristic parameters and combustion state matching.

[0133] S400: Based on the generated feature vector set, the parameter calculation module 240 calculates at least one combustion characteristic parameter from the following: flame centroid position, heat release rate, ignition delay time, flame propagation velocity, and its probability density function. As one implementation, step S400 may include the following steps:

[0134] S401. Calculation of flame centroid position:

[0135] The fiber optic sensor positions are used to define the flame centroid and calculate flame pulsation velocity. In a Cartesian coordinate system, the igniter position is defined as the zero point. The flame centroid position is calculated by weighting the sensor placement positions with dimensionless light intensity; where the weights are the normalized light intensity values ​​at each sensor location. The formula for calculating the flame centroid position is as follows:

[0136] ;

[0137] Where C is the coordinate of the flame's centroid, I i P is the normalized light intensity value at the i-th sensing point. i The coordinates of the point are denoted as N, and the number of effective sensing points is N. The position of the flame centroid is used to characterize the overall position of the flame in the combustion chamber and its dynamic offset.

[0138] S402, Heat release rate estimation:

[0139] Based on spontaneous spectral intensity data and combined with the pre-calibrated correspondence between wavelength and heat release rate, the heat release rate is calculated by multi-wavelength radiation intensity inversion. For example, the intensity of the 300nm light signal corresponding to the hydroxyl main peak OH*(0,0) can be used as a surrogate quantity for heat release rate to sense the heat release; or based on the components in the feature vector that characterize the frequency and amplitude of light intensity pulsation, the instantaneous heat release rate and its changing trend can be estimated in real time through empirical models or data-driven regression models.

[0140] S403, Ignition delay time determination:

[0141] In the feature vector set of the fused time series, the starting time point corresponding to the ignition trigger signal is identified, as well as the time point from that starting time point until the flame intensity first reaches a preset proportion of steady-state intensity (for example, the ignition delay time is defined as the time stamp corresponding to the first rise of the 430nm light signal corresponding to the hydroxyl subpeak OH*(0,2) to 75% of T0+2000us after time T0). Further, the mean and variance of the ignition delay time can be statistically calculated based on the ignition delay time of multiple consecutive ignition processes.

[0142] S404, Calculation of flame propagation speed:

[0143] Based on the time-varying sequence of the flame centroid position, the centroid's moving velocity is calculated as the macroscopic flame propagation velocity; or, based on the time delay estimation of multi-channel light intensity signals, the propagation time of the flame front between adjacent sensors is calculated through cross-correlation analysis, and the local flame propagation velocity is calculated by combining the sensor spacing. The flame propagation velocity is defined as the derivative of the flame centroid position change with time.

[0144] ;

[0145] Where, d ij Let Δt be the distance between sensors i and j. ij Let be the time delay for the flame front to propagate from i to j.

[0146] S405, Probability density function estimation:

[0147] For time-series data of flame propagation velocity, conventional statistical methods are used to calculate its probability density function to characterize the statistical distribution characteristics, stability, and abnormal fluctuations of this probability density. The specific calculation process is as follows: For n collected velocity data points v={v1,v2,…,v…} n The kernel density estimate at point v is:

[0148] ;

[0149] Where h is the bandwidth. σ is the standard deviation of the data.

[0150] S406, Equivalent Ratio:

[0151] Based on the characteristic wavelength radiation intensity ratio in the spontaneous spectral intensity data, and / or based on at least one of the combustion characteristic parameters as the equivalence ratio proxy (such as the collected chemiluminescence intensity), the local equivalence ratio is sensed; combined with a pre-established equivalence ratio and feature correlation model or calibration function, the local or overall equivalence ratio distribution and its trend over time in the combustion area are calculated in real time.

[0152] Equivalent ratio: refers to the ratio of fuel and air that react exactly completely, with no residue remaining; for the hydrogen-air reaction, this ratio is approximately 1:34.3 (mass ratio).

[0153] Φ = 1: This indicates that the fuel and air react completely (chemical equivalent).

[0154] Φ > 1: This indicates high fuel content and low air content, and is called "rich in oil".

[0155] Φ < 1: This indicates low fuel content and high air content, and is called "lean fuel".

