Mask-based map-all test structure

Abstract

The application belongs to the field of dynamic combustion flow field mapping measurement, and specifically discloses a mapping-all-in-one test structure based on a mask technology, and designs a high-resolution, high-spectrum and high-dynamic-range thermal measurement system specially designed for solid rocket engines. The system can accurately measure and monitor the temperature distribution of the exhaust plume of the solid rocket engine by using advanced sensing technology and data processing algorithms. The system has excellent spatial resolution, high-spectrum measurement capability and high-dynamic-range response to rapid temperature changes. By providing accurate temperature data, the system provides key technical support for performance optimization, fault diagnosis and safety control of the solid rocket engine.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of dynamic combustion flow field mapping measurement, in particular to a mapping and spectrum integrated test structure based on mask technology. BACKGROUND

[0002] Solid rocket engines are widely used in various aerospace applications due to their simplicity, reliability, and high thrust-to-weight ratio. However, monitoring the plume temperature in solid rocket engines presents significant challenges. Existing thermal measurement techniques often lack the required accuracy, spatial resolution, and real-time performance, making it difficult to accurately measure the temperature distribution in critical regions of the plume.

[0003] Thermal measurement of solid rocket engines is crucial for optimizing combustion efficiency, preventing malfunctions, and ensuring safety. Existing thermal measurement systems have certain limitations in terms of spatial resolution, spectral range, and dynamic response. The present application aims to address these limitations by introducing an advanced thermal measurement system specifically designed for solid rocket engines, providing a high-resolution, high-spectral, and high-dynamic-range measurement system. SUMMARY

[0004] To solve the above technical problems, a mapping and spectrum integrated test structure based on mask technology is proposed. The test structure improves the sampling speed of the sensor from three aspects: time domain, spatial domain, and frequency domain, thereby obtaining high-resolution information compatible with time, space, and spectrum.

[0005] The technical solution protected by the present application is a mapping and spectrum integrated test structure based on mask technology, which includes an objective lens. A light splitter is arranged behind the objective lens. The light splitter divides the light passing through the objective lens into two beams, which pass through a first mask unit and a second mask unit, respectively. The pinhole arrays of the first mask unit and the second mask unit are arranged in spatial complementarity. The optical structures behind the first mask unit and the second mask unit are the same. The optical structures are arranged in order: a collimating mirror, a dispersive prism, a focusing mirror, and a spectral detection device.

[0006] The spectral detection device includes a complementary metal oxide semiconductor (CMOS). The CMOS is an imaging device. The data collected by the CMOS is transmitted to a field programmable gate array (FPGA) processor. The FPGA processor serves as a controller. It processes the collected image data and stores and transmits the data. The FPGA processor is connected to a host computer through a BNC interface. The host computer processes the received data.

[0007] Further, the mapping and spectrum integrated test structure based on mask technology also includes a narrow-band filter. The narrow-band filter is arranged in front of the objective lens or between the dispersive prism and the collimating mirror.

[0008] Further, the wavelength working range selection process of the narrow-band filter is as follows:

[0009] First, according to the influence characteristics of the wavelength on the spectral differential value of the radiation source term, a wavelength range with higher reconstruction accuracy in the process of inverting the temperature of the combustion field space radiation source term based on hyperspectral is determined;

[0010] Secondly, the gray body and continuous radiation spectrum range of the solid rocket engine combustion field are determined.

[0011] Finally, the spectral range with less influence on the transmission of the radiation source term light from the interior of the combustion field to the boundary is determined.

[0012] Based on the overlapping part of the above three kinds of spectral information, according to the number of combustion field space radiation source term tomography, combined with the size of the detector target surface and the distance between the dispersion spectrometer unit and the detector target surface, the spectral resolution and the number of spectral channels in the hyperspectral detection function are determined, and then the selection of the narrow-band filter light waveband range is completed.

[0013] Further, the host computer performs registration on the received data, and the specific registration process is as follows:

[0014] Step 1) In all the collected images, select one image as the reference image for registration, and the remaining images are used as the registration images.

[0015] Step 2) Feature points are obtained by feature extraction between the to-be-registered image and the reference image, and respective feature sets are formed.

[0016] Step 3) Find the matched feature point pairs by similarity measurement.

