Device and method for measuring ignition delay of solid propellant based on multi-modal data

Abstract

The invention discloses a device and a method for measuring ignition delay of a solid propellant based on multi-modal data. According to the method, laser is used as an ignition source, images, temperature, characteristic spectrum radiation intensity and pressure fluctuation signals in the solid propellant combustion process are synchronously collected, and four independent ignition delay time data are obtained through processing. On the basis, a Dempster-Shafer evidence theory framework is adopted, and nonlinear correlation analysis and fusion are carried out on the multi-mode ignition delay time data. And obtaining correlation coefficients among the data sources through a basic probability distribution function, and finally obtaining the fused ignition delay time with high confidence through an evidence theory synthesis rule. According to the invention, multi-dimensional panoramic observation of the ignition process is realized, a multi-feature criterion with a clear mechanism is provided, the accuracy and robustness of measurement are remarkably improved, a standardized and universal solution is provided, and the system is high in integration level and good in automation degree.

Description

Technical Field

[0001] This invention belongs to the field of solid propellant combustion technology, and relates to a device and method for measuring solid propellant ignition delay based on multimodal data. Background Technology

[0002] The ignition process of solid propellant in the combustion chamber of a rocket engine is a transient process involving strong coupling of multiple physics fields, including heat conduction, pyrolysis, gas-solid two-phase flow, and complex chemical reactions. Precisely defining the ignition timing is a key technical foundation for evaluating engine start-up characteristics, thrust build-up timing, and operational reliability, directly impacting ignition system design, propellant structure safety, and overall performance optimization.

[0003] Current engineering practice and academic research primarily rely on traditional criteria for determining ignition delay time, which involve monitoring a single physical parameter threshold during combustion. These methods stem from different ignition theory models, leading to inconsistent technical standards, scattered measurement results, and poor comparability. The specific limitations of various single-parameter criteria are as follows: 1. Ignition Criteria Based on Visual Images: This method involves recording the moment when a sustained bright spot or characteristic flame pattern first appears on the propellant surface using a high-speed camera. While intuitive, the results heavily rely on the sampling frequency and subjective threshold settings. The discrete nature of image information makes accurately capturing the true ignition moment between two frames extremely difficult, and it is also susceptible to interference from environmental smoke and lens contamination, introducing random errors on the order of milliseconds.

[0004] 2. Ignition Criterion Based on Temperature Measurement: The ignition moment is typically defined as the point at which the propellant surface reaches a specific pyrolysis temperature, as measured by thermocouples or infrared thermometers. While this method is simple in principle, it has a fundamental flaw: the ignition of solid propellants is a dynamic process, not a static temperature point. Simplifying complex unsteady heat transfer and chemical reactions to a single temperature threshold ignores the crucial role of the rate of temperature change, and the thermal decomposition temperature and ignition temperature are not always directly correlated for different propellant formulations.

[0005] 3. Ignition Criterion Based on Spectral Radiation Signals: This method determines ignition based on the moment when the emission intensity of a specific free radical or intermediate product (such as the characteristic spectral line of AlO at 486 nm) first reaches or exceeds a certain percentage (e.g., 10%) of its peak value. This method, based on the gas-phase reaction mechanism, has theoretical advantages. However, its application is severely limited by the signal-to-noise ratio. For propellants with low burning rates and weak flame radiation, the extraction and identification of effective spectral signals become the main technical bottleneck. Furthermore, the selection of characteristic wavelengths, spectrometer calibration, and background radiation subtraction all introduce significant system uncertainties.

[0006] 4. Ignition Criterion Based on Combustion Chamber Pressure Response: This method uses the moment when the intracavity pressure measured by the pressure sensor rises to a certain absolute threshold (e.g., 0.1 MPa) or relative threshold (e.g., 80% of the maximum pressure) as the ignition moment. This method is widely used in engineering, but the pressure signal is a macroscopic, integral representation of the overall thermodynamic state of the combustion chamber, and it lags behind the response to initial surface chemical reactions occurring on a millisecond or even microsecond scale. Using a fixed pressure threshold as the criterion is not applicable to different burning rate propellants, different combustion chamber free volumes, or different ignition energy inputs, lacking universality.

[0007] The core bottleneck of current technology lies in the fact that any single-dimensional physical parameter mentioned above is only an indirect reflection of one aspect or a delayed result in a complex ignition chain reaction, and cannot fully, sensitively, and robustly capture the essential transition process "from solid-phase thermal decomposition to violent gas-phase exothermic reaction." This directly leads to huge differences in the measured ignition delay time of the same sample by different methods, and even the same method in different experiments, resulting in low data confidence and making it difficult to use as reliable design input. In recent years, to overcome the limitations of single-point measurement, multi-dimensional diagnostic technology that simultaneously acquires multi-physics field signals (such as high-speed imaging, multi-band spectroscopy, surface temperature field, and intracavity pressure) has become a research frontier.