[0156] The ratio of collected light intensities can be used as a proxy for the equivalence ratio. For example, in the hydrogen-air reaction, the ratio of the 300nm light signal intensity corresponding to the hydroxyl main peak OH*(0,0) to the 430nm light signal intensity corresponding to the hydroxyl subpeak OH*(0,2) can be used as the equivalence ratio proxy.

[0157] Through the above sub-steps, the system can extract key combustion characteristic parameters with clear physical meaning from the high-dimensional feature vector set, providing a quantitative basis for subsequent state matching, health assessment and fault diagnosis.

[0158] S500: The combustion evolution path set in the historical database is retrieved, and the state matching module 250 performs similarity matching between the feature vector set and the historical paths to generate an operational state index containing flame position, flame intensity, temperature information, and time dynamic characteristic index. As one implementation, step S500 may include the following steps:

[0159] S501, Construction and Indexing of Combustion Evolution Path Database

[0160] A historical database of combustion evolution paths is pre-established, which stores multiple combustion state evolution paths recorded under different operating conditions. Each path consists of a sequence of feature vectors arranged in chronological order, and is labeled with its operating parameters, combustion state labels, flame position, flame intensity, temperature information, and actual observed or high-precision measured values ​​of time dynamic characteristic indices. A multi-dimensional index structure based on key features is established for the set of combustion evolution paths in the historical database. The key features include at least one of the following: average light intensity, mean flame centroid coordinates, heat release rate, and mean flame propagation speed.

[0161] S502, Similarity matching between real-time feature vectors and historical paths:

[0162] The feature vector sequence within the current time window is used as the query sequence, and similarity is calculated with each combustion evolution path in the historical database. The difference in query sequence length is taken into account during the calculation, and the matching distance or similarity score between the query sequence and each historical path is output.

[0163] S503, Candidate Path Selection and Weight Allocation:

[0164] Based on the similarity score, at least one historical path that best matches the query sequence is selected as a candidate path.

[0165] S504. Weighted generation of operating status indicators:

[0166] Based on the flame location, flame intensity, temperature information and other indicators associated with the candidate paths, and combined with their assigned weights, the estimated values ​​of the corresponding operating status indicators within the current time window are generated through weighted averaging or weighted fusion. If multiple candidate paths are selected, the estimated values ​​are the weighted comprehensive results of the corresponding values ​​of each candidate path.

[0167] Through the above steps, the system can quantitatively identify and estimate the current combustion state based on historical combustion evolution experience, providing accurate and reliable real-time status input for subsequent combustion health assessment.

[0168] S600: The diagnostic output module 260 determines the overall ignition and combustion status of the combustion chamber based on a preset combustion chamber combustion health status assessment strategy library, assesses the probability of potential fault risks, and outputs a combustion diagnostic report including the location coordinates of the fault area and the quantitative analysis results of combustion turbulence characteristics. As one implementation, step S600 may include the following steps:

[0169] S601. Construction and invocation of the combustion health status assessment strategy library:

[0170] A combustion health status assessment strategy library is pre-established, which contains multiple assessment rules and logical models. The assessment strategy library is constructed based on at least one of expert knowledge base, historical fault data, and combustion simulation data, and covers the feature criteria and thresholds corresponding to various states such as normal combustion, unstable combustion, partial flameout, thermoacoustic oscillation, and ignition failure. Each assessment rule in the strategy library is associated with one or more operating status indicators and includes corresponding risk assessment weights and warning levels.

[0171] S602. Determination of overall combustion ignition state and combustion state:

[0172] The operational status indicators are input into the combustion health status assessment strategy library, and a step-by-step judgment is performed according to preset rules; the judgment process includes:

[0173] (1) Ignition status determination: Based on the ignition delay time, initial flame propagation speed and initial flame intensity index, determine whether ignition is successful;

[0174] (2) Combustion state determination: Based on the stability of the flame centroid position, the fluctuation characteristics of the heat release rate, and the oscillation frequency and amplitude in the time dynamic characteristic index, the combustion is determined to be in a stable combustion, slow-vibration combustion or violent oscillation state.

