A focusing method and focusing aid system for a spectrometer

Abstract

The application discloses a focusing method and a focusing auxiliary system of a spectrometer. The focusing method comprises the following steps: controlling a linear gradient filter to change the position in the corresponding degree of freedom compared with a line array detector; acquiring the spectral data detected by the line array detector at different positions in the corresponding degree of freedom, and calculating the evaluation function value of the corresponding position in the corresponding degree of freedom according to the spectral data at different positions in the corresponding degree of freedom; determining the focusing position in the corresponding degree of freedom according to the evaluation function value of the corresponding position in the corresponding degree of freedom, and taking the focusing position corresponding to the last degree of freedom as the first focusing position; wherein the corresponding degrees of freedom comprise six degrees of freedom, the linear gradient filter is arranged on a six-axis free platform, the line array detector and the six-axis free platform are fixed on a base, and the first direction is parallel to the direction of the optical axis. Through the focusing method, the problems of difficult focusing, dependence on artificial experience and low efficiency can be solved.

Description

Technical Field

[0001] This invention relates to the field of spectrometers, and more particularly to a focusing method and focusing auxiliary system for a spectrometer. Background Technology

[0002] In miniature near-infrared spectrometers based on the linear variable filter (LVF) dispersive principle, the relative positional relationship between the dispersive element (LVF) and the photodetector (InGaAs linear array) directly determines the overall spectral performance. Ideally, the different wavelength beams dispersed by the LVF should be precisely focused onto the image plane of the detector, and the light spot of each wavelength should accurately fall on the corresponding detector pixel.

[0003] However, precise focusing between the LVF and the detector is one of the most challenging aspects of instrument manufacturing for several reasons. First, there is the multi-degree-of-freedom coupling problem: the coupling between the LVF and the detector involves multiple spatial degrees of freedom (X / Y / Z translation and θx / θy / θz rotation), and these degrees of freedom interact with each other. For example, adjusting the Z-axis (optical axis) distance affects the focal length and magnification, while adjusting the tilt angle (θx / θy) simultaneously changes the center wavelength range and image quality. This optimization process in multi-dimensional parameter space is complex and time-consuming. Second, there is the invisible operating wavelength range: the near-infrared band (900-1700nm) is beyond the range of human vision (380-780nm), making it impossible to determine the focus status visually. Traditional focusing methods for visible light instruments (such as observing sharpness, star point method, etc.) are not applicable here. Third, there is the requirement for minute tolerances: the depth of focus of a typical miniature spectrometer is only on the order of tens of micrometers (depending on the system's numerical aperture NA). For example, for a system with NA = 0.25, the depth of focus is approximately ±2λ / (NA). 2 ≈±35μm@1300nm. This means that the positioning accuracy in the Z direction needs to be controlled within ±20μm to obtain good image quality. The fourth issue is the lack of objective evaluation standards: current focusing operations largely rely on the experience and feel of technicians, gradually approaching the optimal position by repeatedly collecting spectra and subjectively judging "whether it's better." This method is inefficient, inconsistent, and heavily dependent on individual skill levels. The fifth issue is the bottleneck in mass production: each instrument requires individual focusing adjustment, and if the time taken per instrument is too long (e.g., 1-2 hours), it will seriously affect production efficiency and product costs. Automated, rapid focusing is a prerequisite for achieving mass production. The sixth issue is environmental interference: environmental factors such as temperature changes and mechanical vibrations may cause drift in the adjusted position, requiring a focusing method that can quickly recalibrate. Summary of the Invention

[0004] This invention proposes a focusing method and focusing auxiliary system for spectrometers, which can solve the problems of focusing difficulties, reliance on manual experience, and low efficiency in the prior art. It realizes full automation of the focusing process without manual intervention; the focusing accuracy is better than ±5μm in axial positioning repeatability and better than ±0.05° in angular position repeatability; the single focusing time is less than 5 minutes; the focusing results have quantitative evaluation indicators (such as MTF value, peak intensity, resolution, etc.); it is applicable to InGaAs detectors of different specifications such as 256×1, 512×1, or 512×2; and it can be integrated into the production line to achieve batch automated focusing.

[0005] One embodiment of the present invention proposes a focusing method for a spectrometer, the spectrometer comprising a linear graded filter and a linear detector arranged sequentially along a first direction, the method being applied to focusing between the linear graded filter and the linear detector; the focusing method comprising: controlling the linear graded filter to change position relative to the linear detector in a corresponding degree of freedom; acquiring spectral data detected by the linear detector at different positions in the corresponding degree of freedom; calculating an evaluation function value for the corresponding position in the corresponding degree of freedom based on the spectral data at different positions in the corresponding degree of freedom; determining the focusing position in the corresponding degree of freedom based on the evaluation function value for the corresponding position in the corresponding degree of freedom, and taking the focusing position corresponding to the last degree of freedom as the first focusing position; wherein, the corresponding degree of freedom includes six degrees of freedom, the linear graded filter is configured on a six-axis free stage, the linear detector and the six-axis free stage are fixed on a base, and the first direction is parallel to the optical axis direction.

[0006] Optionally, controlling the linear graduated filter to change position relative to the linear detector in corresponding degrees of freedom includes: controlling the linear graduated filter to change position relative to the linear detector in the following focusing sequence: translation in a first direction, translation in a second direction, translation in a third direction, pitch rotation, yaw rotation, and roll rotation, wherein the second direction is parallel to the horizontal direction, and the first direction, the second direction, and the third direction are perpendicular to each other.

[0007] Optionally, when controlling the linear graduated filter to translate relative to the linear detector along a first direction to change its position, wherein the first direction is parallel to the vertical direction; determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: fitting a curve of multiple evaluation function values ​​changing along the first direction, taking the position of the linear graduated filter in the first direction corresponding to the extreme point of the evaluation function value as the coarse focus position in the first direction; based on a hill-climbing algorithm, controlling the linear graduated filter to translate relative to the linear detector within a first range along the first direction with the coarse focus position in the first direction as the center, to determine the focus position in the first direction, wherein in the degree of freedom of translation in the first direction, the evaluation function value is related to the peak intensity evaluation function value, the total energy evaluation function value, the gradient modulus squared evaluation function value, and the spectral bandwidth evaluation function value.