[0008] Therefore, this invention proposes a solid propellant ignition delay measurement device and method based on multimodal data to solve the above-mentioned technical problems. Summary of the Invention

[0009] This invention aims to overcome the shortcomings of existing technologies by proposing a device and method for measuring the ignition delay of solid propellants based on multimodal data. Using a laser as the ignition source, images, temperature, characteristic spectral radiation intensity, and pressure fluctuation signals during the solid propellant combustion process are simultaneously acquired, resulting in four independent ignition delay time data sets. Based on this, the Dempster-Shafer evidence theory framework is employed to perform nonlinear correlation analysis and fusion of the aforementioned multimodal ignition delay time data. The correlation coefficients between the data sources are obtained through a basic probability allocation function, and finally, a high-confidence fused ignition delay time is obtained through evidence theory synthesis rules.

[0010] The solid propellant ignition delay measurement device based on multimodal data includes a high-pressure combustion chamber, a high-speed camera, an ignition laser, a total reflection mirror, a dichroic mirror, an optical fiber beam splitter, a photomultiplier tube with a filter, a three-color thermometer, a spectrometer, a pressure sensor, a flow meter, a gas cylinder, an inlet and outlet pipeline, a computer, and a synchronization controller.

[0011] Optical observation windows are provided around the four sides of the high-pressure combustion chamber and at the center of the top, and the top of the chamber is a detachable structure; during use, the solid propellant is placed at the center of the bottom of the high-pressure combustion chamber. A high-speed camera is installed outside any optical observation window on the side wall of the high-pressure combustion chamber, and the lens of the high-speed camera is coaxially set with the window; the high-speed camera is used to acquire sequential images of the entire combustion process of solid propellant.

[0012] A dichroic mirror and a fiber optic beam splitter are installed outside the optical observation window at the center of the top of the high-pressure combustion chamber. The receiving end of the fiber optic beam splitter, the dichroic mirror, and the window are coaxially arranged. The ignition laser, serving as the ignition source, is fixed outside the combustion chamber. The ignition laser and the dichroic mirror form a coaxial ignition optical path through a total reflection mirror. The light signal radiated during combustion is transmitted through the dichroic mirror and enters the receiving end of the fiber optic beam splitter.

[0013] The fiber optic beam splitter has three output interfaces, which are respectively connected to a photomultiplier tube with filters, a three-color thermometer, and a spectrometer via optical fibers to realize the splitting and transmission of optical signals. The photomultiplier tube with filters and the spectrometer are fixed outside the combustion chamber and are used to selectively collect the temporal variation signals of the radiation intensity of the transition spectral lines of the intermediate products characteristic of solid propellant combustion. The three-color thermometer is fixed outside the combustion chamber and is used for non-contact real-time monitoring of the temperature spatiotemporal distribution in the combustion reaction zone. A pressure sensor for monitoring pressure fluctuations in the sealed combustion chamber is also installed on the side wall of the high-pressure combustion chamber, as well as an inlet pipe for connecting to the gas cylinder and an exhaust pipe that serves as the outlet of the high-pressure combustion chamber. A flow meter for accurately adjusting and measuring the flow rate of oxidizing gas input into the combustion chamber is installed on the inlet pipe. Solenoid valves for controlling shut-off are installed on both the inlet and exhaust pipes.

[0014] The high-speed camera, ignition laser, photomultiplier tube with filter, three-color thermometer, spectrometer, barometer, flow meter, and solenoid valve are all connected to the computer and synchronization controller via synchronization signal lines.

[0015] The computer works in conjunction with the synchronization controller. The synchronization controller sends synchronization trigger signals to the high-speed camera, ignition laser, photomultiplier tube with filter, three-color thermometer, spectrometer, barometer, and flow meter. The computer runs control software to remotely control each instrument, synchronously acquire, store, and subsequently analyze and process multi-channel data.

[0016] Furthermore, the high-pressure combustion chamber includes a main cavity, a top cover, and a base. Based on the simulation requirements of the solid rocket engine combustion environment, the main cavity is a hollow cylinder. The top cover and the base are fastened to the main cavity by a pressure-checked bolt and nut assembly, wherein the mating surface of the top cover and the main cavity is provided with an annular groove for embedding a sealing ring to ensure high-pressure sealing.

