Method for determining multi-parameter ignition criterion of solid propellant based on data fusion theory

Abstract

The present application relates to solid propellant technology, and aims to provide a method for determining a multi-parameter ignition criterion of a solid propellant based on a data fusion theory.The method comprises the following steps: preprocessing and time sequence calibration of original data of a solid propellant ignition experiment, to obtain a time sequence set of multi-source data; according to the characteristics of different parameters, corresponding single-parameter ignition criteria are selected respectively for ignition feature extraction, and an ignition feature set is obtained; data fusion is performed based on the two data sets to obtain a fusion event probability; an ignition decision is established according to the fusion event probability, and after parameter compensation and feature inversion, a fusion ignition delay time is finally obtained.The present application can make up for the test defects of some single sensors by performing fusion calculation and analysis on the test results of multi-source sensors, can quickly extract ignition features and calculate a fusion ignition delay time, and has higher applicability and higher flexibility than existing single-parameter ignition criteria.

Description

Technical Field

[0001] The present invention relates to the technical field of solid propellants, and aims to provide a method for determining a multi-parameter ignition criterion of a solid propellant based on data fusion theory. Background Art

[0002] Solid propellant is an energetic composite material with specific properties. It boasts high energy density, rapid combustion velocity, and high specific impulse. It serves as the fuel for various solid motors used in missiles and spacecraft, and as the power source for solid rocket engines. It plays a crucial role in the development of missile and aerospace technology. Research on the ignition characteristics of solid propellants, which is closely related to the fuel's energy release and combustion kinetics, is a hot topic in defense and aerospace engineering.

[0003] Solid propellant ignition criteria are methods or standards used to determine the starting point of solid propellant ignition. They are an important part of studying the ignition and combustion characteristics of solid propellants and an important basis for establishing chemical kinetic models and simplified models of combustion reactions. Current ignition criteria are mostly based on a single parameter, using parameter characteristics such as temperature, pressure, ignition / non-ignition, or radiation intensity as the basis for determining the ignition starting point. This is usually a threshold value for a particular parameter or a threshold value for its rate of change. For example, in temperature-based ignition criteria, ignition is considered to be complete when the solid propellant gas phase temperature reaches 1500K or the gas phase temperature rise rate is the maximum. The corresponding time at this time is the ignition delay time.

[0004] However, in practical applications, single-parameter ignition criteria have many problems and drawbacks. For example, ignition criteria are not universal across experiments and simulations, parameters such as temperature are difficult to test in experiments, pressure data responds slowly, and simulation studies cannot accurately capture radiation characteristics. Therefore, it is necessary to establish a universal solid propellant ignition criterion. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the deficiencies of the existing single-parameter ignition criterion in terms of applicability and accuracy, and to provide a method for determining the multi-parameter ignition criterion of solid propellant based on data fusion theory.

[0006] To solve the technical problem, the solution of the present invention is:

[0007] A method for determining a multi-parameter ignition criterion for a solid propellant based on data fusion theory is provided, comprising the following steps:

[0008] (1) Conduct ignition experiments on solid propellants to obtain flame image sequences, and raw data on the temporal changes of temperature, pressure, and spectral radiation intensity during the ignition process;

[0009] (2) Preprocessing and time series calibration of the original data to obtain a time series set of multi-source data;

[0010] (3) According to the characteristics of different parameters, the corresponding single-parameter ignition criteria are selected to extract ignition features and obtain an ignition feature set;

[0011] (4) Data fusion is performed based on the multi-source data time series set and the ignition feature set. The fusion process includes six steps, namely, establishing an identification framework, basic probability assignment, probability compensation, reliability analysis, conflict analysis, and Dempster binary iterative fusion, to finally obtain the fusion event probability.

[0012] (5) The ignition decision is established based on the probability of the fusion event. After parameter compensation and feature inversion, the fusion ignition delay time is finally obtained.

[0013] As a preferred embodiment of the present invention, in step (2), the raw data is preprocessed as follows:

[0014] (2.1) Homogenize the raw data obtained from the experiment, including:

[0015] The flame image sequence is converted into a time series of grayscale values ​​through grayscale processing. The temperature data is calibrated and inverted to obtain the temperature time series. The pressure data is fitted and calculated to obtain the pressure time series. The time series of the 486nm AlO characteristic band in the spectral radiation intensity is selected.