[0008] Optionally, when controlling the linear graded filter to translate relative to the linear detector along two directions to change its position, determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: taking the position of the linear graded filter in the second direction corresponding to the minimum value of the spectral centroid offset evaluation function as the focus position in the second direction.

[0009] Optionally, when the linear graduated filter is controlled to change position relative to the linear detector along a third direction, the third direction is perpendicular to the first direction, the third direction is perpendicular to the second direction, the second direction is perpendicular to the first direction, and the second direction is parallel to the horizontal direction; determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: taking the position of the linear graduated filter in the third direction corresponding to the maximum value of the total energy evaluation function as the focus position in the third direction.

[0010] Optionally, when controlling the linear graduated filter to change position relative to the linear detector along the pitch angle within the pitch angle range, determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: taking the position of the linear graduated filter corresponding to the minimum value of the evaluation function value or the minimum value of the ratio of edge to center channel signal within the pitch angle range as the focus position of the pitch angle.

[0011] Optionally, when controlling the linear graded filter to change position relative to the linear detector along the yaw angle within the yaw angle range, determining the focus position under the corresponding degree of freedom based on the evaluation function value of the corresponding position under the corresponding degree of freedom includes: taking the position of the linear graded filter corresponding to the maximum value of the wavelength-channel fitting residual within the yaw angle range as the focus position of the yaw angle.

[0012] Optionally, when controlling the linear graded filter to change position relative to the linear detector within the roll angle range, determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: taking the position of the linear graded filter corresponding to the minimum leakage rate of adjacent channels within the roll angle range as the focus position of the roll angle.

[0013] Optionally, after determining the focus position under the corresponding degree of freedom based on the evaluation function value of the corresponding position under the corresponding degree of freedom, and taking the focus position corresponding to the last degree of freedom as the first focus position, the method further includes: continuing to execute the focusing method of the spectrometer described in the embodiment of the present invention in a loop, and when the difference between the evaluation function value corresponding to the j-th focus position and the evaluation function value corresponding to the (j-1)-th focus position is less than a preset value, the j-th focus position is confirmed as the final focus position, where j is greater than or equal to 2 and j is a positive integer.

[0014] Optionally, in the continued cyclic execution of the focusing method of the spectrometer as described in the embodiments of the present invention, the linear gradient filter is controlled to change its position relative to the linear detector in the corresponding degrees of freedom according to the focusing sequence of translation in the first direction, rotation in the pitch angle, and rotation in the yaw angle.

[0015] Optionally, the spectrometer further includes a light source module, a light-transmitting slit, a collimating lens, and a focusing lens arranged sequentially along the beam transmission direction. The linear graded filter is located between the collimating lens and the focusing lens. Before controlling the linear graded filter to change position relative to the linear detector in a corresponding degree of freedom, the focusing method further includes: focusing the light-transmitting slit, the collimating lens, the focusing lens, and the linear detector. The light source module includes a visible light source module, a near-infrared light source module, and a module switching unit. The light-transmitting slit is a pattern projection template. The module switching unit is connected to both the visible light source module and the near-infrared light source module. When the visible light source module is located in the optical path to achieve focusing of the light-transmitting slit, the collimating lens, the focusing lens, and the linear detector, the linear graded filter is placed in the bypass, or the linear graded filter has a visible light extended band. The module switching unit is used to switch the visible light source module to the near-infrared light source module after focusing.

[0016] Another embodiment of the present invention provides a focusing assistance system for a spectrometer, comprising: a light source module, a light-transmitting slit, a collimating lens, a linear graded filter, a focusing lens, a linear detector, and a controller arranged sequentially along the beam transmission direction. The linear graded filter is located on a six-axis free stage, the six-axis free stage and the linear detector are located on a base, and the controller is electrically connected to the six-axis free stage, the linear detector, and the light source module. The controller is used to execute the focusing method of the spectrometer described in any embodiment of the present invention.

[0017] According to embodiments of the present invention, a focusing method and a focusing auxiliary system for a spectrometer are provided. The spectrometer includes a linear graded filter and a linear detector arranged sequentially along a first direction. The method is applied to focusing between the linear graded filter and the linear detector. The focusing method includes: controlling the linear graded filter to change position relative to the linear detector in a corresponding degree of freedom; acquiring spectral data detected by the linear detector at different positions in the corresponding degree of freedom; calculating the evaluation function value of the corresponding position in the corresponding degree of freedom based on the spectral data at different positions in the corresponding degree of freedom; determining the focusing position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom; and taking the focusing position corresponding to the last degree of freedom as the first focusing position. The corresponding degree of freedom includes six degrees of freedom. The linear graded filter is configured on a six-axis free stage, and the linear detector and the six-axis free stage are fixed on a base. The first direction is parallel to the optical axis. Therefore, this focusing method can solve problems such as focusing difficulties, reliance on human experience, and low efficiency, achieving full automation of the focusing process without human intervention; the focusing accuracy is better than ±5μm in axial positioning repeatability and better than ±0.05° in angular position repeatability; the single focusing time is less than 5 minutes; the focusing results have quantitative evaluation indicators (such as MTF value, peak intensity, resolution, etc.); it is applicable to InGaAs detectors of different specifications such as 256×1, 512×1, or 512×2; and it can be integrated into the production line to achieve batch automated focusing. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the focusing auxiliary system of the spectrometer proposed in an embodiment of the present invention; Figure 2 This is a flowchart of the focusing method for the spectrometer proposed in an embodiment of the present invention; Figure 3 This is a graph showing the relationship between the composite evaluation function value and the Z-axis position in the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 4 This is a graph showing the relationship between the evaluation functions and the Z-axis position in the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 5This is a graph showing the relationship between the composite evaluation function and the Z-axis position under the hill-climbing algorithm in the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 6 This is a graph showing the relationship between the step size and the number of iterations in the hill-climbing algorithm of the focusing method for the spectrometer proposed in this embodiment of the invention. Figure 7 This is a graph showing the relationship between the centroid offset and the X-axis offset in the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 8 This is a graph showing the relationship between the total signal energy and the Y-axis offset in the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 9 This is a comparison diagram of signal intensity before and after XY adjustment in the focusing method of the spectrometer proposed in this embodiment of the invention; Figure 10 This is a comparison chart of the relationship between signal intensity and detector channel signal before and after the elevation angle optimization of the focusing method of the spectrometer proposed in this embodiment of the invention; Figure 11 This is a comparison chart of the channel response uniformity before and after optimization of the pitch angle, yaw angle, and roll angle of the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 12 This is a schematic diagram of the focusing auxiliary system of the spectrometer proposed in an embodiment of the present invention; Figure 13 This is a schematic diagram of the pattern projection template in the focusing auxiliary system of the spectrometer proposed in an embodiment of the present invention; Figure 14 This is a flowchart of a focusing method for a spectrometer according to a specific embodiment of the present invention; Figure 15 This is a schematic diagram of the reporting interface of a focusing method for a spectrometer proposed in a specific embodiment of the present invention. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the invention and not all structures. Various modifications and variations can be made to the present invention without departing from its spirit or scope, which will be apparent to those skilled in the art. Therefore, the present invention is intended to cover modifications and variations falling within the scope of the corresponding claims and their equivalents. It should be noted that the embodiments provided in the present invention can be combined with each other without contradiction.