[0017] Furthermore, each optical observation window includes a window flange cover, a window flange, high-transmittance glass, a sealing ring, and a sealing gasket. The sealing gasket is placed in the groove of the window flange cover, the high-transmittance glass is placed in the groove, and then another sealing gasket is placed on top. The window flange and the window flange cover are connected by threads and the sealing ring is pressed together to form a window sealing unit. Finally, the entire unit is placed into the optical observation window, aligned with the through hole, and fixed with bolts and nuts to achieve high-pressure sealing and optical observation.

[0018] Furthermore, the laser beam emitted by the ignition laser passes through a coaxial ignition optical path, is focused through the upper cover window, and forms a high-energy-density light spot on the surface of the solid propellant sample placed in the combustion chamber to initiate ignition.

[0019] Furthermore, the dichroic mirror is fixed outside the optical observation window at the top center by a mirror frame, and the fiber optic beam splitter is fixed to the mirror frame of the dichroic mirror by optical threads.

[0020] Furthermore, the raw temperature time-series data obtained by the three-color colorimetric thermometer is uploaded after filtering.

[0021] Furthermore, the pressure sensor is connected to the outer wall of the high-pressure combustion chamber via a threaded seal, and the collected pressure-time raw signal is subjected to noise reduction processing.

[0022] Furthermore, a ceramic substrate is provided at the center of the bottom of the high-pressure combustion chamber. In use, solid propellant is placed on the ceramic substrate to support and fix the solid propellant sample to be tested.

[0023] Based on the above-mentioned device, the following method for measuring the ignition delay time of solid propellants is provided, which specifically includes the following steps: First, ignition delay time measurement data based on four different physicochemical principles were obtained simultaneously, as follows: (1) Determination of optical ignition delay time τ1 based on image method: Using a high-speed camera to acquire images of the combustion process, and by analyzing the flame front evolution characteristics in the image sequence, the moment when the first sustainable flame front appears is defined as the optical ignition moment t1, and τ1 = t1 - t0 is calculated, where t0 is the laser ignition trigger moment; (2) Determination of thermodynamic ignition delay time τ2 based on temperature method: Using a three-color colorimetric thermometer to acquire the time series data of the reaction zone temperature, after filtering and differentiation, the moment corresponding to the first inflection point (i.e., dT / dt reaches a local maximum) in the temperature change rate dT / dt curve is defined as the thermodynamic ignition moment t2, and τ2 = t2 - t0 is calculated; (3) Determination of chemical ignition delay time τ3 based on spectroscopic method: Using a spectrometer and a photomultiplier tube with a filter to monitor the intensity of the characteristic spectral lines of the solid propellant to be detected, the moment when the radiation intensity reaches a set threshold is defined as the chemical ignition moment t3, and τ3 = t3 - t0; (4) Determination of gas dynamic ignition delay time τ4 based on gas pressure method: The pressure time curve of the combustion chamber is obtained by pressure sensor. After noise reduction and differentiation, the time corresponding to the first maximum point in the pressure change rate dP / dt curve is defined as the gas dynamic ignition time t4. τ4 = t4 - t0 is calculated.

[0024] Then, the Dempster-Shafer evidence theory framework is used for fusion, as follows: A recognition framework is constructed based on experimental data, and a confidence interval model is built for test data obtained by a single detection method; the confidence interval is defined as follows: A， The dataset of ignition delay times obtained by independent detection methods is constructed by arranging the data in ascending order of time, i.e.: A i =[a i - , a i + ], where a i - and a i + These represent the minimum and maximum values ​​of the ignition delay time data obtained by the independent detection method, respectively. A i of The complement represents the range of inconsistent data: =[-∞, a i - ]∪[a i + [+∞], using A I , A T , AS , A P Let b represent the confidence intervals constructed by sorting the ignition delay time datasets obtained by the image method, temperature method, spectral method, and pressure method in ascending order. The interval number is the number of ignition delay time data measured by each individual method. For example, for ignition delay time data obtained by the image method, the interval number can be represented as b. Ij =[b Ij - , b Ij + ], b Ij - = b Ij + And numerically, it represents the ignition delay time.

[0025] Calculate the interval distance d(a) of test data obtained by different methods i , b j ) and interval similarity S(a i , b j Determine the basic probability assignment (BPA) function for each data source. For any interval number a1 = [a1...] - a1 + ],b1=[b1 - b1 + The distance between interval numbers can be expressed using the following formula: in, , , , .

[0026] , These are the midpoint influence factor and the width influence factor of the interval numbers, respectively. The interval similarity between interval numbers a1 and b1 is S(a1, b1), which can be expressed by the formula: ;in, >0 represents the support coefficient, and d(a1, b1) represents the distance between interval numbers a1 and b1.

