Spectral sampling element and computational spectral measurement device

Abstract

The present invention discloses a spectral sampling element for global spectral sampling in computational spectral measurement, wherein the spectral sampling element is a self-coupling optical waveguide having multiple self-coupling regions, wherein the self-coupling regions are formed by bending and folding the self-coupling optical waveguide itself. The present invention also discloses a computational spectral measurement device. Compared with the prior art, the spectral sampling element proposed in the present invention is based on an ordinary waveguide structure, does not require high process precision like gratings or photonic crystal devices, and does not introduce additional scattering losses. It is easy to design, does not require high processing precision, and has a larger working bandwidth and higher precision. The computational spectral measurement device proposed in the present invention can reconstruct the input spectrum with high precision through a single measurement. Due to the use of the above-mentioned spectral sampling element, the manufacturing difficulty and manufacturing cost are greatly reduced, and the measurement accuracy is effectively improved. Its practicality far exceeds that of existing computational spectrometers.

Description

Technical Field

[0001] The present invention relates to the technical field of spectrum measurement, and in particular to a spectrum sampling element for computational spectrum measurement. Background Art

[0002] In order to detect the information of the target spectrum, spectrometers have come into being. They can recover any unknown spectrum input. Spectrometers are widely used in communications, materials science, astronomy, geography, remote sensing and other fields. With the development of the Internet of Things and smart devices, there is an urgent need for integrated spectrometers that can reconstruct spectra with a single measurement, such as smart wearable devices, portable medical devices, drone remote sensing, etc. Existing integrated spectrometers mostly use narrowband spectrometry, that is, using narrowband filters or spectroscopic gratings to extract different wavelength components of the spectrum to be measured into different channels for separate measurement. The number of channels required is equal to the ratio of the bandwidth and accuracy of the spectrometer. This solution is simple in principle, but in order to obtain large bandwidth and high accuracy, the number of spectroscopic channels must be increased, resulting in a decrease in the signal energy received by each detector, affecting the system size and signal-to-noise ratio. Therefore, it is difficult to strike a balance between bandwidth, accuracy, size and signal-to-noise ratio.

[0003] Computational spectrometers have become a research hotspot as they can effectively solve the above problems. Figure 1 As shown in the figure, the signal is first evenly split into M paths. The signal is then globally sampled through M spectral sampling elements with different transfer functions. The sampling results are converted photoelectrically into electrical signals and then processed using specific algorithms such as convex optimization to reconstruct the unknown spectrum. The core of this type of spectrometer lies in the spectral sampling element. When using high-performance spectral sampling elements, the required number of spectroscopic channels M (and therefore the number of spectral sampling elements) can be much smaller than the ratio of the spectrometer's bandwidth to its accuracy. This effectively improves the spectrometer's signal-to-noise ratio and reduces system size while maintaining the spectrometer's wide bandwidth and high accuracy.

[0004] However, the spectral sampling elements used in existing published solutions are mostly random photonic crystals or random Bragg grating structures. These have the following disadvantages: they require high fabrication precision, introduce additional scattering losses to the signal, and have limited operating bandwidth and spectral accuracy. Therefore, there is an urgent need to find spectral sampling elements with lower fabrication difficulty and implementation cost, while offering better performance, to enhance the practicality of computational spectroscopy. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the deficiencies of the existing technology and provide a spectral sampling element for computational spectral measurement, which greatly reduces the processing difficulty and implementation cost while improving the measurement performance of the system.

[0006] The present invention specifically adopts the following technical solutions to solve the above technical problems:

[0007] A spectrum sampling element is used for global spectrum sampling in computational spectrum measurement. The spectrum sampling element is a self-coupling optical waveguide with multiple self-coupling regions, wherein the self-coupling regions are formed by bending and folding the self-coupling optical waveguide itself.

[0008] Furthermore, the spectrum sampling element further includes at least one ring optical waveguide coupled to the self-coupling optical waveguide.

[0009] The spectrum sampling element is an optical integrated element.

[0010] Based on the same inventive concept, the following technical solutions can also be obtained:

[0011] A computational spectrum measurement device comprises a plurality of spectrum sampling elements with different transfer functions, which are used to perform global spectrum sampling on a light signal to be measured; the spectrum sampling elements are the spectrum sampling elements described in any of the above technical solutions.

