A method for measuring the non-uniformity of the spray distribution of a fuel nozzle using optical measurement techniques

Abstract

The application discloses a kind of optical measurement technology's fuel nozzle atomization unevenness measurement method, comprising the following steps: S1, fuel nozzle atomization image parameter data is collected by optical detection, and fuel nozzle environment is obtained by sensor;S2, the fuel nozzle atomization image parameter data collected is optimized and handled;S3, the fuel nozzle atomization image parameter data after processing is analyzed in depth, and the quantization index of unevenness therein is extracted;S4, the quantization index of unevenness extracted is compared with traditional 12 circumferential method, and verifies statistical equivalence;S5, error compensation mapping table is established, and the accuracy of fuel nozzle atomization unevenness is detected.The application mainly solves the problems of existing measurement method (traditional 12 circumferential method) such as low anti-interference ability, low adaptability and low measurement efficiency.

Description

Technical Field

[0001] This invention relates to the field of optical measurement technology, specifically to a method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology. Background Technology

[0002] Optical measurement is a high-tech application combining optoelectronic technology and mechanical measurement. Utilizing computer technology, it enables rapid and accurate measurements, facilitating recording, storage, printing, and retrieval. Optical measurement is primarily used in modern industrial inspection, checking the form and position tolerances and numerical aperture of products to ensure they meet specifications. Major application areas include metal processing, mold making, plastics, hardware, gears, and mobile phone manufacturing, as well as product development, mold design, mold making, original engraving, rapid prototyping, and circuit testing. Key instruments include 2D measuring machines, tool microscopes, optical image measuring instruments, optical image projectors, 3D measuring machines, coordinate measuring machines, and 3D laser digitizers. In addition, non-contact inspection technologies are used in the machinery manufacturing industry to ensure machined products meet design precision and quality requirements.

[0003] Currently, the measurement of fuel nozzle atomization non-uniformity mainly relies on the 12-circumferential method, which involves collecting droplets through 12 circumferentially distributed sampling cups and then weighing and analyzing them offline.

[0004] Existing technological shortcomings:

[0005] 1. Insufficient capture of dynamic characteristics: The traditional 12-circumferential method relies on static sampling, which cannot monitor the dynamic evolution of the atomization field in real time, resulting in a distortion of the assessment of transient non-uniformity.

[0006] 2. Limited data dimensions: Existing optical technologies are mostly limited to single parameter measurement and lack the ability to integrate and analyze multiple parameters (concentration, particle size, velocity), making it difficult to characterize the complex manifestations of non-uniformity.

[0007] 3. Susceptible to environmental interference: Measurements are easily affected by environmental factors such as vibration, temperature gradients, and background light. Existing technologies lack dynamic compensation mechanisms, resulting in a significant decrease in measurement accuracy under complex working conditions.

[0008] 4. Lack of universality: Existing methods are mostly designed for specific nozzle structures and lack a unified measurement framework and evaluation standard applicable to multiple types of nozzles. Summary of the Invention

[0009] To address the above problems, this invention provides a method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology, comprising the following steps:

[0010] S1. The fuel nozzle atomization image parameter data are acquired through optical detection, and the fuel nozzle environmental conditions are obtained through sensors; the sensors specifically include: a polarization light sensor, an infrared sensor, and a pressure sensor.

[0011] S2. Optimize the collected fuel nozzle atomization image parameter data;

[0012] S3. Perform in-depth analysis on the processed fuel nozzle atomization image parameter data and extract the quantitative index of non-uniformity;

[0013] S4. Compare the extracted quantitative indicators of non-uniformity with the traditional 12-week azimuth method and verify the statistical equivalence.

[0014] S5. Establish an error compensation mapping table to accurately detect the unevenness of fuel nozzle atomization.

[0015] Furthermore, the optical detection in step S1 specifically includes: high-speed imaging, laser diffraction, and particle image velocimetry; the fuel nozzle atomization image parameter data specifically includes: fog field concentration, particle size, velocity distribution, imaging resolution, light source parameters, and sampling frequency.