[0027] The interval similarity assignment (BPA) obtained from the test data based on different testing methods: m In m Tn m Sn m Pn After normalizing BPA, the dynamic weighting coefficient k is obtained. Iw k Tw k Sw kPw Multimodal data weighting fusion is achieved through dynamic weighting coefficients and weighting formulas, where m In m Tn m Sn m Pn These represent BPA obtained by the image method, temperature method, spectral radiometric method, and pressure method, respectively. Similarly, k Iw k Tw k Sw k Pw These represent the dynamic weighted coefficients of BPA normalized by different methods, where the subscripts are... w and n The experimental group representing a single method.

[0028] The final weighted average yields the ignition delay time t. in = k Iw t In + k Tw t Tn + k Sw t Sn + k Pw t Pn , where t In t Tn t Sn t Pn Let t represent the ignition delay time data of the nth group obtained by the image method, temperature method, spectral irradiance method, and pressure method, respectively. in This represents the ignition delay time of the nth group of four single methods after being fused within one of the confidence intervals.

[0029] The present invention has the following beneficial effects: 1. Achieved multi-dimensional panoramic observation of the ignition process: Through innovative optical path and modular design, this invention has for the first time achieved millisecond-level precise synchronous acquisition of four key physicochemical parameters of the solid propellant ignition process in a compact system: visual morphology, temperature field evolution, emission spectra of key chemical species, and cavity pressure dynamics. This overcomes the one-sidedness caused by single-point measurement and information asynchrony.

[0030] 2. A multi-feature criterion with a clear mechanism is proposed: The four selected criteria (flame front propagation, temperature change rate inflection point, characteristic spectral intensity threshold, and pressure change rate peak) are closely related to the physicochemical nature of different stages of the ignition process (surface flame establishment, thermal equilibrium transition, gas phase reaction activation, and system mechanical response). Compared with the traditional fixed threshold criteria, it has a more solid theoretical basis and physical clarity.

[0031] 3. Significantly improved measurement accuracy and robustness: By introducing the Dempster-Shafer evidence theory—a rigorous mathematical framework for handling uncertainty and conflicting information—this method solves the core challenge of how to scientifically fuse multimodal data. It automatically assesses and balances the reliability of each piece of evidence, effectively suppressing random errors and systematic biases inherent in single methods, ultimately outputting a fusion ignition delay time with higher confidence and lower uncertainty, resulting in a qualitative improvement in data repeatability and reliability.

[0032] 4. Provides a standardized and universal solution: The complete technical process established by this invention, from data acquisition to fusion decision, does not rely on specific propellant formulations or the experience judgment of experimental personnel. It provides an objective, reproducible, and comparable standardized method for measuring the ignition delay time of solid propellants, and has broad engineering applicability and promotional value.

[0033] 5. High system integration and high degree of automation: The entire system is highly integrated and achieves "one-click" operation through central synchronous control, reducing human intervention, ensuring the consistency of experimental conditions, greatly improving testing efficiency and standardization, and providing strong tool support for the research and development and quality assessment of high-performance solid propellants. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the overall structure of the present invention; Figure 2 This is a schematic diagram illustrating the working principle of the device of the present invention; Figure 3 This is a schematic diagram of an explosion in a high-pressure combustion chamber. Figure 4 This is a front sectional view of the internal structure of the high-pressure combustion chamber; Figure 5 This is a schematic diagram of the overall structure of the high-pressure combustion chamber; Figure 6 This is a bar chart showing the error between the test data obtained by the four methods and the average of the fused data from the four confidence intervals. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of the present invention clearer, the implementation methods of the technical solutions of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. 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.

[0036] like Figure 1 and 2As shown, the solid propellant ignition delay time measurement device based on multi-sensor data fusion includes a high-pressure combustion chamber 1, a high-speed camera 2, an ignition laser 3, a total reflection mirror 4, a photomultiplier tube with a narrow band filter 5, a three-color thermometer 6, a dichroic mirror 7, an optical fiber beam splitter 8, a spectrometer 9, a flow meter 10, a gas cylinder 11, an inlet and outlet pipeline 12, a pressure sensor 13, a computer 14, and a synchronization controller 15.

[0037] like Figure 3 As shown, the high-pressure combustion chamber 1 includes a hollow cylindrical main cavity 20, a top cover 19, and a base 23. To enable multi-angle observation, four optical observation windows are symmetrically opened at equal angles (90° intervals) on the side wall of the main cavity 20, and another optical observation window of the same specification is set at the center of the top cover 19.