[0020] Figure 1This is a schematic diagram of the focusing auxiliary system of the spectrometer proposed in an embodiment of the present invention. Figure 1 As shown, the spectrometer includes a linear graded filter and a linear detector arranged sequentially along the first direction Z. The method is applied to the focusing between the linear graded filter and the linear detector. The linear graded filter is configured on a six-axis free stage, and the linear detector and the six-axis free stage are fixed on a base. The first direction Z is parallel to the optical axis. By adjusting the six-axis free stage, the linear graded filter and the linear detector can be aligned. It should be noted that the alignment between the linear graded filter and the linear detector means that the wavelength gradient axis of the linear graded filter must be precisely parallel to the pixel row or column direction of the linear detector. Whether it is parallel to the row or column direction depends mainly on the distribution direction of the wavelength gradient axis of the linear graded filter. For example, if the wavelength gradient axis of a linear gradient filter extends horizontally (i.e., the second direction X) and is arranged vertically (i.e., the third direction Y), then the pixels of the linear detector extend horizontally and are arranged vertically (i.e., extending horizontally in the X direction and arranged in the Y direction). Alternatively, if the wavelength gradient axis of the linear gradient filter extends vertically (i.e., the third direction Y) and is arranged horizontally (i.e., the second direction X), then the pixels of the linear detector extend vertically and are arranged horizontally (i.e., extending vertically in the Y direction and arranged in the X direction). The linear detector in this embodiment can be an InGaAs detector.

[0021] The following explains the specific focusing method. Figure 2 This is a flowchart of the focusing method for a spectrometer proposed in an embodiment of the present invention. Figure 2 As shown, the focusing methods include: S101 controls the linear gradient filter to change position relative to the linear detector in the corresponding degrees of freedom.

[0022] The corresponding degrees of freedom include six. That is, the linear graded filter can be controlled to change position in six dimensions. These six degrees of freedom can be translation in three dimensions and rotation along three axes. For example, position changes can be made along the optical axis, such as translation from a base point along the optical axis, or rotation about the optical axis. The step size for each position change can be preset.

[0023] S102, acquire the spectral data detected by the linear detector at different positions under the corresponding degrees of freedom, and calculate the evaluation function value of the corresponding position under the corresponding degrees of freedom based on the spectral data at different positions under the corresponding degrees of freedom.

[0024] In other words, each time the position of the linear graduated filter changes, spectral data detected by the linear detector is acquired. Then, the spectral data is evaluated based on the corresponding evaluation function for that degree of freedom to infer whether the linear graduated filter is at the desired focus position. The evaluation function may differ for each degree of freedom during focusing. For example, the evaluation function is a composite function for translation along the optical axis; for translation along the horizontal axis, it is a spectral centroid shift function; for translation along the vertical axis, it is a total energy evaluation function; for rotation around the horizontal axis, it is a composite evaluation function or a function of the signal ratio between the edge and center channels; for rotation around the vertical axis, it is a wavelength-to-channel fitting residual function; and for rotation around the optical axis, it is an adjacent channel leakage rate function. Thus, when the linear graduated filter is changed in each degree of freedom, the spectral data detected by the linear detector can be judged to meet the focusing requirements using the preset corresponding evaluation function.

[0025] S103, determine the focus position under the corresponding degree of freedom based on the evaluation function value of the corresponding position under the corresponding degree of freedom, and take the focus position corresponding to the last degree of freedom as the first focus position.

[0026] In other words, adjustments are made for each degree of freedom (DOF), with each DDF searching for its optimal position independently. Once the focus position for that DDF is determined, the linear graduated filter is kept unchanged at that DDF. The process then continues for the next DDF, searching for and determining its focus position, until all six DDFs have been adjusted. The focus position determined for the last DDF is the final first focus position obtained in this autofocus process. This method iteratively completes the focus search for each DDF sequentially, eliminating the need to adjust multiple DDFs simultaneously. This reduces algorithm complexity, improves focusing efficiency, and avoids parameter coupling interference caused by simultaneous adjustments of multiple DDFs, ensuring focusing accuracy.

[0027] Optionally, controlling the linear graduated filter to change position relative to the linear detector in corresponding degrees of freedom includes: controlling the linear graduated filter to change position relative to the linear detector in the following focusing sequence: translation in the first direction, translation in the second direction, translation in the third direction, pitch angle rotation, yaw angle rotation, and roll angle rotation, wherein the second direction is parallel to the horizontal direction, and the first direction, the second direction, and the third direction are perpendicular to each other.

[0028] It should be noted that the aforementioned first direction is the optical axis, the second direction is the horizontal direction, and the third direction is the vertical direction. The pitch angle is the angle of rotation around the second direction, the yaw angle is the angle of rotation around the third direction, and the roll angle is the angle of rotation around the first direction. Since the offset of the linear graduated filter in the optical axis direction has a significant impact on the detection of the entire linear detector, the offset in the optical axis direction can be adjusted first. Then, the horizontal direction, the vertical direction, and each rotational degree of freedom are adjusted sequentially. This adjustment sequence conforms to the adjustment logic of correcting large errors before small errors and translation before rotation, which can converge to the optimal focus position more quickly and further improve the efficiency of focus adjustment.