[0017] Step 4) Given the search range of the transformation parameters, perform full search on the lowest resolution layer; take out the transformation parameters in the search space, perform geometric transformation on the corresponding layer of the to-be-registered image, obtain the preliminary transformation parameter estimation of the optimal solution under the resolution by using the gray-based cross-correlation registration method, and take the estimation as the search center for the processing of the next image layer.

[0018] Step 5) Based on the search center of one layer, search the transformation parameters at a higher resolution, gradually refine the transformation parameters from coarse to fine, and finally obtain the registration parameters that meet the accuracy requirements on the original registration image. Finally, the image is registered by the coordinate transformation parameters.

[0019] Compared with the prior art, the present application has the following advantages:

[0020] 1. For high spatial resolution detection part, based on light energy beam splitter to make the combustion field the same scene information into different detectors, by placing a mask unit in front of the detector, the pinhole array in different mask structure is arranged in space complementary, so as to make up the spatial resolution loss caused by avoiding spectral overlap in single mask plate discrete sampling.

[0021] 2. For spectral imaging structure, there is a competitive relationship between space-time resolution and spectral resolution. The present application uses the spectral range less affected by the extreme harsh environment of the combustion field, and combines the demand of the number of radiation source items, and uses narrowband filtering to spread the anti-interference spectral information on the fixed size image target surface through the dispersion prism, so as to improve the spectral resolution.

[0022] 3. The high-speed detection structure based on FPGA and high-frequency response CMOS sensor are used to further improve the sampling speed of the sensor, so as to obtain high-resolution information compatible with space-time spectrum. BRIEF DESCRIPTION OF DRAWINGS

[0023] The present application will be described in further detail below with reference to the accompanying drawings.

[0024] Figure 1 The figure is a schematic diagram of the spectral detection structure of the present application.

[0025] Figure 2 The figure is a schematic diagram of the spectral detection device.

[0026] Figure 3 The figure is a schematic diagram of the pinhole array of two mask units.

[0027] Figure 4 The figure is a schematic diagram of the narrowband filter.

[0028] Figure 5 The figure is a schematic diagram of the feature-based image registration process.

[0029] Figure 6 The figure is a schematic diagram of the pixel registration effect. DETAILED DESCRIPTION

[0030] In order to make the purpose, features and advantages of the present application obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0031] In order to better understand the principle of the algorithm of the present application, the principle of the 3D point cloud virtual structured light coding of geometric rearrangement will be briefly described below.

[0032] As Figure 1As shown, the atlas test structure based on mask technology includes an objective lens, a light splitter arranged behind the objective lens, the light splitter divides the light passing through the objective lens into two beams, the first mask unit and the second mask unit respectively, the pinhole arrays of the first mask unit and the second mask unit are arranged in space complementation, the optical structures behind the first mask unit and the second mask unit are the same, and the optical structures are arranged in sequence as a collimating mirror, a dispersive prism, a focusing mirror and a spectral detection device.

[0033] As shown in the figure, Figure 2 The spectral detection device includes a complementary metal oxide semiconductor (CMOS) as an imaging device, the CMOS is connected with the FPGA processor and transmits the collected data to the FPGA processor, the FPGA processor serves as a controller, processes the collected image data and stores and transmits the image data, the FPGA processor is connected with an upper computer through a BNC interface, and the upper computer processes the received data.

[0034] In the snapshot spectral imaging structure, there is a physical competitive relationship between the spatial resolution and the spectral resolution, and the mask technology further restricts the image spatial resolution. The time-space-frequency joint filtering technology is used to realize high-frequency and high-resolution spectral information acquisition. According to the Planck radiation law, spectral temperature measurement does not require full-wavelength information, further use of the spectral range less affected by the harsh environment of the solid rocket engine combustion field, and combination of the requirement of the number of layers of the combustion field radiation source term to be analyzed, the invention discards the invalid spectral information by adding a narrow-band filter in front of the detector, thereby improving the spectral resolution on the basis of fixed detector size. The narrow-band filter can be arranged in front of the objective lens, or between the dispersive prism and the collimating mirror.