[0038] The assembly of the optical observation window is performed according to the following steps: First, a sealing gasket is placed into the receiving groove of the optical window flange cover 16, followed by the placement of the high-transmittance glass 17, and then another identical sealing gasket is placed on the upper surface of the glass. Next, the optical window flange 18 is screwed onto the flange cover 16 carrying the glass assembly using threaded connections to form a pre-sealed window unit. After placing a sealing gasket in the assembly groove of the window in the main cavity 20, the entire window unit is embedded, aligned with the bolt through holes, and then secured using M5 bolts, gaskets, and M5 nuts to complete the side wall window installation. The sealing method for the center window of the upper cover 19 is similar; after placing the sealing gasket and the high-transmittance glass 17 in sequence, it is secured by the optical window flange 18.

[0039] like Figure 4 As shown, the sample installation process is as follows: First, the ceramic substrate 25 is cleaned with alcohol to remove surface contamination, and then it is positioned at the geometric center inside the main cavity 20. The solid propellant sample 24 to be tested is precisely placed in the central area of ​​the clean ceramic substrate 25, ensuring that its entirety is within the field of view of the side wall observation window.

[0040] After the internal sample arrangement is completed, a sealing gasket is placed in the annular sealing groove on the upper end face of the main cavity 20. Then, multiple sets of evenly distributed M6 bolts, sealing gaskets and M6 nuts are used to press and seal the upper cover 19 to the main cavity 20.

[0041] After wrapping sealing PTFE tape around the threaded portions of the intake manifold interface 21 and the exhaust manifold interface 22, screw them into the corresponding threaded holes on the side wall of the main cavity 20 and tighten them. The threaded interface at the front end of the pressure sensor 13 is also sealed with PTFE tape before being screwed into the pressure sensor interface 26 (e.g., welded and fixed to the outer wall of the main cavity 20). Figure 5 As shown in the figure, this enables communication with the combustion chamber.

[0042] The fully assembled high-pressure combustion chamber 1 is securely mounted on the optical platform. The dichroic mirror 7 is fixed directly above the top cover 19 using a precision mirror frame, and its optical center is finely adjusted to be coaxial with the center of the high-transmittance glass 17 inside the observation window of the top cover.

[0043] Infrared laser 3 and total reflection mirror 4 are fixed on an optical platform. Through precise adjustment, the optical paths of the laser 3 output port, the center of the reflective surface of the total reflection mirror 4, and the center of the dichroic mirror 7 are made coaxial, forming a complete coaxial ignition optical path. The laser beam finally passes through the observation window of the upper cover and, after being focused, acts on the surface of the solid propellant sample 24.

[0044] The ignition laser 3 is fixed on the platform. Its output laser beam is reflected by the total reflection mirror 4 and then transmitted through the dichroic mirror 7 into the combustion chamber. Its orientation can be determined by its relative position to the reflecting mirror.

[0045] The fiber beam splitter 8 is mounted on the same frame as the fixed dichroic mirror 7 via the optical thread interface on its barrel, and its signal receiving end is finely adjusted to ensure alignment with the aforementioned coaxial optical path in order to efficiently collect the radiation light signal generated during the combustion process.

[0046] The high-speed camera 2, computer 14, synchronization controller 15, photomultiplier tube 5, three-color thermometer 6, and spectrometer 9 are sequentially arranged and fixed on the optical platform. Power supplies for each device are connected. The data acquisition synchronization lines of the high-speed camera 2, photomultiplier tube 5, three-color thermometer 6, and spectrometer 9 are each connected to the corresponding acquisition card of computer 14 and synchronization controller 15. Synchronization controller 15 is responsible for sending a unified and precise timing trigger signal to all data acquisition devices and laser 3. The three output fiber optic interfaces of fiber optic beam splitter 8 are connected to the optical input port of photomultiplier tube 5, the fiber optic probe of three-color thermometer 6, and the incident slit fiber of spectrometer 9, respectively, to complete the splitting and directional transmission of optical signals.

[0047] Gas cylinder 11 is connected to the inlet of flow meter 10 via inlet and outlet pipes 13, and the outlet of flow meter 10 is connected to the intake port 21 of combustion chamber via a pipe. Exhaust port 22 is connected to an external exhaust pipe. Both the inlet and outlet pipes are equipped with electromagnetic control valves (not shown in the figure) for precisely controlling the filling of gas into the combustion chamber, pressure maintenance, and safe discharge after the experiment.

[0048] Application examples Taking aluminum-based solid propellant as an example, the specific implementation process of this invention is illustrated by monitoring the intensity of characteristic spectral lines of AlO radicals using a spectrometer and a photomultiplier tube with a center wavelength of 488 nm and a full width at half maximum (FWHM) of ±10 nm: 1. Sample Preparation and Installation: Cubic specimens with geometric dimensions of 5 mm × 2.5 mm × 2.5 mm (tolerance ±5%) were cut from the same batch of materials. High-precision positioning fixtures were used to assemble the specimens onto the surface of the ceramic substrate, ensuring tight contact at the interface. After calibration, the substrate module carrying the specimens was placed in the core reaction zone of the high-pressure combustion chamber.