[0016] (2.2) The time series of the four types of data are calibrated to complete the time series synchronization and obtain the time series set of multi-source data.

[0017] As a preferred solution of the present invention, in step (3), the single-parameter ignition criteria are divided into three categories: data threshold type, data extreme value type and data change rate maximum value type. According to the properties of the four types of original data obtained from the experiment, the adapted single-parameter ignition criteria are selected from the corresponding types to extract the ignition features and obtain the ignition feature set.

[0018] As a preferred embodiment of the present invention, step (4) specifically includes the following steps:

[0019] (4.1) In the solid propellant ignition and combustion process, ignition and combustion are considered as two independent propositions. Using ignition characteristics as a node, the process is divided into two basic elements: the ignition delay phase and the combustion phase, and a basic identification framework is established:

[0020] (4.2) Based on the established basic identification framework, basic probability assignment is performed according to the ignition characteristic results;

[0021] Based on the data preprocessing results, the time when each parameter initially increases is analyzed and recorded. The ignition starting point is extracted and determined from the ignition feature set using a single-parameter ignition criterion. The combustion end time is obtained from the time series set of multi-source data. The ignition starting point is used as the ignition fuzzy boundary for fuzzy boundary probability allocation.

[0022] (4.3) Calculate the average value of the data from the four sources as the fifth data source; this will be used for probability compensation in subsequent steps;

[0023] (4.4) Study the trust metric Bel and plausibility metric Pl of each data source to verify the reliability of the original data;

[0024] (4.5) Calculate the conflict coefficient k, which is used to evaluate the degree of conflict between the focal elements of each data source. The larger the k value, the greater the conflict;

[0025] (4.6) According to Dempster’s binary iterative fusion rule, the five data sources are fused in pairs to obtain the fused probability distribution.

[0026] The present invention further provides a system for determining multi-parameter ignition criteria for solid propellants, comprising a sensor module, a data preprocessing module, an ignition feature extraction module, a data fusion module, and a decision and inversion module arranged in sequence;

[0027] The sensor module includes a high-speed microscopic camera, a fiber optic spectrometer, a temperature meter, and a pressure sensor installed in the solid propellant ignition experimental device, which are used to obtain the flame image sequence, temperature, pressure, and raw data of the time-varying changes of spectral radiation intensity during the ignition process;

[0028] The data preprocessing module, ignition feature extraction module, data fusion module and decision and inversion module are all software function modules set in the computer, which are used to execute the contents of the above steps (2) to (5) respectively.

[0029] As a preferred solution of the present invention, synchronous control is achieved between the igniter in the ignition experimental device and the various devices in the sensor module, so that the various data measured in the experiment can correspond to each other in time.

[0030] Description of the invention principle:

[0031] This paper utilizes a multi-source sensor test system for the solid propellant ignition process to obtain raw data, which is then used to establish a multi-parameter solid propellant ignition criterion based on data fusion, using a modified DS evidence theory. The entire system consists of five components: a multi-source sensor test module, a data preprocessing module, an ignition feature extraction module, a data fusion module, and a decision and inversion module.

[0032] Multi-source data fusion technology is a multi-level, multi-faceted process that automatically detects, connects, correlates, estimates, and combines information and data from multiple sources. Through various methods, it automatically or semi-automatically transforms information from different sources and at different time points into a single form to provide effective support for related research and automated decision-making. This technology integrates multiple disciplines, including signal processing, information theory, statistical estimation, and artificial intelligence, to effectively enhance data credibility and validity.

[0033] Dempster-Shafer evidence theory (DS) is a commonly used method for analyzing, processing, and modeling uncertain information during multi-source information fusion. Essentially, it is a generalization of probability theory, expanding the basic event space in probability theory into the probability space of basic events and establishing a basic probability assignment function based on this space. Compared with traditional probability theory, DS evidence theory can not only effectively express cumulative uncertainty, but also incomplete information and subjective uncertainty. Furthermore, DS evidence theory provides an effective Dempster combination rule for information fusion. This rule satisfies the excellent properties of commutativity and associativity, enabling the fusion of evidence without prior information, effectively reducing system uncertainty.

[0034] Existing multi-source data fusion technology and DS evidence theory have the defect of unclear compatibility when applied to new fields. If they are directly applied, it will lead to problems such as large data conflicts, fusion failure or abnormal fusion results.