[0029] Optionally, when controlling the linear graduated filter to translate relative to the linear detector along a first direction to change its position, wherein the first direction is parallel to the vertical direction; determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: fitting a curve of multiple evaluation function values ​​changing along the first direction, taking the position of the linear graduated filter corresponding to the extreme point of the evaluation function value in the first direction as the coarse focus position in the first direction; based on the hill climbing algorithm, controlling the linear graduated filter to translate relative to the linear detector within a first range along the first direction with the coarse focus position in the first direction as the center, to determine the focus position in the first direction, wherein in the degree of freedom of translation in the first direction, the evaluation function value is related to the peak intensity evaluation function value, the total energy evaluation function value, the gradient modulus squared evaluation function value, and the spectral bandwidth evaluation function value.

[0030] Specifically, before focusing on the first degree of freedom of translation, initialization and pre-positioning steps are required, namely, mounting the optical module (including LVF) to be assembled on a six-degree-of-freedom precision displacement stage; the displacement stage has: X-axis translational travel ≥ 5 mm / resolution ≤ 1 μm, driven by a stepper motor or piezoelectric; Y-axis translational travel ≥ 5 mm / resolution ≤ 1 μm; Z-axis translation (optical axis direction) travel ≥ 3 mm / resolution ≤ 0.5 μm (most critical); θx-axis rotation (around the X-axis): range ± 5°, resolution ≤ 0.01°; θy-axis rotation (around the Y-axis): range ± 5°, resolution ≤ 0.01°; θz-axis rotation (around the Z-axis): range ± 10°, resolution ≤ 0.02°.

[0031] Next, the InGaAs detector is fixedly mounted on the base, and the circuitry is connected to the readout system and the data acquisition computer. The near-infrared light source (halogen tungsten lamp or LED) is turned on to ensure the optical path operates normally. The six-degree-of-freedom stage is moved to the mechanical zero position or a preset initial position (within ±0.5mm of a roughly estimated optimal area). Then, focusing is performed on the first degree of freedom based on the image quality evaluation function.

[0032] This invention uses the following composite evaluation function as a quantitative indicator of focus quality: (a) Peak intensity evaluation function This refers to the maximum signal strength across all channels. When well-focused, the energy is concentrated, and the peak intensity reaches its maximum. The calculation is simple, but it is sensitive to defocusing. Among these, For channel number, For the first Channel signal strength.

[0033] (b) Total Energy Evaluation Function This is the sum of the signal strengths of all channels. Better focusing results in a greater total light energy entering the effective photosensitive area of ​​the linear detector, thus enhancing its noise immunity. For channel number, For the first Channel signal strength, where N is the total number of channels.

[0034] (c) Gradient modulus squared evaluation function (Tenengrad function). This is the sum of the squares of the differences between adjacent pixels. When well focused, the absorption peaks and other characteristic edges on the spectral curve are steep, with a large gradient. This function is sensitive to subtle changes in focal length and is suitable for fine focusing. For channel number, For the first Channel signal strength, where N is the total number of channels.

[0035] (d) Spectral bandwidth evaluation function ,in The full width at half maximum (FWHM) is used as a reference for the characteristic peak. Better focusing results in narrower peaks and a smaller FWHM. This function has a clear physical meaning and is directly related to spectral resolution.

[0036] (e) Composite evaluation function ,in, , , , The values ​​of each normalization subfunction, Weighting coefficients (default) =0.3, =0.2, =0.3, =0.2).

[0037] Next, a coarse scan is performed in the first direction, i.e., the Z-axis. Here, the other five degrees of freedom are kept constant, only the Z-axis position is changed; within a preset search range (usually within ±1 mm of the initial position), one frame of spectral data is acquired point-by-point with a large step size (e.g., 50 μm); the composite evaluation function value F at each position is calculated, and the FZ curve is plotted (approximately bell-shaped); the curve is fitted using a quadratic polynomial, and the extreme point position Z* is found as the coarse focusing result (e.g., ...). Figure 3 and Figure 4 As shown, Figure 3 This is a graph showing the relationship between the composite evaluation function value and the Z-axis position in the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 4 This is a graph showing the relationship between the evaluation functions and the Z-axis position in the focusing method of the spectrometer proposed in this embodiment of the invention.

[0038] Furthermore, Figure 5 This is a graph showing the relationship between the composite evaluation function and the Z-axis position under the hill-climbing algorithm in the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 6 This is a graph showing the relationship between the step size and the number of iterations in the hill-climbing algorithm of the focusing method for the spectrometer proposed in this embodiment of the invention; as shown... Figure 5 and Figure 6 As shown, fine focusing along the Z-axis is achieved through local optimization based on a hill-climbing algorithm. The search range and step size are narrowed using the extreme point Z* as the center. The search range can be Z* ± 100 μm, and the step size can be 5-10 μm. An improved hill-climbing algorithm is used, with the current position z_cur and the evaluation value F_cur. Specifically, the improved hill-climbing algorithm proceeds as follows: It attempts to move forward one step z_try = z_cur + step, then acquires the spectrum and calculates F_try. If F_try > F_cur, it moves to z_try (i.e., z_cur = z_try, F_cur = F_try), maintaining the same direction, and continues to attempt in the same direction. If F_try ≤ F_cur, it attempts in the opposite direction, z_try = z_cur - step. If F_try > F_cur, it continues to attempt in the opposite direction. If F_try ≤ F_cur, it reduces the step size to step = step / 2 and attempts in the positive direction. The iteration continues until step < 1 μm, and the final z_opt is the optimal position along the Z-axis. Figure 5 The text is for illustrative purposes only. In other words, if the composite evaluation function value F_try corresponding to the trial step is greater than the composite evaluation function value F_cur at the current position, the trial continues in the direction of the previous trial. If the composite evaluation function value F_try corresponding to the trial step is less than or equal to the composite evaluation function value F_cur at the current position, the trial continues in the opposite direction of the previous trial. Furthermore, when the trial steps are repeated in an even number of times, the trial step size is reduced to half of the previous step size.