[0049] 2. Experimental Environment and Ignition: The experiment was conducted in a constant room temperature (23±1℃) and relative humidity (50±3%) environment. Ignition was performed by the laser ignition module according to preset parameters, simultaneously triggering the multi-dimensional data acquisition system.

[0050] 3. Construction of Confidence Intervals for Different Methods: Under normal pressure (0.1 MPa), 80 sets of ignition experiments with aluminum-based solid propellants were conducted. Eighty sets of ignition delay time data were obtained using four different methods. Through statistical analysis, outliers that significantly deviated from the overall distribution, or those potentially caused by human error, equipment transient failures, or external random interference, were removed. Then, confidence intervals were constructed using the image method, temperature method, spectral method, and pressure method, respectively. In this example, the confidence intervals for the test data using each method are as follows: Image method... A I The ignition delay time data range is [75, 590], temperature method A T The ignition delay time data range is [75, 605], spectral method A S The ignition delay time data range is [135.5, 785.5], pressure method A P The ignition delay time data range is [200, 745], and the unit is ms.

[0051] 4. Data fusion calculation based on DS evidence theory: (1) Interval similarity calculation: A set of measured ignition delay time data was randomly selected as an example: image method 76.4 ms, temperature method 78.0 ms, spectral method 148.6 ms, pressure method 327.6 ms. The midpoint influence factor of the interval number was taken. The influence factor of the width of the interval number is 0.95. The support coefficient is 0.05. The reliability interval for ignition delay time data obtained using the temperature method is 0.1. A TUsing [75,605] as a baseline, the interval number distance d and interval similarity S between the test data obtained by the four different testing methods and the confidence interval of the temperature method were calculated. The calculated interval number distances d for the four methods were 282.94, 281.53, 221.01, and 119.13, respectively; the corresponding interval similarities S were 0.034, 0.0343, 0.0433, and 0.0774, respectively.

[0052] (2) BPA allocation and fusion: Based on the interval similarity S, the basic probability values ​​were assigned to each evidence source, and the BPA values ​​before normalization were 0.966, 0.966, 0.957, and 0.923, respectively. After normalizing the BPA, the corresponding dynamic weighting coefficients were 0.180, 0.181, 0.229, and 0.409, respectively.

[0053] (3) Fusion result: The ignition delay time after fusion is calculated using the weighted formula as: t i = 76.4×0.180+ 78.0×0.181 + 148.6×0.229 + 327.6×0.409 = 196.05 ms.

[0054] This result indicates that, within the fusion framework based on the temperature-based confidence interval, the final ignition delay time obtained from this set of experimental data is 196.05 ms.

[0055] As shown in Table 1, the ignition delay time data obtained by the four different test methods vary due to differences in ignition mechanisms and reference criteria. The differences in ignition delay time data obtained by each of the ten test methods are due to the non-uniform composition of the test samples. Local chemical imbalances in the solid propellant can affect its ignition delay time. If the ignition laser preferentially acts on the oxidizer-rich region of the solid propellant, it may lead to an accelerated local combustion rate and a shorter ignition delay time. If the ignition laser preferentially acts on the fuel-rich region, excessively high fuel concentrations may result in incomplete combustion due to insufficient oxidation, leading to a decrease in local temperature and a prolonged ignition delay time.

[0056] Table 1. Ignition delay data obtained by four different methods in 10 typical groups.

[0057] As shown in Table 2, the ignition delay time data obtained by fusing the four confidence intervals are close to the same value. This indicates that the interval similarity and weighting coefficients obtained from the four feasible intervals are reasonable. The ignition delay time data obtained by fusing each individual confidence interval is compared with the ignition delay time test data obtained by the four methods. Ten sets of fused data are close to the same value, while ten sets of test data from the four methods show some fluctuation due to differences in testing methods and testing errors. Overall, the fused data is more stable.

[0058] Table 2 shows the data obtained by fusing 10 typical ignition delay data obtained from four different methods using the four methods to establish confidence intervals.

[0059] like Figure 6 As shown in the figure, this figure compares the mean performance of traditional test data and fused data in 10 experimental groups (G1-G10). The vertical axis represents the ignition delay time in milliseconds, and the horizontal axis represents the experimental group number. The means of the two sets of data across all experimental groups are very close, indicating that the fusion method does not significantly change the average value of the original test results. Although the means of the fused data and the test data are close to the same, the error bars show that the error bars of the fused data are shorter and the volatility is smaller, indicating that the fusion method improves the stability of the data. This shows that the confidence intervals obtained by the four methods and the BPA and weighting coefficients obtained by the DS evidence fusion theory are effective.