[0035] Therefore, the present invention has made improvements in terms of data homogenization and probabilistic processing of real events. Among them, the improvement in data homogenization can expand the compatibility of the data fusion algorithm, so that it can be applied to more forms of data fusion scenarios. The improvement in the probabilistic processing of real events can enable accurate numerical representation of each stage of the solid propellant ignition and combustion process to improve the accuracy and rationality of data fusion. The above-mentioned improvement means were all developed in the context of solid propellant ignition research, and the improvement means were applied to the multi-source data fusion technology based on DS evidence theory, breaking through the conventional thinking mode of those skilled in the art and broadening the application boundaries of multi-source data fusion technology.

[0036] Solid propellant ignition is a complex process involving combustion, flow, heat transfer, and fluid-solid coupling. Its operating conditions are extremely complex and accompanied by a high-temperature, high-pressure combustion environment, making the measurement of parameters such as temperature challenging. Measuring this process using multiple sensors and fusing the resulting multi-source data is a significant testing method that can complement the strengths of various test equipment and has significant application potential.

[0037] Compared with the traditional single-parameter ignition criterion technology, the present invention has the following advantages:

[0038] (1) This invention applies data fusion theory to the study of solid propellant ignition criteria for the first time. The test results of multiple sensors are integrated and analyzed through data fusion algorithms, which can largely make up for some of the test defects of single sensors.

[0039] (2) The data fusion stage of the present invention uses a probability compensation method and a binary iterative fusion algorithm to improve the fusion accuracy and efficiency of the traditional Ds evidence theory;

[0040] (3) The multi-parameter ignition criterion system for solid propellants disclosed in the present invention realizes programmed operation, and can quickly extract ignition characteristics and calculate fusion ignition delay time;

[0041] (4) The multi-parameter ignition criterion obtained by the present invention can change the type and number of sensors for subsequent data fusion according to the needs and conditions of the actual research scenario. Compared with the existing single-parameter ignition criterion, it is more applicable and more flexible to use. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Flowchart of multi-parameter ignition criteria for solid propellant based on data fusion.

[0043] Figure 2 is the reliability interval distribution of proposition A.

[0044] Figure 3 This is an image of the flame from a solid propellant ignition experiment.

[0045] Figure 4 Spectral radiation intensity data of solid propellant ignition experiment.

[0046] Figure 5 These are the electric fire excitation signals, temperature, and pressure data for the solid propellant ignition experiment.

[0047] Figure 6 Solid propellant ignition and combustion data preprocessing results.

[0048] Figure 7 Multi-source basic probability assignment results and data fusion probability results at each stage. DETAILED DESCRIPTION

[0049] First of all, it should be noted that the present invention relates to database technology, which is an application of computer technology in the field of information security technology. In the process of implementing the present invention, the application of multiple software functional modules will be involved. The applicant believes that after carefully reading the application documents and accurately understanding the implementation principles and purpose of the present invention, and combining the existing known technology, those skilled in the art can fully use their software programming skills to implement the present invention. The aforementioned software functional modules include but are not limited to: scanning module, bottom detection module, comparison module, etc. All those mentioned in the application documents of the present invention fall into this category, and the applicant will not list them one by one.

[0050] Those skilled in the art will appreciate that, in addition to implementing a portion of the system and its various devices, modules, and units provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same functions of the system and its various devices, modules, and units provided by the present invention in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; the devices, modules, and units for implementing various functions can also be considered as both software modules implementing the method and structures within the hardware component.

[0051] It should also be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0052] Combine Figure 1 The flow chart of the multi-parameter ignition criterion for solid propellant based on data fusion is shown in FIG. The specific implementation process of the present invention is as follows.

[0053] like Figure 1As shown, the present invention proposes a system for determining the multi-parameter ignition criterion of solid propellant, comprising a sensor module, a data preprocessing module, an ignition feature extraction module, a data fusion module and a decision and inversion module arranged in sequence; wherein the sensor module is a hardware functional module, including a high-speed microscopic camera (High-speed Camera), a thermometer (Thermometer), a pressure transducer (Pressure transducer), and a fiber optic spectrometer (FTIR) provided in a solid propellant ignition experimental device, and is respectively used to obtain the flame image sequence, temperature, pressure and spectral radiation intensity over time during the ignition process. The data preprocessing module, ignition feature extraction module, data fusion module and decision and inversion module are all software functional modules provided in a computer, and are respectively used to execute the contents of steps (2) to (5) in the following scheme.