[0039] Optionally, when controlling the linear graded filter to translate relative to the linear detector along the second direction to change its position, determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: taking the position of the linear graded filter in the second direction corresponding to the minimum value of the spectral centroid offset evaluation function as the focus position in the second direction.

[0040] In this process, based on the Z-axis optimization, the XY positions are optimized to center the spectral energy distribution. Next, optimization is performed in the X-direction (dispersion direction), scanning within a range of ±200 μm from the current position (x_opt). The evaluation metric is changed to minimizing the spectral centroid shift, where the spectral centroid shift is... , N is the weighted centroid channel number, N is the total number of channels, and N / 2 is the geometric center channel number. The optimization objective is to concentrate the spectral energy near the center of the effective region of the linear detector.

[0041] In other words, once the optimized Z-axis position is obtained, the X-axis position can be adjusted. During adjustment, the spectral centroid position (i.e., the weighted centroid channel number) is continuously obtained from the spectral data of the linear detector. Then, the spectral centroid offset is calculated. When the spectral centroid offset is minimized, the current position can be considered the focus position in the X-axis (e.g., ...). Figure 7 As shown, Figure 7 This is a graph showing the relationship between the centroid offset and the X-axis offset of the focusing method of the spectrometer proposed in this embodiment of the invention.

[0042] Optionally, when the linear graded filter is controlled to change position relative to the linear detector along a third direction, the third direction is perpendicular to the first direction, the third direction is perpendicular to the second direction, the second direction is perpendicular to the first direction, and the second direction is parallel to the horizontal direction; determining the focus position in the corresponding degree of freedom according to the evaluation function value of the corresponding position in the corresponding degree of freedom includes: taking the position of the linear graded filter in the third direction corresponding to the maximum value of the total energy evaluation function as the focus position in the third direction.

[0043] After optimizing the X-direction, optimization is performed in the third direction (Y-direction, perpendicular to the dispersion direction and the Z-axis), specifically the slit length direction. Using y_opt as the current position, a point-by-point scan is conducted within ±500μm of the current position. The Y-direction primarily affects the total signal intensity (whether the slit image falls completely within the photosensitive width of the linear detector). This optimization process uses the total energy evaluation function as the criterion: the better the focusing effect, the more effective incident light energy the linear detector receives, and the larger the value of the total energy evaluation function. Therefore, the position in the Y-direction corresponding to the maximum value of the total energy evaluation function is the optimal focusing position in the third direction (e.g., ...). Figure 8As shown, Figure 8 This is a graph showing the relationship between the total signal energy and the Y-axis offset in the focusing method of the spectrometer proposed in this embodiment of the invention.

[0044] After finding the optimal focus position z_opt in the Z direction, the corresponding range is scanned using the current position's x_opt and y_opt as the current position. This reduces the scanning range, decreases the spectral acquisition and calculation of useless positions, and significantly improves the overall efficiency of focus optimization.

[0045] In other embodiments, focusing can be performed first in the X direction, or in the Y direction, and then other degrees of freedom can be adjusted based on the current position after focusing.

[0046] Figure 9 This is a comparison diagram of signal intensity before and after XY adjustment in the focusing method of the spectrometer proposed in this embodiment of the invention; as shown. Figure 9 As shown, for the XY plane, the adjusted centroid is more central than the original centroid.

[0047] Optionally, when controlling the linear graduated filter to change position relative to the linear detector along the pitch angle within the pitch angle range, determining the focus position under the corresponding degree of freedom based on the evaluation function value of the corresponding position under the corresponding degree of freedom includes: taking the position of the linear graduated filter within the pitch angle range corresponding to the minimum value of the evaluation function value or the minimum value of the ratio of edge to center channel signal as the focus position of the pitch angle.

[0048] After focusing in the X, Y, and Z directions, tilt optimization can be performed to eliminate astigmatism and distortion. θx affects the effective wavelength coverage of the LVF and the imaging quality of the edge channels. When performing θx optimization, i.e., tilt angle optimization, a scan is performed within ±2° of the current position, with a step size of 0.1°. The evaluation index during the optimization process is the minimization of the overall F-value (calculated using all channel data) or the edge / center channel signal ratio. Figure 10 This is a comparison chart of the relationship between signal intensity and detector channel signal before and after elevation angle optimization for the focusing method of the spectrometer proposed in this embodiment of the invention. Figure 10 As shown, after θx correction, compared to before correction, the left edge channel is improved by -18%, and the right edge channel is improved by +457%.

[0049] Optionally, when controlling the linear graduated filter to change position relative to the linear detector along the yaw angle within the yaw angle range, determining the focus position under the corresponding degree of freedom based on the evaluation function value of the corresponding position under the corresponding degree of freedom includes: taking the position of the linear graduated filter corresponding to the maximum value of the wavelength-channel fitting residual within the yaw angle range as the focus position of the yaw angle.

[0050] θy affects the spectral linearity of the dispersion direction. After focusing in the X, Y, and Z directions, and after θx is calibrated, θy optimization (i.e., yaw angle optimization) is performed. During θy optimization, a scan is performed within ±1° of the current position, with a scan step size of 0.05°. The evaluation metric during optimization is the wavelength-channel fitting residual R0. 2 Maximize (requires the use of characteristic spectral lines from a reference light source).

[0051] Optionally, when controlling the linear gradient filter to change position relative to the linear detector along the roll angle within the roll angle range, determining the focus position under the corresponding degree of freedom based on the evaluation function value of the corresponding position under the corresponding degree of freedom includes: taking the position of the linear gradient filter corresponding to the minimum leakage rate of adjacent channels within the roll angle range as the focus position of the roll angle.

[0052] θz affects the alignment between the slit image direction and the pixel array of the linear detector. After focusing in the X, Y, and Z directions, and θx and θy, θz optimization, i.e., roll angle optimization, is performed. During θz optimization, a scan is performed within ±3° of the current position, with a scan step size of 0.1°. The evaluation metric during optimization is minimizing crosstalk between channels (which can be assessed by measuring the leakage rate of adjacent channels under a narrowband light source).