Claims

1. A solid propellant ignition delay measurement device based on multimodal data, characterized in that, It includes a high-pressure combustion chamber, a high-speed camera, an ignition laser, a total reflection mirror, a dichroic mirror, a fiber optic beam splitter, a photomultiplier tube with a filter, a three-color thermometer, a spectrometer, a pressure sensor, a flow meter, gas cylinders, inlet and outlet pipelines, a computer, and a synchronization controller. Optical observation windows are provided around the four sides of the high-pressure combustion chamber and at the center of the top, and the top of the chamber is a detachable structure; during use, the solid propellant is placed at the center of the bottom of the high-pressure combustion chamber. A high-speed camera is installed outside any optical observation window on the side wall of the high-pressure combustion chamber, and the lens of the high-speed camera is coaxially set with the window; the high-speed camera is used to acquire sequential images of the entire combustion process of solid propellant; A dichroic mirror and a fiber optic beam splitter are installed outside the optical observation window at the top center of the high-pressure combustion chamber. The receiving end of the fiber optic beam splitter, the dichroic mirror, and the window are coaxially arranged. The ignition laser serves as the ignition source and is fixed outside the combustion chamber. The ignition laser and the dichroic mirror form a coaxial ignition optical path. The light signal radiated during the combustion process is transmitted through the dichroic mirror and enters the receiving end of the fiber optic beam splitter. The fiber optic beam splitter has three output interfaces, which are respectively connected to a photomultiplier tube with filters, a three-color thermometer, and a spectrometer via optical fibers to realize the splitting and transmission of optical signals. The photomultiplier tube with filters and the spectrometer are fixed outside the combustion chamber and are used to selectively collect the temporal variation signals of the radiation intensity of the transition spectral lines of the intermediate products characteristic of solid propellant combustion. The three-color thermometer is fixed outside the combustion chamber and is used for non-contact real-time monitoring of the temperature spatiotemporal distribution in the combustion reaction zone. A pressure sensor for monitoring pressure fluctuations in the sealed combustion chamber is also installed on the side wall of the high-pressure combustion chamber, as well as an inlet pipe for connecting the gas cylinder and an exhaust pipe that serves as the outlet of the high-pressure combustion chamber; a flow meter for precisely adjusting and measuring the flow rate of oxidizing gas input into the combustion chamber is installed on the inlet pipe; and a solenoid valve for controlling shut-off is installed on both the inlet and exhaust pipes. The high-speed camera, ignition laser, photomultiplier tube with filter, three-color thermometer, spectrometer, barometer, flow meter, and solenoid valve are all connected to the computer and synchronous controller via synchronous signal lines. The computer works in conjunction with the synchronization controller. The synchronization controller sends synchronization trigger signals to the high-speed camera, ignition laser, photomultiplier tube with filter, three-color thermometer, spectrometer, barometer, and flow meter. The computer runs control software to remotely control each instrument, synchronously acquire, store, and subsequently analyze and process multi-channel data.

2. The solid propellant ignition delay measurement device based on multimodal data as described in claim 1, characterized in that, The high-pressure combustion chamber includes a main cavity, an upper cover, and a base; the main cavity is a hollow cylinder; the upper cover and the base are fastened to the main cavity by a pressure-checked bolt and nut assembly, wherein the mating surface of the upper cover and the main cavity is provided with an annular groove for embedding a sealing ring to ensure high-pressure sealing.

3. The solid propellant ignition delay measurement device based on multimodal data as described in claim 1, characterized in that, Each optical observation window includes a window flange cover, a window flange plate, high-transmittance glass, a sealing ring, and a sealing gasket. The sealing gasket is placed in the groove of the window flange cover, the high-transmittance glass is placed in the groove, and then another sealing gasket is placed on top. The window flange plate and the window flange cover are connected by threads and the sealing ring is pressed together to form a window sealing unit. Finally, the entire unit is placed into the optical observation window, aligned with the through hole, and fixed with bolts and nuts to achieve high-pressure sealing and optical observation.

4. The solid propellant ignition delay measurement device based on multimodal data as described in claim 1, characterized in that, The laser beam emitted by the ignition laser passes through a coaxial ignition optical path and is focused through the upper cover window to form a high-energy-density light spot on the surface of the solid propellant sample placed in the combustion chamber, thereby initiating ignition.

5. The solid propellant ignition delay measurement device based on multimodal data as described in claim 1, characterized in that, The dichroic mirror is fixed outside the optical observation window at the top center by a mirror frame, and the fiber optic beam splitter is fixed to the mirror frame of the dichroic mirror by optical threads.