[0016] Furthermore, step S3 specifically includes the following sub-steps:

[0017] S31. Perform in-depth analysis of the processed fuel nozzle atomization image parameter data:

[0018] S311. Spectral analysis based on Raman scattering effect; S312. Segmentation of spray particles using edge detection; S313. Removal of noise from fuel nozzle atomization image parameter data through morphological operations; S314. Statistical analysis of equivalent diameter and spatial distribution density of spray particles; S315. Analysis of the correlation between environmental parameters and fuel spray uniformity;

[0019] S32. Extract non-uniformity metrics from the data after in-depth analysis;

[0020] S33. Correct the errors of the extracted non-uniformity quantification index.

[0021] Further, step S32 specifically involves: calculating the spray uniformity using the coefficient of variation, and then performing a correlation analysis between the calculated particle size dispersion of the spray and the obtained coefficient of variation.

[0022] Furthermore, the formula for calculating the coefficient of variation is as follows: In the formula, The standard deviation of fuel concentration within the spray area; μ c Indicates the average fuel concentration;

[0023] The formula for calculating the particle size dispersion of the spray is: In the formula, Dv90, Dv50, and Dv10 are the particle size distributions of the sprayed particles that were actually measured.

[0024] Furthermore, a coefficient of variation of less than 5% indicates uniform spraying, while a coefficient of variation of more than 10% indicates non-uniform spraying.

[0025] Furthermore, step S33 specifically includes the following sub-steps:

[0026] S331. Correction of non-uniformity metrics based on DNN model and genetic algorithm:

[0027] The directional gradient histogram of the spray image, the current temperature and pressure parameters, and their original coefficients of variation are input into the DNN model for convergence. The converged coefficients of variation are output, and multiple sets of DNN weight matrices are randomly generated. Based on their fitness function, excellent individuals are selected for cross mutation.

[0028] S332. Reduce non-uniformity metrics through polarized light emission and multi-sensor fusion.

[0029] Furthermore, the formula for calculating the fitness function is as follows: Where MSE represents the mean square error between the corrected CV value and the true value, the smaller the MSE, the larger f is.

[0030] Further, step S332 specifically includes: detecting changes in light intensity using the birefringence effect of fuel molecules, then measuring phase delay, fuel temperature, and injection back pressure using a polarized light sensor, an infrared sensor, and a pressure sensor respectively, and fusing the detected data using Kalman filtering combined with time-series data;

[0031] The formula for calculating the change in light intensity is: δ represents the phase delay. This indicates the intensity of the incident light.

[0032] Further, step S5 specifically involves: constructing an error compensation mapping table based on ambient temperature, injection back pressure, and their corresponding compensation coefficients; collecting the ambient temperature and fuel pressure of the fuel injector in real time; obtaining the corresponding compensation coefficients through the error compensation mapping table; correcting the original spectral peak intensity; and then using the corrected data to calculate the concentration distribution to obtain the fuel injector atomization non-uniformity.

[0033] This invention provides a method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology, which has the following advantages:

[0034] This invention is based on optical measurement technology for non-uniformity of distribution. It achieves accurate quantification, dynamic monitoring and real-time optimization of non-uniformity of the atomization field through a high spatiotemporal resolution optical measurement system, a multi-dimensional fusion model and a feedback correction mechanism. It mainly solves the problems of poor anti-interference ability, low adaptability and low measurement efficiency of existing measurement methods (traditional 12-circumferential method).

[0035] This invention is adaptable to different nozzle structures, provides real-time measurement and visual feedback, reduces trial and error costs, enables dynamic adjustment of process parameters, and improves product consistency; it achieves non-contact, high spatiotemporal resolution full-field measurement, meeting the needs for rapid, accurate, and multi-dimensional atomization performance evaluation. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0037] Figure 1 The method flowchart provided by the present invention. Detailed Implementation

[0038] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0039] The following detailed description of the implementation method of the present invention is in conjunction with the accompanying drawings. The description is only a partial embodiment and not all embodiments. For clarity, representations and descriptions unrelated to the present invention are omitted in the drawings and description.