[0175] S603. Potential Fault Risk Probability Assessment and Location: When the judgment result indicates the presence of combustion abnormalities, the following steps may be performed, including but not limited to:

[0176] (1) Risk probability calculation: Based on the degree of deviation between the abnormal indicator and the corresponding threshold, and combined with the contribution weight of the indicator in historical failures, the probability of occurrence of each potential failure mode is evaluated.

[0177] (2) Fault area location: Based on the spatial distribution anomalies of multi-channel spontaneous spectral intensity data and the abrupt change points of the flame centroid trajectory, combined with the location coordinates of the fiber optic sensor deployment, the spatial coordinates or area identifiers of the abnormal area are obtained.

[0178] S604, Quantitative Analysis of Combustion Turbulence Characteristics:

[0179] Based on the time dynamic characteristic indicators in the feature vector set, especially the fluctuation distribution of flame propagation speed and the pulsating spectrum characteristics of light intensity signal, the combustion turbulence characteristics are analyzed to characterize the strength and stability of the flow-combustion coupling effect in the combustion chamber.

[0180] S605. Generation and output of combustion diagnostic reports:

[0181] Integrating the aforementioned step-by-step judgment results, fault risk probability assessment, fault area location, and quantitative analysis of combustion turbulence characteristics, the diagnostic output module 260 generates and outputs a structured combustion diagnostic report; the report content includes, but is not limited to, the following:

[0182] (1) Classification and confidence level of overall combustion state;

[0183] (2) Potential fault types, risk probabilities, and location information;

[0184] (3) Comparative analysis of key combustion parameters with standard operating conditions;

[0185] (4) Quantitative indicators of turbulence characteristics and their impact on combustion stability assessment;

[0186] (5) Maintenance suggestions or control adjustment prompts.

[0187] Through the above steps, the system completes the entire combustion diagnosis process from signal acquisition to status assessment and fault early warning, realizing comprehensive real-time monitoring and intelligent health management of the engine combustion process.

[0188] This invention is well-conceived and enables non-contact, multimodal, distributed real-time monitoring of key parameters in the combustion process. Through spatiotemporal feature extraction and historical path matching, it significantly improves the accuracy of combustion status identification and fault early warning capabilities. It can acquire the three-dimensional structure, thermochemical information, and dynamic characteristics of the flame without disturbing the flame, providing reliable data support for the design optimization, condition assessment, and health management of the engine combustion chamber.

[0189] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A combustion diagnosis and identification system for engine combustion chambers based on multimodal fiber optic sensing and monitoring, characterized in that: The diagnostic identification system includes a signal acquisition module (210), a signal processing module (220), a feature extraction module (230), a parameter calculation module (240), a state matching module (250), and a diagnostic output module (260). The signal acquisition module (210) is used to acquire continuous sensing signals collected by distributed optical fiber sensors deployed around the combustion chamber igniter. The signal processing module (220) is used to filter and perform photoelectric conversion on the continuous sensing signal to obtain spontaneous spectral intensity data; The feature extraction module (230) is used to extract spatiotemporal features from the spontaneous spectral intensity data and generate a feature vector set; The parameter calculation module (240) is used to calculate combustion characteristic parameters based on the feature vector set; The state matching module (250) is used to perform similarity matching between the feature vector set and the historical combustion evolution path to generate operating state indicators; The diagnostic output module (260) is used to output a combustion diagnostic report based on the combustion health status assessment strategy library.

2. An engine combustion diagnosis and identification system and method based on fiber optic sensing monitoring, based on the engine combustion diagnosis and identification system based on multi-mode fiber optic sensing monitoring in claim 1, characterized in that, Specifically, the following steps are included: 1) Distributed fiber optic sensors are installed at specific locations of the igniter in the combustion chamber, and the signal acquisition module (210) synchronously acquires continuous sensing signals during the combustion process through multiple channels. 2) The signal processing module (220) performs filtering and photoelectric conversion processing on the collected continuous sensing signal to obtain spontaneous spectral intensity data within a specific wavelength range; 3) The feature extraction module (230) extracts spatiotemporal features from the acquired spontaneous spectral intensity data and generates a feature vector set characterizing the micro-variation features of the multi-physics field in the combustion chamber; 4) Based on the generated feature vector set, the parameter calculation module (240) calculates at least one of the following combustion characteristic parameters: flame centroid position, heat release rate, ignition delay time, flame propagation speed and its probability density function; 5) Call the set of combustion evolution paths in the historical database, and the state matching module (250) performs similarity matching between the feature vector set and the historical paths to generate an operating status index containing flame position, flame intensity, temperature information and time dynamic characteristic index. 6) The diagnostic output module (260) determines the overall ignition and combustion status of the combustion chamber based on the preset combustion chamber combustion health status assessment strategy library, assesses the probability of potential fault risks, and outputs a combustion diagnostic report containing the fault area location coordinates and the quantitative analysis results of combustion turbulence characteristics.

3. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 2, characterized in that, Step 1) specifically includes the following steps: 1.1) Determine the deployment locations and channel configurations of the distributed fiber optic sensors. The sensor array probes are deployed within a predefined monitoring area around the combustion chamber igniter to achieve continuous sensing of the spatial distribution of the combustion process in a multi-point, multi-channel manner. Each fiber optic sensor channel corresponds to an independent sensing unit, or a fiber optic coupler is used to split the signal collected by a single fiber optic sensor into multiple channels for collecting self-emitting signals within a specific wavelength or band range. The fiber optic coupler is used to split the signal collected by a single fiber optic sensor into multiple channels, and a synchronous triggering mechanism is used between the channels to ensure the spatiotemporal alignment of the signal acquisition, so as to facilitate subsequent multiphysics coupling analysis. 1.2) Acquire multi-channel synchronous continuous sensor signal stream During engine operation, the flame self-luminous signal generated in the combustion chamber is captured in real time by fiber optic sensors arranged at specific points to form a raw sensing signal containing time series information; the signal acquisition module (210) synchronously and continuously acquires the self-luminous signal of each fiber optic sensor channel at a preset sampling frequency to obtain a multi-channel synchronous continuous sensing signal stream, ensuring that rapid dynamic changes that may occur during combustion can be captured. 1.3) Signal preprocessing and quality verification The obtained multi-channel synchronous continuous sensing signal stream is subjected to preliminary signal preprocessing and quality verification.

4. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 3, characterized in that, The specific process of step 1.3) is as follows: check the validity of the self-emitting signals of each fiber optic sensor channel, identify and mark the signal abnormality segments that may be caused by fiber optic sensor failure or external interference; calibrate and correct the original sensing signals formed in step 1.2) to eliminate noise components such as zero drift introduced by ambient light or fiber optic sensor background noise; record the trigger and acquisition timestamps of each channel in the multi-channel signal bundled by a fiber optic coupler and the corresponding engine operating condition parameters, so as to lay the data foundation for subsequent information processing.

5. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 4, characterized in that, The specific process for calibrating and correcting the original sensing signal formed in step 1.2) is as follows: 1.2.1) Calibration Using a standard cold light source: Install filters to perform photoelectric conversion on the signals collected by all channels of the fiber optic sensor probe, and calibrate the gain and voltage response of the components so that each channel outputs the same voltage value under the same light intensity. Check fiber optic status: If the fiber optic cable corresponding to the fiber optic sensor is found to be damaged during the calibration process, resulting in no signal, then the data of that channel is excluded. The signals of the remaining channels are valid and can be used in subsequent tests. Gain adjustment and recalibration: After the first engine ignition test, the voltage is measured, and the gain of the photomultiplier tube is adjusted according to the voltage value. After the gain is adjusted, the voltage response is recalibrated using a standard cold light source to ensure signal quality. 1.2.2) Signal background correction during the experiment A cold flow test was performed before each ignition test, and the background signal was recorded. During data processing, the median value of the cold flow test data of each channel was taken as the background signal, and the self-ignition signal was zero-drift corrected to eliminate the inherent noise of the system. 1.2.3) Signal normalization processing The maximum value of the data after zero drift correction for each channel is taken as the reference value, and the signal is processed to be dimensionless, so that the signal range is normalized to [0, 1]. The normalized data is used for subsequent calculations of ignition delay, heat release rate, equivalence ratio, flame centroid, and pulsation velocity.

6. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 2, characterized in that, Step 2) After acquiring spontaneous spectral intensity data within a specific wavelength range, it is also necessary to extract spontaneous spectral temporal features. The specific process is as follows: calculate the temporal statistical features of the spontaneous spectral intensity data within a specific wavelength range acquired by each channel, and extract temporal waveform features from the spontaneous spectral intensity data.

7. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 2, characterized in that, Step 3) specifically includes the following steps: 3.1) Extraction of Multi-channel Spatial Distribution Features Based on spontaneous spectral intensity data acquired synchronously through multiple channels, a two-dimensional or three-dimensional light intensity distribution map of the combustion chamber monitoring area is constructed, and spatial distribution features are extracted from the light intensity distribution map. 3.2) Spatiotemporal joint feature fusion The extracted temporal and spatial features are fused to generate a feature vector set; The specific process is as follows: the temporal features of each channel are spliced ​​together in channel order and fused with the spatial distribution features. The fused feature vectors are associated with the corresponding timestamps and operating condition labels and arranged in time series to form a feature vector set representing the multi-physics field of the combustion chamber.

8. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 7, characterized in that, Step 4) specifically includes the following steps: 4.1) Calculation of the flame centroid position The fiber optic sensor positions are used to define the flame centroid and calculate flame pulsation velocity. In a Cartesian coordinate system, the igniter position is defined as the zero point. The flame centroid position is calculated by weighting the sensor placement positions with dimensionless light intensity; the weights are the normalized light intensity values ​​at each sensor location. The formula for calculating the flame centroid position is as follows: ； In the above formula, C represents the coordinates of the flame's center of mass; I i P is the normalized light intensity value at the i-th sensing point; i is the spatial coordinate of the point; N is the number of effective sensing points; the flame centroid position is used to characterize the overall position of the flame in the combustion chamber and its dynamic offset; 4.2) Estimation of heat release rate Based on the spontaneous spectral intensity data, combined with the pre-calibrated correspondence between wavelength and heat release rate, the local heat release rate is characterized by light signals of a specific wavelength, and the intensity of the 300nm light signal corresponding to the hydroxyl main peak OH*(0,0) is used as the heat release rate proxy to sense the heat release. 4.3) Determination of ignition delay time In the feature vector set of the fused time series, the starting time point corresponding to the ignition trigger signal is identified, as well as the time point from the starting time point to the first time when the flame intensity reaches the preset ratio of steady-state intensity, which is the ignition delay time; then the mean and variance of the ignition delay time of multiple consecutive ignition processes are statistically analyzed. 4.4) Calculation of flame propagation speed Based on the time-varying sequence of the flame centroid position, the centroid's moving velocity is calculated as the macroscopic flame propagation velocity; or, based on the time delay estimation of multi-channel light intensity signals, the propagation time of the flame front between adjacent sensors is calculated through cross-correlation analysis, and the local flame propagation velocity is calculated by combining the sensor spacing. The flame propagation velocity is defined as the derivative of the flame centroid position change with time. ； In the above formula, d ij Let Δt be the distance between sensors i and j. ij Let i be the time delay for the flame front to propagate from i to j; 4.5) Probability density function estimation For time series data of flame propagation speed, calculate its probability density function to characterize the statistical distribution characteristics, stability and abnormal fluctuations of the probability density. 4.6) Equivalent ratio Based on the characteristic wavelength radiation intensity ratio in the spontaneous spectral intensity data, and / or based on at least one of the combustion characteristic parameters as the equivalence ratio proxy, the local equivalence ratio is sensed; combined with a pre-established equivalence ratio and feature correlation model or calibration function, the local or overall equivalence ratio distribution and its changing trend over time within the combustion area are calculated in real time.

9. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 2, characterized in that, Step 5) specifically includes the following steps: 5.1) Construction and index generation of the combustion evolution path database A historical database of combustion evolution paths is pre-established, which stores multiple combustion state evolution paths recorded under different operating conditions. Each path consists of a sequence of feature vectors arranged in chronological order, and is labeled with its operating parameters, combustion state label, flame position, flame intensity, temperature information, and actual observed or high-precision measured values ​​of time dynamic characteristic index. A multi-dimensional index structure based on key features is established for the set of combustion evolution paths in the historical database. The key features include at least one of the following: average light intensity, mean flame centroid coordinates, heat release rate, and mean flame propagation speed. 5.2) Similarity matching between real-time feature vectors and historical paths The feature vector sequence within the current time window is used as the query sequence, and the similarity is calculated with each combustion evolution path in the historical database. The difference in query sequence length is considered during the calculation process, and the matching distance or similarity score between the query sequence and each historical path is output. 5.3) Candidate Path Selection and Weight Allocation Based on the similarity score, at least one historical path that best matches the query sequence is selected as a candidate path; 5.4) Weighted generation of operational status indicators: Based on the flame location index, flame intensity index, and temperature information index associated with the candidate path, and combined with their assigned weights, an estimated value of the corresponding operating status index within the current time window is generated by weighted averaging. If multiple candidate paths are selected, the estimated value is a weighted sum of the values ​​corresponding to each candidate path.

10. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 2, characterized in that, Step 6) specifically includes the following steps: 6.1) Construction and Invocation of the Combustion Health Status Assessment Strategy Library A combustion health status assessment strategy library is pre-established, which contains multiple assessment rules and logical models. The assessment strategy library is constructed based on at least one of expert knowledge base, historical fault data, and combustion simulation data, and covers the feature criteria and thresholds corresponding to various states such as normal combustion, unstable combustion, partial flameout, thermoacoustic oscillation, and ignition failure. Each assessment rule in the assessment strategy library is associated with one or more operating status indicators and includes corresponding risk assessment weights and warning levels. 6.2) Determination of overall combustion ignition state and combustion state: The operational status indicators are input into the combustion health status assessment strategy library and judged step by step according to preset rules. 6.3) Potential Failure Risk Probability Assessment and Location When the determination result indicates that there is an abnormal combustion, the following steps may be performed, including but not limited to: 6.3.1) Risk probability calculation: Based on the degree of deviation between the abnormal indicator and the corresponding threshold, and combined with the contribution weight of the indicator in historical failures, assess the probability of occurrence of each potential failure mode. 6.3.2) Fault area location: Based on the spatial distribution anomalies of multi-channel spontaneous spectral intensity data and the abrupt change points of the flame centroid trajectory, combined with the location coordinates of the fiber optic sensor deployment, the spatial coordinates or area identifiers of the abnormal area are obtained. 6.4) Quantitative Analysis of Combustion Turbulence Characteristics Based on the time dynamic characteristic indicators in the feature vector set, especially the fluctuation distribution of flame propagation speed and the pulsating spectrum characteristics of light intensity signal, the combustion turbulence characteristics are analyzed to characterize the strength and stability of the coupling effect between flow and combustion in the combustion chamber. 6.5) Generation and Output of Combustion Diagnostic Reports By integrating the aforementioned step-by-step judgment results, fault risk probability assessment, fault area location and combustion turbulence characteristic quantitative analysis, the diagnostic output module (260) generates and outputs a structured combustion diagnostic report.

11. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 10, characterized in that, The process of step 6.2) involving step-by-step determination based on preset rules includes: 6.2.1) Ignition status determination: Based on the ignition delay time, initial flame propagation speed and initial flame intensity, determine whether ignition is successful; 6.2.2) Combustion state determination: Based on the stability of the flame centroid position, the fluctuation characteristics of the heat release rate, and the oscillation frequency and amplitude in the time dynamic characteristic index, the combustion is determined to be in a stable combustion, slow-vibration combustion, or violent oscillation state.

12. The engine combustion diagnostic and identification system and method based on fiber optic sensing monitoring according to claim 10, characterized in that, The structured combustion diagnostic report in step 6.5) includes, but is not limited to: 6.5.1) Classification and confidence level of overall combustion status; 6.5.2) Potential fault types, risk probabilities, and location information; 6.5.3) Comparative analysis of key combustion parameters with standard operating conditions; 6.5.4) Quantitative indicators of turbulence characteristics and their impact on combustion stability assessment; 6.5.5) Maintenance suggestions or control adjustment prompts.

Images