6. The solid propellant ignition delay measurement device based on multimodal data as described in claim 1, characterized in that, The raw temperature time-series data obtained by the three-color colorimetric thermometer is uploaded after filtering.

7. The solid propellant ignition delay measurement device based on multimodal data as described in claim 1, characterized in that, The pressure sensor is connected to the outer wall of the high-pressure combustion chamber via a threaded seal, and the collected pressure-time raw signal is noise-reduced.

8. The solid propellant ignition delay measurement device based on multimodal data as described in claim 1, characterized in that, A ceramic substrate is provided at the center of the bottom of the high-pressure combustion chamber. In use, solid propellant is placed on the ceramic substrate to support and fix the solid propellant sample to be tested.

9. A method for measuring the ignition delay time of a solid propellant, based on the apparatus according to any one of claims 1 to 8, characterized in that, Specifically, the steps include the following: First, ignition delay time measurement data based on four different physicochemical principles were obtained simultaneously, as follows: (1) Determination of optical ignition delay time τ1 based on image method: Using a high-speed camera to acquire images of the combustion process, and by analyzing the flame front evolution characteristics in the image sequence, the moment when the first sustainable flame front appears is defined as the optical ignition moment t1, and τ1 = t1 - t0 is calculated, where t0 is the laser ignition trigger moment; (2) Determination of thermodynamic ignition delay time τ2 based on temperature method: Using a three-color colorimetric thermometer to acquire the time series data of the reaction zone temperature, after filtering and differentiation, the moment corresponding to the first inflection point in the temperature change rate dT / dt curve is defined as the thermodynamic ignition moment t2, and τ2 = t2 - t0 is calculated; (3) Determination of chemical ignition delay time τ3 based on spectroscopic method: Using a spectrometer and a photomultiplier tube with a filter to monitor the intensity of the characteristic spectral lines of the solid propellant to be tested, the moment when the radiation intensity reaches the set threshold is defined as the chemical ignition moment t3, and τ3 = t3 - t0; (4) Determination of gas dynamic ignition delay time τ4 based on gas pressure method: The combustion chamber pressure time curve is obtained by using a pressure sensor. After noise reduction and differentiation, the time corresponding to the first maximum point in the pressure change rate dP / dt curve is defined as the gas dynamic ignition time t4. τ4 = t4 - t0 is calculated. Then, the Dempster-Shafer evidence theory framework is used for fusion, as follows: A recognition framework is constructed based on experimental data, and a confidence interval model is built for test data obtained by a single detection method; the confidence interval is defined as follows: A， The dataset of ignition delay times obtained by independent detection methods is constructed by arranging the data in ascending order of time, i.e.: A i =[a i - , a i + ], where a i - and a i + These represent the minimum and maximum values ​​of the ignition delay time data obtained by the independent detection method, respectively. A i of The complement represents the range of inconsistent data: =[-∞, a i - ]∪[a i + [+∞], using A I , A T , A S , A P The confidence interval is constructed by arranging the ignition delay time datasets obtained by the image method, temperature method, spectral method, and pressure method in ascending order, where the ignition delay time data measured by each individual method is used as the interval number. Calculate the interval distance d(a) of test data obtained by different methods i , b j ) and interval similarity S(a i , b j Determine the basic probability allocation function BPA for each data source, for any interval number a1=[a1 - a1 + ],b1=[b1 - b1 + The distance between interval numbers can be expressed using the following formula: in, , , , ; , These are the midpoint influence factor and the width influence factor of the interval numbers, respectively; the interval similarity between interval numbers a1 and b1 is S(a1, b1), expressed by the formula: ;in, >0 represents the support coefficient, and d(a1, b1) represents the distance between interval numbers a1 and b1. The interval similarity assignment (BPA) obtained from the test data based on different testing methods: m In m Tn m Sn m Pn After normalizing BPA, the dynamic weighting coefficient k is obtained. Iw k Tw k Sw k Pw Multimodal data weighting fusion is achieved through dynamic weighting coefficients and weighting formulas, where m In m Tn m Sn m Pn These represent BPA obtained by the image method, temperature method, spectral radiometric method, and pressure method, respectively. Similarly, k Iw k Tw k Sw k Pw These represent the dynamic weighted coefficients of BPA normalized by different methods, where the subscripts are... w and n Experimental groups representing a single method; The final weighted average yields the ignition delay time t. in = k Iw t In + k Tw t Tn + k Sw t Sn + k Pw t Pn , where t In t Tn t Sn t Pn Let t represent the ignition delay time data of the nth group obtained by the image method, temperature method, spectral irradiance method, and pressure method, respectively. in This represents the ignition delay time of the nth group of four single methods after being fused within one of the confidence intervals.

Images