[0035] The wavelength working range selection principle of the narrow-band filter is as follows: first, according to the influence characteristics of wavelength on the spectral differential value of the radiation source term, determine the wavelength range with higher reconstruction accuracy in the process of reconstructing the temperature of the combustion field spatial radiation source term based on hyperspectral inversion; second, determine the gray body and continuous radiation spectral range of the solid rocket engine combustion field; and finally, determine the spectral range less affected by the transmission of the radiation source term light from the harsh environment inside the combustion field to the boundary. Based on the overlapping part of the above three kinds of spectral information, according to the requirement of the number of combustion field spatial radiation source term tomography, and in combination with the size of the detector target surface and the distance between the dispersive spectrometer unit and the detector target surface, the spectral resolution and the number of spectral channels in the hyperspectral detection function are determined, and the selection of the light transmission wavelength range of the narrow-band filter is completed on the basis of the above steps.

[0036] As shown in the figure, Figure 2As shown, for the high time resolution sampling part, a multi-channel FPGA-based high-speed spectral detection device is designed. A high-frequency complementary metal oxide semiconductor (CMOS) is used as the imaging device; a FPGA processor is used as the core controller of the overall system to preliminarily process the collected image data and store and transmit the image data; a CoaXPress communication protocol is used to perform high-speed transmission of the data, and high-speed transmission can be realized by increasing the communication cable; in order to realize high frame rate data acquisition speed, data storage is performed in the form of in-machine cache; a BNC interface is designed to connect with a computer terminal, and an upper computer is designed to perform parameter control and data transmission of the high-speed spectral detection device.

[0037] For the high spatial resolution detection part, in order to realize high spatial resolution of spectral information, a light beam splitter is used to make the same light field from the combustion field scene enter different detectors. A mask unit is placed in front of each detector, such as Figure 3 As shown, the mask structures are complementary in space to compensate for the loss of spatial resolution caused by the overlapping of spectral information when the wavelength light splitting element expands the spectral information in space. After passing through the mask unit, the optical system structures behind each mask are the same.

[0038] The spectral space-time high-resolution imaging structure of the application comprises two spectral detection devices, and in fact, different detectors are not strictly coaxial optical systems. Therefore, it is necessary to analyze the geometric distortion of each component image, and to use geometric transformation to normalize the image to a unified coordinate system for gray level and spatial registration. The corresponding registration technology process is as shown in Figures 5-6 As shown, one image is selected as the reference standard for registration, and the remaining images are used as registration images. Among them, the application mainly studies the relative registration of large-format images, and how to determine the registration function mapping relationship between different images is the key. A suitable polynomial is used to fit the translation, rotation and affine transformation between two images, thereby converting the image registration function mapping relationship into how to determine the coefficients of the polynomial, and finally converting it into how to determine the registration control points.

[0039] Step 1) In all collected images, one image is selected as the reference image for registration, and the remaining images are used as registration images;

[0040] Step 2) Feature points are obtained by feature extraction between the to-be-registered images and the reference image to form respective feature sets;

[0041] Step 3) Find the matched feature point pairs by similarity measurement;

[0042] Step 4) Given the search range of transform parameters, full search is performed on the lowest resolution layer; the transform parameters in the search space are taken out in turn, the corresponding layers of the images to be registered are geometrically transformed, the gray-based cross-correlation registration method is adopted, the preliminary transform parameter estimation of the optimal solution at this resolution is obtained, and the estimation is taken as the search center for the processing of the image layer at the next level;

[0043] Step 5) Based on the search result of one layer, search the transform parameters at a higher resolution level, gradually refine the transform parameters from coarse to fine, and finally obtain the registration parameters that meet the accuracy requirements on the original registration image. Finally, the image is registered by the coordinate transformation parameters.

[0044] A two-dimensional matrix I1(x,y) and I2(x,y) of a given size are used to represent the images obtained by the two detectors, respectively, and the mapping relationship is

[0045] I2(x,y)=g(I1(f(x,y)))

[0046] Where f and g are two-dimensional coordinate transformation and gray transformation, respectively.

[0047] Based on the complementary information between multiple images, a single image is restored based on multiple images, and its degradation process can be represented as

[0048] k=1,2,…,N

[0049] Where N is the number of existing degradation samples, Yk is the kth sample, is the point spread function of the kth sample, Fk is the transform function of the kth sample, and Vk is the noise of the kth sample.