[0053] After optimizing and adjusting all degrees of freedom, the optimal position and angle parameters of each axis are stored and used as the final focusing parameters for the linear graded filter relative to the linear detector, thus completing the autofocus process. Compared to traditional manual focusing, which relies on operator experience and suffers from poor consistency and low efficiency, this multi-degree-of-freedom autofocus method optimizes each degree of freedom sequentially and combines targeted evaluation functions corresponding to different degrees of freedom to quickly and stably achieve high-precision focusing, ensuring the energy utilization and spectral detection accuracy of the spectral imaging system.

[0054] Figure 11 This is a comparison chart of the channel response uniformity of the focusing method of the spectrometer proposed in this embodiment of the invention before and after optimization of the pitch angle, yaw angle, and roll angle; as shown. Figure 11 As shown, after optimizing θx, θy, and θz, the uniformity of the channel response was improved.

[0055] Optionally, after determining the focus position under the corresponding degree of freedom based on the evaluation function value of the corresponding position under the corresponding degree of freedom, and taking the focus position corresponding to the last degree of freedom as the first focus position, the method further includes: continuing to execute the focusing method of the spectrometer of the present invention in a loop, and when the difference between the evaluation function value corresponding to the j-th focus position and the evaluation function value corresponding to the (j-1)-th focus position is less than a preset value, the j-th focus position is confirmed as the last focus position, where j is greater than or equal to 2 and j is a positive integer.

[0056] In other words, due to the coupling between the degrees of freedom, multiple rounds of iterative optimization are required, typically 2-3 rounds of global iteration. In one embodiment, the focusing method of the spectrometer as described in this embodiment of the invention is continuously executed in a loop, controlling the linear gradient filter to change position relative to the linear detector in the corresponding degrees of freedom according to the focusing sequence of translation in the first direction, rotation in the pitch angle, and rotation in the yaw angle.

[0057] For example, Round 1: Z→X→Y→θx→θy→θz; Round 2: Z→θx→θy (XY and θz generally do not change significantly); Round 3: Z fine-tuning (final precise focusing). The evaluation function value is recorded after each round. The iteration terminates when the improvement in the F-value between two adjacent rounds is less than 0.5%. In other words, after the first round of focusing, the focus position corresponding to the last degree of freedom can be taken as the final focus position. Alternatively, several more iterations can be repeated according to the first round of focusing to obtain the final focus position. Since XY and θz generally do not change significantly, iteration can also be performed in the manner of Z→θx→θy, thereby reducing the number of degrees of freedom that need to be adjusted during the iteration process and further improving the focusing optimization efficiency. Multiple iterations can effectively eliminate optimization errors caused by coupling between different degrees of freedom, allowing the final focusing parameters to approach the optimal state, further improving focusing accuracy while ensuring efficiency.

[0058] Then, after finding the focus position, the displacement stage can be locked in the optimal position (mechanical locking or UV adhesive curing); the verification spectrum is collected to confirm that all indicators meet the requirements; a focus report is generated, which includes the optimal position coordinates (values ​​of 6 degrees of freedom), the final evaluation value F_final, various performance indicators (peak intensity, total energy, half bandwidth, etc.), comparison pictures before and after focusing, and the qualification judgment conclusion.

[0059] Optionally, Figure 12 This is a schematic diagram of the focusing auxiliary system of the spectrometer proposed in an embodiment of the present invention; as shown. Figure 12As shown, the spectrometer also includes a light source module, a light-transmitting slit, a collimating lens, and a focusing lens arranged sequentially along the beam transmission direction. A linear graded filter is located between the collimating lens and the focusing lens. Before controlling the linear graded filter to change position relative to the linear detector in the corresponding degrees of freedom, the focusing method further includes: focusing the light-transmitting slit, collimating lens, focusing lens, and linear detector. The light source module includes a visible light source module, a near-infrared light source module, and a module switching unit. The light-transmitting slit serves as a pattern projection template. The module switching unit is connected to both the visible light source module and the near-infrared light source module. When the visible light source module is located in the optical path to achieve focusing of the light-transmitting slit, collimating lens, focusing lens, and linear detector, the linear graded filter is placed in the bypass, or the linear graded filter has a visible light extension band. The module switching unit is used to switch the visible light source module to the near-infrared light source module after focusing.

[0060] The visible light source module can be a visible light laser module with a wavelength of 650nm (red light, easy to observe with the naked eye), a power of less than 5mW (Class 3R safety level), and a TEM beam mode. It is a Gaussian beam with a collimated beam diameter of 1-2mm, installed at the source location or as a substitute for the source. A variable density attenuation filter group is used to attenuate the laser power to within the safe operating range of the linear detector. Neutral density ND2-ND4 is optional; InGaAs still has a weak response at 650nm (approximately 1% peak response). The variable density attenuation filter can be integrated into the visible light laser module. The module switching unit can be an electrically operated flipping mechanism. In the working position, it acts as the laser injection path; in the retracted position, the laser is removed, restoring normal near-infrared illumination. The switching time is less than 1 second. The pattern projection template (optional enhancement function) is a special mask placed at the entrance slit position, containing known geometric patterns (such as cross lines, equidistant stripes, pinhole arrays, etc.). The focus status is determined by observing the image of these patterns on the linear detector.

[0061] Figure 13 This is a schematic diagram of the pattern projection template in the focusing auxiliary system of the spectrometer proposed in this embodiment of the invention; as shown... Figure 13As shown, the partitions of the scribing mask (typically 15mm×12mm) are arranged as follows: Area A: Crosshairs (top left) used to detect XY plane offset and locate the center through the crosshairs; Area B: Equally spaced stripes (top center) used to evaluate the dispersion linearity of the stripes and the known wavelength difference of the observed wavelength; Area C: Pinhole array (top right) used to measure MTF and resolution point spread function (PSF) analysis; Area D: Concentric circles (left center) used to detect roll angle and circle center offset, reflecting rotational deviation; Area E: USAF resolution target (middle center) used to quantify spatial resolution and detect different line pairs corresponding to different frequencies; Area F: Checkerboard (right center) used to detect geometric distortion, distortion, and field curvature aberrations; Area G: Multi-width slit group (bottom left) with different widths used for sensitivity calibration to simulate different spectral resolution requirements; Area H: Radiation star map (bottom right) used for all-around characteristic detection to verify the consistency of imaging in all directions.