[0040] To provide a clearer understanding of the technical features, objectives, and beneficial effects of this invention, the following detailed description of the technical solution is provided. Obviously, the described embodiments are only a portion of the embodiments of this invention, not all of them, and should not be construed as limiting the scope of implementation of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the protection scope of this invention.

[0041] This invention, based on optical measurement technology for uneven distribution, delves into its detection principles and manifestations, applying it to the research of performance indicators for uneven distribution in fuel nozzle atomization. Through in-depth analysis of data obtained from optical measurements, it accurately extracts quantitative indicators of unevenness and compares them with the traditional 12-circumferential method to verify the accuracy of optical measurement technology for uneven distribution. Simultaneously, it studies methods for calibrating and correcting uneven distribution, conducting experimental verification and application expansion for different structures. The invention also analyzes the influence of complex environmental factors on optical measurement and unevenness results, evaluating the applicability of different nozzle structures to the measurement of uneven distribution, thus achieving accurate evaluation of the measurement results.

[0042] The 12-circumferential method for nozzle atomization is a test method for measuring the fuel distribution non-uniformity of fuel nozzles. It analyzes the fuel distribution characteristics of the nozzle by collecting droplets in circumferentially divided areas. Twelve equally divided collection areas are evenly arranged circumferentially on the spray cone, and the amount of fuel collected in each area within the same time period is recorded. The fuel distribution non-uniformity of the nozzle is determined by calculating the difference in fuel quantity between the areas. This method uses the spray cone angle measured by optical or mechanical probe methods as an auxiliary parameter.

[0043] This invention, based on optical measurement technology for uneven distribution, delves into its detection principles and manifestations, applying it to the research of performance indicators for uneven distribution in fuel nozzle atomization. Through in-depth analysis of the data obtained from optical measurements, quantitative indicators of unevenness are accurately extracted and compared with the traditional 12-circumferential method. Bland-Altman consistency analysis verifies the statistical equivalence between the optical measurement results and the 12-circumferential method. An error compensation mapping table is established to verify the accuracy of the optical measurement technology for uneven distribution.

[0044] This invention also investigates methods for calibrating and correcting distribution non-uniformity, conducting experimental verification and application expansion for nozzles with different structures. Furthermore, it analyzes the influence of complex environmental factors on optical measurements and non-uniformity results, evaluates the applicability of different nozzle structures for measuring distribution non-uniformity, and achieves accurate evaluation of measurement results.

[0045] like Figure 1 As shown, the present invention provides a method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology, comprising the following steps:

[0046] S1. The fuel nozzle atomization image parameter data is acquired through optical detection, and the environmental conditions of the fuel nozzle are obtained through sensors; specifically, the sensors include: a polarization light sensor, an infrared sensor, and a pressure sensor. The optical detection specifically includes: high-speed imaging, laser diffraction, and particle image velocimetry; the fuel nozzle atomization image parameter data specifically includes: fog field concentration, particle size, velocity distribution, imaging resolution, light source parameters, and sampling frequency.

[0047] S2. Optimize the collected fuel nozzle atomization image parameter data.

[0048] S3. Perform in-depth analysis on the processed fuel nozzle atomization image parameter data to extract the quantitative index of non-uniformity:

[0049] S31. A combined system of a high-resolution spectrometer (such as Ocean Insight HR4000, wavelength range 400-1100nm, spectral resolution 0.1nm) and a high-speed industrial camera (such as Basler ace acA1920 - 150um, frame rate 1000fps, pixel 1920×1080) is used to perform in-depth analysis of the processed fuel nozzle atomization image parameter data:

[0050] S311. Spectroscopic analysis based on Raman scattering effect using a spectrometer (laser source wavelength 532nm, power 50mW): Based on the Raman scattering effect, fuel molecules (such as alkane CH bonds, aromatic hydrocarbon C=C bonds) will produce characteristic shifts (Δν) and peak intensities (I) after being excited by a laser. For example, " ( The higher the peak intensity of the "symmetric stretching peak," the higher the fuel concentration in the corresponding region; The presence of "(aromatic hydrocarbon C=C peak)" indicates the presence of aromatic hydrocarbon components.