[0054] Based on this system, the present invention proposes a method for determining the multi-parameter ignition criteria for solid propellants based on data fusion theory. The main implementation process includes the following:

[0055] (1) Obtain raw data from solid propellant ignition experiments.

[0056] (2) Obtain a multi-source data time series set through the data preprocessing module.

[0057] (3) According to the requirements, a single parameter ignition criterion is selected to complete the ignition feature extraction and obtain the ignition feature set.

[0058] (4) The multi-source data time series and the ignition feature set are input into the data fusion module, and the steps of establishing the recognition framework are carried out in sequence to finally complete the data fusion and obtain the fusion event probability.

[0059] (5) Make decisions and inversion based on the probability of fusion events, and finally obtain the fusion ignition delay time.

[0060] The following is a detailed description of the implementation process of this method:

[0061] 1. Carry out ignition experiments to obtain raw data on the changes in ignition over time;

[0062] A multi-source sensor module is used to observe various parameters during the solid propellant ignition process. Using four test devices—a high-speed microscopic camera, a thermometer, a pressure transmitter, and a fiber optic spectrometer—they obtain raw data: flame image sequences, temperature, pressure, and spectral radiation intensity. Synchronous control is implemented between the igniter in the ignition experiment device and the various devices in the sensor module, ensuring temporal alignment of the measured data.

[0063] 2. Transmit the four types of raw data to the computer through signal lines for data processing.

[0064] First, enter the preprocessing module. In this functional module, the flame image sequence is converted into a time series of grayscale values ​​L(t) through grayscale processing to characterize the brightness of the flame in the image. The temperature data is calibrated and inverted to obtain the time series of temperature values ​​T(t). The pressure data is fitted to obtain the time series of pressure values ​​P(t). The spectral radiation intensity selects the time series of radiation intensity values ​​of the AlO characteristic band (486nm) I(t).

[0065] After time calibration, the four data time series are synchronized to obtain the time series set X = [L(t), I(t), T(t), P(t)] of the preprocessed data.

[0066] 3. Input the time series set X of preprocessed data into the ignition feature extraction module. The ignition feature extraction standard is the single parameter ignition criterion.

[0067] There are dozens of commonly used single parameter ignition criteria according to the different dimensional forms of the parameters, which can be divided into three categories according to the data type. The first category is the data threshold X j =X j，lim , it is considered that ignition is completed when the data reaches the set threshold, and the corresponding time is the ignition delay time t i ; The second category is data extreme values It is assumed that ignition is completed when the data reaches the extreme value, and the corresponding time is the ignition delay time t i ; The third category is the maximum value of data change rate It is assumed that ignition is completed when the data change rate reaches the maximum value, and the corresponding time is the ignition delay time t i .

[0068] For the three types of ignition criteria mentioned above, the four pre-processed parameters (data) can be used. Taking the first type of ignition criteria as an example (using the parameter threshold as the ignition feature extraction standard): the moment when the image grayscale L first develops to 0.1 times the maximum value is taken as the ignition starting point, that is, when L = 0.1L max The ignition is completed at , which corresponds to the ignition delay time based on the image data. In the specific application examples provided later, the four parameters (data) will all use the data threshold type ignition criteria to extract ignition features.

[0069] After extracting the ignition features, the ignition feature set t is obtained. i，X =[t i，L , t i，I , t i，T , t i，P ].

[0070] 4. The above processed time series set X and ignition feature set t i，x Input data fusion module.

[0071] The processing content of the data fusion module includes six steps, namely, establishing the identification framework, basic probability assignment, probability compensation, reliability analysis, conflict analysis, and Dempster binary iterative data fusion.

[0072] (4.1) Establishing an identification framework

[0073] The basic identification framework Θ establishes a mutually exclusive set of elements, establishing two independent propositions in the solid propellant ignition and combustion process: ignition is Proposition A, and combustion is Proposition B. Using ignition characteristics as nodes, it is divided into two basic elements: θ1: ignition delay phase and θ2: combustion phase.

[0074] Θ = {ignition delay phase θ1, combustion phase θ2}

[0075] Power set of Θ2 Θ It can be expressed as:

[0076]

[0077] Where, Refers to blank events;

[0078] (4.2) Basic probability assignment

[0079] Based on the above basic identification framework Θ, basic probability assignment is performed according to the ignition feature results.