[0062] In one embodiment, the focusing assistance process (using laser assistance) is as follows: the visible light laser is flipped into the optical path, and the near-infrared main light source is turned off; the attenuated laser passes through a slit, collimating lens, LVF, and focusing lens to form a spot or pattern on a linear detector (if the slit is a pattern projection template, a pattern is formed). Since 650nm is located at the short end of the LVF passband (if the LVF starts from 900nm, then 650nm is cut off), the following two strategies can be adopted at this time. The first is to use an extended band LVF (such as 600-1700nm) for focusing, that is, to customize a "tuning-grade LVF" specifically for focusing, whose passband extends to the visible light range; after focusing is completed, it is replaced with a formal near-infrared LVF (ensuring that the thickness of both is consistent). The second method involves setting up a removable LVF support, temporarily removing the LVF during focusing; directly using a 4f system consisting of a collimating lens and a focusing lens to image the slit onto the linear detector; when the slit image is at its sharpest, the collimating lens, focusing lens, and linear detector have achieved confocal focus; then the LVF is placed back in and fine-tuned. The high contrast of the laser spot / pattern is utilized to accelerate the calculation of the evaluation function; after focusing is complete, the laser is flipped back and the system returns to normal operation.

[0063] Therefore, the focusing method proposed in this invention can achieve the following beneficial effects: Fully automated focusing process: From initialization to locking, no manual intervention is required, eliminating the influence of human factors and ensuring consistency and repeatability of each focusing attempt. Scientific and quantitative evaluation system: A composite evaluation function and multi-index comprehensive judgment are adopted to avoid ambiguity and subjective arbitrariness in the focusing results; each focusing step has a clear quantitative basis. Efficient multi-dimensional collaborative optimization: A reasonable optimization sequence and multi-round iteration strategy are designed for the coupling characteristics of the six degrees of freedom, achieving the optimal focusing state with the fewest position adjustments. Significantly reduced focusing time: Compared to the 1-2 hours required for traditional manual focusing, the automatic focusing of this method can be completed within 5 minutes, improving efficiency by more than an order of magnitude. Applicable to mass production: The automation characteristics allow this method to be directly integrated into the production line, achieving unmanned batch assembly and adjustment in conjunction with conveyor mechanisms and automatic loading and unloading equipment. Laser assistance improves reliability: The introduction of visible light lasers provides additional focusing means, especially when the initial state deviates significantly, it can quickly find the approximate focus position, reducing the global search time. Complete document traceability capabilities: Automatically generated focus reports provide complete data support for quality control, meeting the traceability requirements of industrial production.

[0064] It should be noted that the focusing method proposed in this embodiment of the invention is applicable to focusing during the assembly and packaging process of a spectrometer.

[0065] In one specific embodiment, the device configuration includes: a six-DOF stage: a Physik Instrumente (PI) H-811 series six-legged stage; a controller: a PIC-887 digital controller; a data acquisition card: an NIUSB-6363; software: an autofocus program developed using LabVIEW; a linear detector: a 256×1 InGaAs (SU256-L-1.7RT); and LVF specifications: 15mm long, 950-1650nm.

[0066] The execution process is shown in Table 1 (in conjunction with...) Figure 14 It can be seen that, Figure 14 (This is a flowchart of a focusing method for a spectrometer according to a specific embodiment of the present invention). Table 1 The focusing results are as follows: the composite evaluation function F value increased by 42% (relative to the initial position), the peak signal intensity increased from 12000DN to 18500DN, the average half bandwidth of the spectrum decreased from 12.3nm to 7.8nm, and the edge channel (960nm) signal increased from 58% of the center channel to 76%.

[0067] Based on Example 1, a focusing assistance system is enabled: using a 650nm laser and a bypass LVF, 4f system focusing is first completed under visible light (approximately 1 minute). After placing the LVF, only Z-axis and θx / θy fine adjustments are needed (approximately 2 minutes), reducing the total time to approximately 3 minutes. Traditional manual focusing methods, operated by skilled technicians relying on experience and visual observation to acquire the spectral shape, take an average of 1.5 hours. The standard deviation of the results varies between different technicians / times, approximately ±15μm (Z-axis) and ±0.3° (angle), requiring a specialized darkroom environment and auxiliary tools. Therefore, it can be seen that the focusing method of this invention significantly improves both efficiency and accuracy compared to traditional focusing methods.

[0068] Another embodiment of the present invention provides a focusing assistance system for a spectrometer, comprising: a light source module, a light-transmitting slit, a collimating lens, a linear graded filter, a focusing lens, a linear detector, and a controller arranged sequentially along the beam transmission direction. The linear graded filter is located on a six-axis free stage, the six-axis free stage and the linear detector are located on a base, and the controller is electrically connected to the six-axis free stage, the linear detector, and the light source module. The controller is used to execute the focusing method of the spectrometer described in any embodiment of the present invention.

[0069] The focusing auxiliary system based on the above spectrometer implements the following focusing method: (a) the LVF is mounted on a six-degree-of-freedom precision displacement stage, the linear detector is fixedly mounted, the near-infrared light source is turned on, and the system is moved to the initial position; (b) a coarse scan of the Z-axis is performed, and the coarse focusing position of the Z-axis is determined based on the composite evaluation function; (c) fine climbing optimization of the Z-axis is performed with the coarse focusing position as the center to determine the optimal Z-axis position; (d) the XY plane is centered to make the spectral energy distribution in the center of the effective area of ​​the linear detector; (e) the θx / θy / θz three-axis tilt angle is optimized to eliminate astigmatism and distortion; (f) 2-3 rounds of multi-degree-of-freedom iterative optimization are performed until the evaluation function converges; (g) the optimal position is locked and a focusing report is generated (e.g., ...). Figure 15 As shown, Figure 15This is a schematic diagram of the reporting interface of the focusing method of the spectrometer proposed in a specific embodiment of the present invention. The fine hill-climbing optimization in step (c) adopts an improved hill-climbing algorithm: it moves tentatively along the current gradient direction; if the evaluation value improves, it continues to move in the same direction; otherwise, it tries in the opposite direction. The step size is halved after each extreme point encountered, and iteration stops when the step size is less than 1 μm. In step (d), the XY plane centering adjustment uses the distance between the spectral weighted centroid channel number and the center channel number of the linear detector as the evaluation index in the X direction, and the total signal energy as the evaluation index in the Y direction. The device includes a visible light laser (650 nm), a variable density attenuation filter group, and an electric flipping mechanism; the laser is flipped into the optical path during focusing and can selectively bypass the LVF, utilizing the high contrast characteristics of visible light to accelerate the focusing process. The focusing auxiliary device also includes a pattern projection template placed at the entrance slit position. The template has cross-shaped lines, equally spaced stripes, or pinhole array patterns. The focusing status is aided by observing the imaging clarity of these patterns on the linear detector.