[0051] S312. Use edge detection (Canny operator) to segment spray particles; S313. Remove noise from fuel nozzle atomization image parameter data through morphological operations (erosion + dilation); S314. Statistically analyze the equivalent diameter and spatial distribution density of spray particles. The larger the diameter deviation, the higher the spray non-uniformity; S315. Analyze the correlation between environmental parameters and fuel spray uniformity, as shown in Table 1 below.

[0052] Table 1. Correlation between Key Parameter Types and Non-uniformity

[0053]

[0054] S32. Extract non-uniformity quantification indicators from the data after depth analysis: calculate the spray uniformity using the coefficient of variation, and then perform correlation analysis between the calculated particle size dispersion of the spray and the obtained coefficient of variation.

[0055] The formula for calculating the coefficient of variation is: In the formula, The standard deviation of fuel concentration within the spray area (calculated by weighting spectral peak intensities, such as...) , (where i is the concentration of the i-th pixel). Indicates the average fuel concentration ( When the coefficient of variation is less than 5%, it is considered uniform spraying; when the coefficient of variation is greater than 10%, it is considered non-uniform spraying. (When the coefficient of variation is greater than 10%, the spray has obvious oil channels and poor distribution uniformity; when the coefficient of variation is less than or equal to 10%, the spray quality is high and the distribution uniformity is good.)

[0056] The particle size distribution (Dv10, Dv50, Dv90) of the sprayed particles was measured using a laser particle size analyzer (Malvern Mastersizer 3000), and the particle size dispersion was calculated. Correlation analysis was performed between the CV value and the Pearson correlation coefficient (r=0.92). The same spray sample was measured 10 times, and the relative standard deviation (RSD) of the CV value was 2.1% (far lower than the industry requirement of 5%), proving that the indicator is stable.

[0057] S33. Error correction was performed on the extracted non-uniformity quantification index. The main errors are shown in Table 2 below:

[0058] Table 2 Main sources of error

[0059] Error type Source Description Impact on measurement results Ambient temperature error The quantum efficiency of the spectrometer detector varies with temperature (sensitivity changes by ±8% when the temperature ranges from 20 to 40°C). A fluctuation of ±10% in spectral peak intensity results in a concentration calculation error of ±8%. fuel pressure error The spray cone angle changes by ±15° when the injector atomization pressure changes from 0.5 to 2 MPa. A change in particle distribution density of ±12% affects the uniformity assessment. Sensor nonlinearity The spectrometer's response curve deviates from linearity under strong light (nonlinear error ±5% when I > 10000 counts). Low values ​​were observed in high-concentration areas.

[0060] S331. Correction of non-uniformity metrics based on DNN model and genetic algorithm:

[0061] The DNN model architecture adopts an "input layer-hidden layer-output layer" structure, with specific parameters as follows: Input layer: 256 nodes (containing "HOG features of the spray image (128-dimensional) + temperature / pressure parameters (16-dimensional) + original CV value (1-dimensional)"); Hidden layer 1: 128 nodes, activation function ReLU (accelerates convergence); Hidden layer 2: 64 nodes, activation function ReLU; Output layer: 1 node, activation function Sigmoid (outputs the corrected CV value, ranging from 0 to 1).

[0062] The directional gradient histogram of the spray image, current temperature and pressure parameters, and their original coefficients of variation are input into a DNN model for convergence using the ReLU activation function. The converged CV value (range 0-1) is output. Multiple sets of DNN weight matrices are then randomly generated (each set contains weights for input layer → hidden layer 1, hidden layer 1 → hidden layer 2, and hidden layer 2 → output layer). Based on the fitness function, high-performing individuals are selected for crossover (two-point crossover, crossover probability 0.8) and mutation (Gaussian mutation, mutation probability 0.1). The fitness function is calculated as follows: Where MSE represents the mean square error between the corrected CV value and the true value, the smaller the MSE, the larger f is.