[0012] Preferably, the sampling parameters of the plurality of spectral sampling elements are optimized by a multi-objective optimization method with the optimization goal of minimizing the correlation between the spectral sampling elements and maximizing the randomness of the transfer function of a single spectral sampling element in the wavelength domain.

[0013] Further preferably, the correlation is measured by a mutual correlation coefficient.

[0014] Further preferably, the randomness of the transmission function in the wavelength domain is measured by the number of poles and / or the autocorrelation coefficient.

[0015] Preferably, the sampling parameters include at least one of the following parameters: the number of self-coupling regions in the self-coupling optical waveguide, the distance between the respective coupling regions in the self-coupling optical waveguide, and the coupling coefficient of the respective coupling regions in the self-coupling optical waveguide.

[0016] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0017] The spectral sampling element proposed in the present invention is based on an ordinary waveguide structure. It does not require high process precision like grating or photonic crystal devices, and does not introduce additional scattering losses. It is easy to design, does not require high processing precision, and has a larger working bandwidth and higher precision.

[0018] The computational spectral measurement device proposed in the present invention can reconstruct the input spectrum with high precision through a single measurement. Due to the use of the above-mentioned spectral sampling element, the manufacturing difficulty and cost are greatly reduced, while the measurement accuracy is effectively improved. Its practicality far exceeds that of existing computational spectrometers. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a schematic diagram of the structural principle of the computational spectrum measurement device;

[0020] Figure 2 is a specific embodiment of the spectrum sampling element of the present invention;

[0021] Figure 3 is another specific embodiment of the spectrum sampling element of the present invention;

[0022] Figure 4 for Figure 3 The transfer function of the spectral sampling element shown;

[0023] Figure 5 This is a measurement effect diagram of the computational spectrum measurement device of the present invention. DETAILED DESCRIPTION

[0024] In response to the shortcomings of the existing technology, the solution of the present invention is to use multi-point self-coupled optical waveguides as sampling elements for computational spectroscopy measurement, which not only improves the measurement performance but also significantly reduces the processing difficulty and implementation cost of the components.

[0025] The technical solutions proposed in the present invention are as follows:

[0026] A spectrum sampling element is used for global spectrum sampling in computational spectrum measurement. The spectrum sampling element is a self-coupling optical waveguide with multiple self-coupling regions, wherein the self-coupling regions are formed by bending and folding the self-coupling optical waveguide itself.

[0027] In order to introduce more degrees of freedom in transmission function design and improve the randomness of the transmission function of the spectrum sampling element in the wavelength domain; further, the spectrum sampling element also includes at least one annular optical waveguide coupled to the self-coupling optical waveguide.

[0028] The computational spectrum measurement device proposed in the present invention includes multiple spectrum sampling elements with different transfer functions, which are used to perform global spectrum sampling on the light signal to be measured; the spectrum sampling elements are the spectrum sampling elements described in any of the above technical solutions.

[0029] Preferably, the sampling parameters of the plurality of spectral sampling elements are optimized by a multi-objective optimization method with the optimization goal of minimizing the correlation between the spectral sampling elements and maximizing the randomness of the transfer function of a single spectral sampling element in the wavelength domain.

[0030] Further preferably, the correlation is measured by a mutual correlation coefficient.

[0031] Further preferably, the randomness of the transmission function in the wavelength domain is measured by the number of poles and / or the autocorrelation coefficient.

[0032] Preferably, the sampling parameters include at least one of the following parameters: the number of self-coupling regions in the self-coupling optical waveguide, the distance between the respective coupling regions in the self-coupling optical waveguide, and the coupling coefficient of the respective coupling regions in the self-coupling optical waveguide.