[0080] According to the data preprocessing results, the time t0 when each parameter first increases is analyzed and recorded, and the ignition starting point t1 is determined by the single parameter ignition criterion (t1 is determined from the ignition characteristics t i，x Extracted), the combustion end time t2 is obtained from the preprocessed data time series X, and t1 is used as the ignition fuzzy boundary to perform fuzzy boundary probability distribution as follows:

[0081]

[0082] m(θ1)=t1 / t2

[0083] m(θ2)=(t2-t1) / t2

[0084] m(θ1∩θ2)=(t1-t0) / t2

[0085] If satisfied ∑m=1, which meets the DS basic probability assignment requirements.

[0086] The probability assigned to θ1 is the basic probability m(A) of the ignition proposition A, or the confidence level of proposition A. If m(A) > 0, proposition A can be considered a focal element. m(θ2) is the basic probability of proposition B, and m(θ1∩θ2) is the basic probability of proposition A∩B. The probability assignment results are shown in Table 1.

[0087] Table 1 Basic probability assignment results

[0088]

[0089] (4.3) Probability compensation

[0090] Probability compensation is an important process to make up for abnormal data, that is, to obtain the average value of the basic probability assignment results of the above four source data through mathematical calculations as the fifth data source.

[0091] This calculation process performs mean compensation on the aforementioned probability assignments. The result is a dimensionless value in the range [0, 1], which will be used for mean compensation in the subsequent reliability analysis, conflict analysis, and data fusion steps. This method reduces conflicts in multi-source data during calculation and application, and is generally known in the art.

[0092] (4.4) Reliability analysis

[0093] The trust metric Bel and plausibility metric P1 of each data source are studied through reliability analysis to verify the reliability of the original data.

[0094] Among them, the trust function requires Bel: 2 Θ →[0, 1] satisfies the following conditions:

[0095]

[0096] A∈2 Θ

[0097] Where Bel(A) is the belief measure that proposition A is true, and m(B) refers to the basic probability assignment result of combustion event B (see Table 1);

[0098] At this time, since A is a single-element proposition, it satisfies

[0099] Bel(A)=m(A)

[0100]

[0101] Bel(Θ)=1

[0102] The plausible function requires P1:2 Θ →[0, 1], and the following conditions are met:

[0103]

[0104] A∈2 Θ

[0105] Where Pl is the plausible measure on Θ, P1(A) represents the degree of non-objection to proposition A, Indicates the degree of belief that A is false.

[0106] The relationship between the plausibility function and the basic probability assignment and the belief function is as follows: Figure 2 shown.

[0107] It can be seen from this that the confidence interval of proposition A is [Bel(A), Pl(A)]. This is the upper and lower limits of the degree of confidence in proposition A, which can also reflect its degree of uncertainty to a certain extent.

[0108] Table 2 Probability distribution of ignition proposition A from multi-source data

[0109]

[0110] The present invention is mainly carried out for ignition events (ie, Proposition A), and Proposition B is used as auxiliary data for conflict analysis and fusion calculation.

[0111] (4.5) Conflict Analysis

[0112] Since there is conflicting information between the evidences, it is necessary to calculate the conflict coefficient k according to the following formula to evaluate the degree of conflict between the focal elements of each data source.

[0113]

[0114] Where A i Refers to the ignition event of the i-th data source; B j Refers to the combustion event of the jth data source; m(A i ) refers to the basic probability value of the ignition event of the i-th data source; m(B j ) refers to the basic probability value of the ignition event of the j-th data source.

[0115] The k value is usually in the range of [0, 1]. The larger the value, the greater the conflict. Larger conflicts will lead to abnormal fusion results. When k = 1, it means that the Dempster combination rule cannot be used.

[0116] (4.6) Dempster binary iterative fusion

[0117] According to the following fusion rules, the above five data sources are fused in pairs:

[0118]

[0119] Where m(A∩B) refers to the basic probability assignment of the overlapping phase of ignition and combustion events;

[0120] The obtained fusion results are shown in Table 3. The fusion probability of step 4 is used as the final fusion probability, that is, m 点火 、m 燃烧 、m 点火，燃烧 , corresponding to m(A), m(B), and m(A∩B) respectively.

[0121] Table 3 Probability distribution after Dempster binary iterative fusion

[0122]

[0123] 5. Decision-making and inversion based on the probability of fusion events

[0124] The purpose of ignition decision is to obtain the fusion ignition delay time t i，Fusion , fusion burning time t c，Fusion and the fusion ignition combustion overlap time t h，Fusion .