[0063] S332. Reduce non-uniformity quantification indicators through polarized light emission and multi-sensor fusion: Detect light intensity changes by utilizing the birefringence effect of fuel molecules, and then measure phase delay, fuel temperature, and injection back pressure using polarized light sensors, infrared sensors, and pressure sensors respectively. Kalman filtering is used in conjunction with time-series data to fuse the detected data.

[0064] Utilizing the birefringence effect of fuel molecules, incident linearly polarized light (λ=532nm, polarization direction along the x-axis) becomes elliptically polarized after passing through the spray due to the disordered molecular orientation. The change in light intensity is detected by an analyzer (polarization direction along the y-axis), and the formula is: δ represents the phase delay. The value δ represents the intensity of the incident light. The larger the value δ is, the more disordered the molecular orientation and the more uneven the microstructure of the spray.

[0065] The detected data are fused using Kalman filtering combined with time-series data. The state equation and observation equation are as follows:

[0066] Equations of state: ;in, (The state vector at time k, including phase delay, temperature, and pressure); (State transition matrix, assuming δ changes slowly with T); (Control matrix, u_k is the fuel injection pulse signal); For process noise (covariance matrix) ).

[0067] Observation equation: ;in, (Estimated non-uniformity after fusion at time k); (Observation matrix, δ contributes the most to CV); The observation noise is represented by the covariance matrix R = 0.1.

[0068] S4. The extracted quantitative index of non-uniformity is compared with the traditional 12-axis method, and the statistical equivalence between the optical measurement results and the 12-axis method is verified by Bland-Altman consistency analysis.

[0069] S5. Establish an error compensation mapping table to accurately detect the atomization non-uniformity of fuel injectors: Based on ambient temperature, injection back pressure and their corresponding compensation coefficients, an error compensation mapping table is constructed as shown in Table 3 below. The ambient temperature and fuel pressure of the fuel injector are collected in real time. The corresponding compensation coefficients are obtained through the error compensation mapping table. The original spectral peak intensity is corrected (I = I0 × k). The concentration distribution is then calculated using the corrected data to obtain the atomization non-uniformity of the fuel injectors.

[0070] Table 3 Error Compensation Mapping Table

[0071]

[0072] The compensation coefficient is obtained through "standard sample calibration" (e.g., by measuring fuel spray of known concentration at different T / P values ​​and then back-calculating the k value).

[0073] Comparing the CV values ​​before and after compensation: without compensation, CV = 12.5%, and after compensation, CV_comp = 7.8% (close to the true value of 8.0%), with the error decreasing from ±4.5% to ±0.2%.

[0074] This invention is based on optical measurement technology for non-uniformity of distribution. It achieves accurate quantification, dynamic monitoring and real-time optimization of non-uniformity of the atomization field through a high spatiotemporal resolution optical measurement system, a multi-dimensional fusion model and a feedback correction mechanism. It mainly solves the problems of poor anti-interference ability, low adaptability and low measurement efficiency of existing measurement methods (traditional 12-circumferential method).

[0075] This invention is adaptable to different nozzle structures, provides real-time measurement and visual feedback, reduces trial and error costs, enables dynamic adjustment of process parameters, and improves product consistency; it achieves non-contact, high spatiotemporal resolution full-field measurement, meeting the needs for rapid, accurate, and multi-dimensional atomization performance evaluation.

[0076] This invention employs multi-parameter dynamic fusion measurement technology, achieving for the first time simultaneous high-resolution measurement of concentration, particle size, and velocity field. It constructs a comprehensive non-uniformity index, overcoming the limitations of single-dimensional static measurement and reducing non-uniformity assessment error to <5% (compared to >12% for traditional methods). Utilizing intelligent calibration and real-time feedback correction mechanisms, based on a DNN-based dynamic calibration model and genetic algorithm optimization strategy, it achieves closed-loop control of "measurement-analysis-correction," improving nozzle non-uniformity correction efficiency by over 50%. Through polarized light emission technology and multi-sensor fusion compensation, it maintains a measurement error of <3% even under harsh environments with vibration intensity of 5G and background illuminance of 20,000 Lux.