[0033] To facilitate public understanding, the technical solution of the present invention is described in detail below through two specific embodiments:

[0034] Figure 2 The basic structure of the spectrum sampling element proposed in the present invention is shown in FIG. Figure 2 As shown, the spectral sampling element is a self-coupling optical waveguide, which forms self-coupling regions at multiple locations through its own bending and folding (the example in the figure shows a total of 12 self-coupling regions). For a single self-coupling waveguide, the multi-point self-coupling effect provides multiple optical paths for light propagation. The light components reaching the output end from different optical paths interfere with each other at the output end, causing the transmission function to vary dramatically with wavelength. Parameters such as the coupling coefficient of each self-coupling region, the number and distribution of self-coupling regions, and the distance between each coupling region all affect the transmission function of the self-coupling waveguide, providing multiple degrees of freedom in the design of the transmission function. Based on these parameters, it is easy to make the transmission function of a single spectral sampling element extremely random in the wavelength domain and the correlation between the transmission functions of any two spectral sampling elements extremely small, thereby significantly improving the measurement accuracy of computational spectral measurements and reducing the number of spectral sampling elements required.

[0035] Figure 3 Another improved structure of the spectrum sampling element of the present invention is shown in FIG. Figure 2 Based on the structure shown, a series of annular optical waveguides coupled with the self-coupling optical waveguide are added. These annular optical waveguides and the self-coupling optical waveguide actually form a series of micro-ring resonant cavities. The added annular optical waveguide can be directly coupled with the self-coupling optical waveguide, or indirectly coupled with the self-coupling optical waveguide through other annular optical waveguides. As an additional device, the micro-ring resonant cavity also has a coupling effect with the self-coupling optical waveguide, enabling the backward transmission of the optical signal, further improving the diversity of the optical path and the randomness of the transmission function. The coupling coefficient between the annular optical waveguide and the self-coupling optical waveguide and the radius of the annular optical waveguide can be used as additional design freedom for the transmission function to ensure that the transmission function of each spectral sampling element varies dramatically with wavelength and the correlation between the transmission functions of any two sampling elements is small, thereby making it easier for the spectral sampling element to meet the requirements of computational spectral measurement.

[0036] The spectral sampling element is based on a common waveguide structure, thus eliminating the need for complex fabrication. Furthermore, because it utilizes a multi-point self-coupling waveguide as the primary light transmission path, the optical signal undergoes multiple coupling effects, ensuring that the majority of its energy reaches the output port. This minimizes propagation losses.

[0037] When constructing a computational spectral measurement device using the aforementioned spectral sampling elements, the multiple spectral sampling elements used should satisfy the requirements of ensuring that the transfer function of each spectral sampling element varies as dramatically as possible with wavelength and that the correlation between the transfer functions of any two sampling elements is as minimal as possible. The sampling parameters of the multiple spectral sampling elements (and the aforementioned design degrees of freedom) can be determined through actual experimentation. Alternatively, the sampling parameters can be optimized using multi-objective optimization methods such as simulated annealing, particle swarm optimization, and genetic algorithms, with the goal of minimizing the correlation between the spectral sampling elements while maximizing the randomness of the transfer function of a single spectral sampling element in the wavelength domain. This allows for the use of a minimum number of spectral sampling elements, photodetectors, and other components, while achieving higher spectral measurement accuracy. The correlation between the spectral sampling elements can be measured using the mutual correlation coefficient between the transfer functions, while the randomness of the transfer function of a single spectral sampling element in the wavelength domain can be measured using the number of poles and / or the autocorrelation coefficient.

[0038] In order to verify the effectiveness of the technical solution of the present invention, we conducted a simulation in Lumerical and used the designed spectrum sampling element to simulate the process of measuring the spectrum by a computational spectrometer.

[0039] Let the width of the waveguide be 450nm and the waveguide structure be as follows: Figure 2 There are 12 coupling points in total. To adjust the coupling coefficient k0 at the multi-point self-coupling waveguide coupling point, the length of each coupling point is changed between 0.5μm and 1μm, and the coupling spacing is changed between 0.1μm and 0.2μm. Figure 3 As shown in the figure, six ring optical waveguides with different radii are placed near the multi-point self-coupling waveguide to adjust the coupling coefficient k between each micro-ring resonator and the self-coupling waveguide. i and the radius ri of each micro-ring resonant cavity, the coupling spacing is set to randomly vary between 0.1 μm and 0.2 μm, and the radius ri of each micro-ring resonant cavity is set to randomly take a value between 4 μm and 5 μm.