[0125] First, calculate according to the following formula to obtain the basic ignition decision:

[0126] t i,Fusion =m 点火 ×t 总

[0127] t c,Fusion =m 燃烧 ×t 总

[0128] t h,Fusion =t i,Fusion -m 点火，燃烧 ×t 总

[0129] Where m 点火 Refers to the probability of ignition event after fusion; m 燃烧 Refers to the probability of combustion events after fusion; m 点火，燃烧 Refers to the probability of ignition and combustion overlapping time after fusion; t 总 Refers to the total ignition and combustion time, which is the sum of the ignition delay time and the combustion time: t 总 =t i +t c .

[0130] Then, the compensation coefficients (α and β, respectively) of the ignition temperature and ignition pressure of the main research object (ignition delay time) are calculated according to the following formula:

[0131] α=t i，Fusion / t i , p

[0132] β=t i，Fusion / t i，T

[0133] Finally, the characteristic inversion process is repeated and iterated to obtain the compensated ignition temperature Ti and ignition pressure P i :

[0134] t i =αt i，p =βt i，T

[0135] T i =T(t i )

[0136] P i =P(t i )

[0137] Where, t i Refers to the ignition delay time, corresponding to the ignition moment; T(t i ) refers to the temperature corresponding to the temperature value time series data T(t) at the ignition moment; P(t i ) refers to the pressure corresponding to the pressure numerical time series data P(t) at the ignition moment; t i，p is the ignition delay time obtained using the pressure ignition criterion; t i，T is the ignition delay time obtained using the temperature ignition criterion. i is the ignition temperature, P i is the ignition pressure.

[0138] At this point, the workflow of the multi-parameter ignition criterion is completed, and the ultimate goal is to obtain the fusion ignition delay time t with the minimum deviation. i，Fusion .

[0139] A specific application example:

[0140] A multi-parameter ignition test experiment for solid propellant was conducted based on a laser ignition test bench. The experimental object was a 5×5×10mm HTPB solid propellant strip. The experimental environment was normal temperature, normal pressure, argon atmosphere, and a constant volume combustion chamber. The test equipment included a high-speed microscopic camera, a fiber optic spectrometer, a temperature meter, and a pressure sensor. The raw data obtained was as follows: Figure 3-5 As shown, Figure 3 For flame images, Figure 4 is the spectral radiation intensity, Figure 5 It is the electric fire excitation signal and the original data of temperature and pressure.

[0141] Based on the above raw data, preprocessing and time series synchronization are performed, and a 0.1X threshold is used to extract ignition features and perform subsequent probability assignment. The preprocessing data results are as follows: Figure 6 The ignition feature extraction and probability assignment results are shown in Table 4.

[0142] Table 4 Ignition characteristics and basic probability assignment results

[0143]

[0144] After the basic probability assignment is performed, the subsequent data fusion processing is carried out. The multi-source basic probability assignment results and the data fusion probability results at each stage are as follows: Figure 7 shown.

[0145] Table 5 Data fusion results at each stage

[0146]

[0147] The initial A event probability after the four data sources and probability compensation is analyzed together with the A event fusion probability after the four-step iterative fusion, as shown in the following example: Figure 7 As shown in the figure, the initial event probability for the P source data is significantly different from that of the other data sources. However, as the iterative fusion process progresses, the final fused event probability stabilizes at 0.2560. The ignition delay time decision is 0.8135s. The final fused event probability is comparable to the original event probabilities from the I, G, and T data sources, but significantly different from the P source data. This phenomenon demonstrates that this data fusion method can effectively eliminate the influence of abnormal data and achieve more accurate decisions.

[0148] In general, the multi-parameter ignition criterion for solid propellant based on data fusion provided by the present invention combines five functional modules, namely multi-source sensors, data preprocessing, ignition feature extraction, data fusion, and decision-making and inversion, and pioneers the application of multi-source data fusion theory to solid propellant ignition research. The present invention makes basic decisions based on the results of binary iterative data fusion, and performs parameter compensation to obtain correction coefficients of relevant ignition parameters. Then, repeated fusion iterations are performed through the feature inversion process to finally obtain the data fusion result with the smallest deviation, that is, the ignition starting point of multi-source fusion. Moreover, a more accurate ignition starting point is finally obtained through mutual compensation between multiple parameters and the deviation attenuation characteristics of the iterative process. Due to the differences in sensor types and quantities in different studies, researchers can choose the type and quantity of sensor data to be fused for use in different research scenarios.