[0077] Establish a cross-technology platform comparison and verification system, and establish a consistency analysis framework for optical measurement and the 12-axis method to solve the problem of mutual recognition between traditional methods and optical data, providing technical basis for the upgrading of industry standards; propose an adjustable optical support design scheme to support rapid adaptation measurement of more types of nozzle structures.

[0078] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology, characterized in that, Includes the following steps: S1. The fuel nozzle atomization image parameter data are acquired through optical detection, and the fuel nozzle environmental conditions are obtained through sensors; the sensors specifically include: a polarization light sensor, an infrared sensor, and a pressure sensor. S2. Optimize the collected fuel nozzle atomization image parameter data; S3. Perform in-depth analysis on the processed fuel nozzle atomization image parameter data and extract the quantitative index of non-uniformity; S4. Compare the extracted quantitative indicators of non-uniformity with the traditional 12-week azimuth method and verify the statistical equivalence. S5. Establish an error compensation mapping table to accurately detect the fuel injector atomization unevenness. Step S3 specifically includes the following sub-steps: S31. Perform in-depth analysis of the processed fuel nozzle atomization image parameter data: S311. Spectral analysis based on Raman scattering effect; S312. Use edge detection to segment spray particles; S313. Remove noise from fuel nozzle atomization image parameter data through morphological operations; S314. Statistical analysis of the equivalent diameter and spatial distribution density of sprayed particles; S315. Analyze the correlation between environmental parameters and fuel spray uniformity; S32. Extract non-uniformity metrics from the data after in-depth analysis; S33. Correct the errors in the extracted non-uniformity quantification index; The specific steps of S32 are as follows: the uniformity of the spray is calculated using the coefficient of variation, and then the correlation analysis is performed between the calculated particle size dispersion of the spray and the obtained coefficient of variation. The formula for calculating the coefficient of variation is: In the formula, This represents the standard deviation of fuel concentration within the spray area; Indicates the average fuel concentration; The formula for calculating the particle size dispersion of the spray is: In the formula, Dv90, Dv50, and Dv10 are the actual measured particle size distributions of the sprayed particles. Step S33 specifically includes the following sub-steps: S331. Correction of non-uniformity metrics based on DNN model and genetic algorithm: The directional gradient histogram of the spray image, the current temperature and pressure parameters, and their original coefficients of variation are input into the DNN model for convergence. The converged coefficients of variation are output, and multiple sets of DNN weight matrices are randomly generated. Based on their fitness function, excellent individuals are selected for cross mutation. S332. Reduce non-uniformity metrics through polarized light emission and multi-sensor fusion; The S332 step specifically includes: detecting light intensity changes using the birefringence effect of fuel molecules, then measuring phase delay, fuel temperature, and injection back pressure using a polarized light sensor, an infrared sensor, and a pressure sensor respectively, and fusing the detected data using Kalman filtering combined with time-series data; The formula for calculating the change in light intensity is: δ represents the phase delay. This indicates the intensity of the incident light.

2. The method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology according to claim 1, characterized in that, The optical detection in step S1 specifically includes: high-speed imaging, laser diffraction, and particle image velocimetry; the fuel nozzle atomization image parameter data specifically includes: fog field concentration, particle size, velocity distribution, imaging resolution, light source parameters, and sampling frequency.

3. The method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology according to claim 1, characterized in that, When the coefficient of variation is less than 5%, it is considered uniform spraying; when the coefficient of variation is greater than 10%, it is considered non-uniform spraying.

4. The method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology according to claim 1, characterized in that, The fitness function is calculated using the following formula: Where MSE represents the mean square error between the corrected CV value and the true value, the smaller the MSE, the larger f is.

5. The method for measuring the non-uniformity of fuel nozzle atomization distribution using optical measurement technology according to claim 1, characterized in that, The S5 step is as follows: an error compensation mapping table is constructed based on ambient temperature, injection back pressure and their corresponding compensation coefficients; the ambient temperature and fuel pressure of the fuel injector are collected in real time; the corresponding compensation coefficients are obtained through the error compensation mapping table; the original spectral peak intensity is corrected; and the concentration distribution is calculated using the corrected data to obtain the fuel injector atomization non-uniformity.

Images