[0040] By adjusting the design parameters, we can construct M spectrum sampling elements that meet the requirements of the computational spectrometer for sampling the spectrum. In the Lumerical simulation software, the transfer function of a simulated spectrum sampling element is as follows: Figure 4 As shown, it can be seen that it has great randomness in the wavelength domain.

[0041] Assume the spectral range is λ min ~λ max The optical signal to be measured is evenly divided into M paths and then input into each spectral sampling element constructed above. Then, M photodetectors perform photoelectric detection on the output optical signals of the M spectral sampling elements respectively and convert the optical signals into electrical signals I.

[0042] In order to achieve a resolution of Δλ, the spectrometer should have spectral sampling points. Since the transfer function of the spectral sampling element we designed has an extremely small autocorrelation width and can ensure that any two different spectral sampling elements used in the computational spectrometer have extremely low correlation, it is possible to accurately reconstruct the spectrum when N>>M.

[0043] The above process can be expressed mathematically as:

[0044]

[0045] Among them I M×1 =[I1; I2…; I M ], I M Represents the output result of the Mth photodetector; S N×1 =[S1; S2…; S N ], S N It represents the amplitude of the Nth point of the spectrum to be measured, which can be regarded as the unknown number to be solved; H M×N =[H1(λ);H2(λ)…;H M (λ)] is the transmission matrix of M spectral sampling elements, where H i (λ)=[H i1 ,H i2 …,H iN ] is the i-th row of the transfer matrix, that is, the transfer function of the i-th spectral sampling element, i = 1, 2, ..., N; n is the normalization coefficient after measurement calibration.

[0046] The above equations are solved using a convex optimization algorithm. That is, through M equations, N unknowns can be solved and the spectrum can be reconstructed with a small error.

[0047] Compared with traditional spectrometers, which require M = N filters to achieve the target resolution, the broadband filter we designed maintains fewer sampling channels, N>>M, improving the system signal-to-noise ratio and significantly reducing the system size.

[0048] A computational spectrometer with 20 global sampling channels was constructed using the two spectral sampling elements mentioned above. Figure 5As shown in the figure, the spectrum with a bandwidth of 100nm can be accurately reconstructed with a resolution of 0.5nm. The predicted spectrum 1 and predicted spectrum 2 in the figure are based on Figure 2 、 Figure 3 From the reconstructed spectrum obtained by the spectral sampling element, it can be seen that when the parameters of the self-coupled optical waveguide are the same, the spectral reconstruction result of the spectral sampling element structure with the microring resonator added is better.

Claims

1. A spectrum sampling element for global spectrum sampling in computational spectrum measurement, characterized in that: The spectrum sampling element includes a self-coupling optical waveguide having multiple self-coupling regions, wherein the self-coupling regions are formed by bending and folding the self-coupling optical waveguide itself; the spectrum sampling element also includes at least one annular optical waveguide coupled to the self-coupling optical waveguide.

2. The spectrum sampling element according to claim 1, wherein: It is an optical integrated component.

3. A computational spectrum measurement device comprising a plurality of spectrum sampling elements with different transfer functions for respectively performing global spectrum sampling on a light signal to be measured; characterized in that: The spectrum sampling element is the spectrum sampling element according to claim 1 or 2.

4. The computational spectrum measuring device according to claim 3, wherein: The sampling parameters of multiple spectral sampling elements are optimized by a multi-objective optimization method with the goal of minimizing the correlation between the spectral sampling elements and maximizing the randomness of the transfer function of a single spectral sampling element in the wavelength domain.

5. The computational spectrum measuring device according to claim 4, wherein: The correlation is measured as the cross-correlation coefficient.

6. The computational spectrum measuring device according to claim 4, wherein: The randomness of the transfer function in the wavelength domain is measured as the number of poles and / or the autocorrelation coefficient.

7. The computational spectrum measuring device according to claim 4, wherein: The sampling parameters include at least one of the following parameters: the number of self-coupling regions in the self-coupling optical waveguide, the distance between the respective coupling regions in the self-coupling optical waveguide, and the coupling coefficient of the respective coupling regions in the self-coupling optical waveguide.

Images