[0050] Through detailed analysis of the image restoration model, the single image is deblurred using multiple images. On the basis of norm minimization and robust recovery of regularization parameters, the search gradient direction optimization of the descent algorithm is used, and the image restoration algorithm of the CGD algorithm is used. Through a large number of experimental comparisons, the optimal algorithm is determined and the clearest combustion field image is obtained.

[0051] The system adopts advanced high-resolution temperature sensors, which can achieve accurate temperature measurement in the key components of solid rocket engines. The sensors have excellent spatial resolution capability, which can capture local temperature changes and temperature gradients; the system has multi-spectral measurement function, which can obtain more extensive temperature information. By using sensors of different wavelengths, the temperature of different fuels and combustion products can be accurately measured, so as to realize more comprehensive temperature distribution analysis; the system has high dynamic range response capability, which can quickly and accurately capture the rapid change of temperature. Whether in the process of starting, accelerating or decelerating, the system can provide accurate temperature measurement results, helping to monitor and control the temperature change of solid rocket engine. The system is equipped with high-speed data acquisition and processing technology, which can acquire and process temperature data in real time. This can realize real-time monitoring and analysis of temperature data, support real-time control and fault diagnosis of the engine; the sensors and components of the system are carefully designed, which can work stably in high temperature, high pressure and harsh environment. Its high reliability and durability ensure that the system can run stably for a long time, meet the requirements of solid rocket engine exhaust temperature measurement in harsh environment.

[0052] The above describes the embodiments of the application in detail in combination with the drawings, but the application is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the purpose of the application.

Claims

1. A combined image and spectrum testing structure based on mask technology, including an objective lens, characterized in that: A beam splitter is provided behind the objective lens. The beam splitter splits the light passing through the objective lens into two beams, which then pass through a first mask unit and a second mask unit respectively. The pinhole arrays of the first mask unit and the second mask unit are spatially complementary. The optical structures behind the first mask unit and the second mask unit are the same. The optical structures include a collimating lens, a dispersive prism, a focusing lens, and a spectral detection device arranged in sequence. The spectral detection device includes a complementary metal-oxide-semiconductor (CMOS) sensor, which is an imaging device. It is connected to an FPGA processor and the acquired data is transmitted to the FPGA processor. The FPGA processor acts as a controller, processing, storing, and transmitting the acquired image data. The FPGA processor is connected to a host computer via a BNC interface, and the host computer processes the received data. It also includes a narrowband filter, which is positioned in front of the objective lens or between the dispersive prism and the collimating lens. The process for selecting the wavelength operating range of the narrowband filter is as follows: First, based on the influence characteristics of wavelength on the differential value of the radiation source term spectrum, determine the wavelength range with high reconstruction accuracy in the process of reconstructing the temperature of the spatial radiation source term of the combustion field based on hyperspectral inversion. Secondly, determine the ash body and continuous radiation spectrum range of the solid rocket motor combustion field; Finally, the spectral range of the radiation source term that was least affected during the transmission of light from the harsh environment inside the combustion field to the boundary was determined. Based on the overlapping portion of the above three types of spectral information, the spectral resolution and number of spectral channels in the hyperspectral detection function are determined according to the number of spatial radiation source terms in the combustion field, combined with the size of the detector target surface and the distance between the dispersive spectral unit and the detector target surface, and then the screening of the light transmission band range of the narrowband filter is completed.

2. The image-data integration test structure based on mask technology according to claim 1, characterized in that: The host computer performs registration on the received data. The specific registration process is as follows: Step 1) Select one image from all the acquired images as the reference image for registration, and use the remaining images as the registration images; Step 2) Feature points are obtained by extracting features between the image to be registered and the reference image, forming their respective feature sets; Step 3) Find matching feature point pairs by performing a similarity measurement; Step 4) Given the search range of transformation parameters, perform a full search on the layer with the lowest resolution; sequentially extract the transformation parameters in the search space, perform geometric transformation on the layer corresponding to the image to be registered, and use the gray-level cross-correlation registration method to obtain the preliminary transformation parameter estimate of the optimal solution at this resolution, and use this estimate as the search center for the next level of image layer processing. Step 5) Based on the search results of the first layer as the search center, search for transformation parameters at a higher level of resolution, gradually refine the transformation parameters from coarse to fine, and finally obtain the registration parameters that meet the accuracy requirements on the original registered image. Finally, perform image registration using coordinate transformation parameters.

Images