[0149] The multi-parameter ignition criterion based on data fusion theory proposed in this invention has significant advantages. It can eliminate the influence of defective data and abnormal data, unify data dimensions and data forms, solve data conflict problems, and ultimately realize data combination and matching, thereby achieving accurate decision-making on ignition events.

Claims

1. A method for determining multi-parameter ignition criteria for solid propellants based on data fusion theory, characterized in that: The following steps are involved: (1) Conduct ignition experiments on solid propellants to obtain flame image sequences, raw data on the time-varying changes in temperature, pressure, and spectral radiation intensity during the ignition process; (2) Preprocess the original data and calibrate the time series to obtain a time series set of multi-source data; (3) According to the characteristics of different parameters, the corresponding single-parameter ignition criteria are selected to extract ignition features and obtain the ignition feature set; (4) Data fusion is performed based on the multi-source data time series set and the ignition feature set. The fusion process includes six steps, namely, establishing an identification framework, basic probability assignment, probability compensation, reliability analysis, conflict analysis, and Dempster binary iterative fusion, to finally obtain the fusion event probability. (5) Establish an ignition decision based on the probability of the fusion event, and finally obtain the fusion ignition delay time after parameter compensation and feature inversion.

2. The method according to claim 1, characterized in that In step (2), the raw data is preprocessed as follows: (2.1) Homogenize the raw data obtained from the experiment, including: The flame image sequence was converted into a time series of grayscale values ​​through grayscale processing. The temperature data was calibrated and inverted to obtain the temperature time series. The pressure data was fitted and calculated to obtain the pressure time series. The time series of the 486 nm AlO characteristic band in the spectral radiation intensity was selected. (2.2) The time series of the four types of data are calibrated to complete the time series synchronization and obtain the time series set of multi-source data.

3. The method according to claim 1, characterized in that In step (3), the single-parameter ignition criteria are divided into three categories: data threshold type, data extreme value type, and data change rate maximum value type. According to the properties of the four types of original data obtained in the experiment, the adapted single-parameter ignition criteria are selected from the corresponding types to extract the ignition features and obtain the ignition feature set.

4. The method according to claim 1, wherein The step (4) specifically includes the following steps: (4.1) In the solid propellant ignition and combustion process, ignition and combustion are considered two independent propositions. Using ignition characteristics as a node, the process is divided into two basic elements: the ignition delay phase and the combustion phase, and a basic identification framework is established: (4.2) Based on the established basic identification framework, basic probability assignment is performed according to the ignition characteristic results; Based on the data preprocessing results, the time when each parameter initially increases is analyzed and recorded. The ignition starting point is extracted and determined from the ignition feature set using a single-parameter ignition criterion. The combustion end time is obtained from the time series set of multi-source data. The ignition starting point is used as the ignition fuzzy boundary for fuzzy boundary probability allocation. (4.3) Calculate the average of the data from the four sources as the fifth data source; this will be used for probability compensation in subsequent steps; (4.4) Study the trust metric Bel and plausibility metric Pl of each data source to verify the reliability of the original data; (4.5) Calculate the conflict coefficient k, which is used to evaluate the degree of conflict between the focal elements of each data source. k The larger the value, the greater the conflict; (4.6) According to Dempster’s binary iterative fusion rule, the five data sources are fused in pairs to obtain the fused probability distribution.

5. A system for determining multi-parameter ignition criteria for solid propellants, characterized in that: It includes a sensor module, a data pre-processing module, an ignition feature extraction module, a data fusion module and a decision and inversion module which are arranged in sequence; The sensor module includes a high-speed microscopic camera, a fiber optic spectrometer, a temperature meter, and a pressure sensor installed in the solid propellant ignition experimental device, which are used to obtain the flame image sequence, temperature, pressure, and raw data of the time-varying changes of spectral radiation intensity during the ignition process; The data preprocessing module, ignition feature extraction module, data fusion module and decision and inversion module are all software function modules provided in a computer, and are respectively used to execute the contents of steps (2) to (5) in claim 1.

6. The system according to claim 5, characterized in that Synchronous control is achieved between the igniter in the ignition experimental device and the various devices in the sensor module, so that the various data measured in the experiment can correspond to each other in time.

Images