[0070] In the focusing assist system, the controller can integrate a data acquisition unit connected to the output of the InGaAs detector; the controller can also integrate a computation control unit to run the focusing algorithm and output the displacement stage motion command; the computation control unit uses focusing software developed with LabVIEW, Python or C++ to provide a graphical interface to display the real-time evaluation function curve and optimization progress, and automatically generates a report file containing the optimal coordinates, evaluation value and pass / fail judgment after focusing is completed.

[0071] It should be noted that in this embodiment of the invention, X-axis translation is a lateral shift along the dispersion direction, mainly affecting the coverage of the spectral wavelength range. The final optimized X-axis result ensures uniform signal distribution in each channel. Y-axis translation is a shift along the slit length direction, mainly affecting the total light energy entering the detector. The final optimized Y-axis result ensures that the slit image falls completely within the detector's photosensitive area. Z-axis translation is a distance adjustment along the optical axis, directly determining the imaging sharpness / depth of field ±35μm. The θx pitch angle, the rotation angle around the X-axis, mainly affects the effective wavelength coverage of the LVF, and the optimized result improves the uniformity of edge channel signals. The θy yaw angle, the rotation angle around the Y-axis, mainly affects the spectral linearity in the dispersion direction, and the optimized result reduces wavelength-channel nonlinearity error. The θz rotation angle, the rotation angle around the Z-axis, mainly affects the alignment of the slit image with the pixel column, and the optimized result reduces inter-channel crosstalk.

[0072] In summary, according to the focusing method and focusing auxiliary system of the spectrometer proposed in the embodiments of the present invention, the spectrometer includes a linear graded filter and a linear detector arranged sequentially along a first direction. The method is applied to focusing between the linear graded filter and the linear detector. The focusing method includes: controlling the linear graded filter to change position relative to the linear detector in a corresponding degree of freedom; acquiring spectral data detected by the linear detector at different positions in the corresponding degree of freedom; calculating the evaluation function value of the corresponding position in the corresponding degree of freedom based on the spectral data at different positions in the corresponding degree of freedom; determining the focusing position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom; and taking the focusing position corresponding to the last degree of freedom as the first focusing position. The corresponding degree of freedom includes six degrees of freedom. The linear graded filter is configured on a six-axis free stage, and the linear detector and the six-axis free stage are fixed on a base. The first direction is parallel to the optical axis. Therefore, this focusing method can solve problems such as focusing difficulties, reliance on human experience, and low efficiency, achieving full automation of the focusing process without human intervention; the focusing accuracy is better than ±5μm in axial positioning repeatability and better than ±0.05° in angular position repeatability; the single focusing time is less than 5 minutes; the focusing results have quantitative evaluation indicators (such as MTF value, peak intensity, resolution, etc.); it is applicable to InGaAs detectors of different specifications such as 256×1, 512×1, or 512×2; and it can be integrated into the production line to achieve batch automated focusing.

[0073] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A focusing method for a spectrometer, characterized in that, The spectrometer includes a linear graded filter and a linear detector arranged sequentially along a first direction, and the method is applied to focusing between the linear graded filter and the linear detector; The focusing method includes: The linear gradient filter is controlled to change position relative to the linear detector in the corresponding degrees of freedom; Acquire the spectral data detected by the linear detector at different positions under the corresponding degrees of freedom, and calculate the evaluation function value of the corresponding position under the corresponding degrees of freedom based on the spectral data at different positions under the corresponding degrees of freedom; The focus position under the corresponding degree of freedom is determined based on the evaluation function value of the corresponding position under the corresponding degree of freedom, and the focus position corresponding to the last degree of freedom is taken as the first focus position. The corresponding degrees of freedom include six degrees of freedom. The linear gradient filter is configured on a six-axis free stage. The linear array detector and the six-axis free stage are fixed on a base. The first direction is parallel to the optical axis.

2. The focusing method for the spectrometer according to claim 1, characterized in that, The control of the linear gradient filter to change position relative to the linear detector in the corresponding degree of freedom includes: The linear gradient filter is controlled to change position relative to the linear detector in the following order of focusing: translation in the first direction, translation in the second direction, translation in the third direction, pitch angle rotation, yaw angle rotation, and roll angle rotation. The second direction is parallel to the horizontal direction, and the first direction, the second direction, and the third direction are perpendicular to each other.

3. The focusing method for the spectrometer according to claim 1 or 2, characterized in that, When the linear gradient filter is controlled to translate relative to the linear detector along the first direction to change its position, determining the focus position in the corresponding degree of freedom based on the evaluation function value of the corresponding position in the corresponding degree of freedom includes: Fit a curve of multiple evaluation function values ​​changing along the first direction, and take the position of the linear gradient filter corresponding to the extreme point of the evaluation function value in the first direction as the coarse focus position in the first direction; Based on the hill-climbing algorithm, the linear gradient filter is controlled to translate relative to the linear detector within a first range along the first direction with the coarse focusing position in the first direction as the center, to determine the focusing position in the first direction. Under the degree of freedom of translation in the first direction, the evaluation function value is related to the peak intensity evaluation function value, the total energy evaluation function value, the gradient modulus squared evaluation function value, and the spectral bandwidth evaluation function value.

4. The focusing method of the spectrometer according to claim 1 or 2, characterized in that, When the linear gradient filter is controlled to change position relative to the linear detector along the second direction... The process of determining the focus position in the corresponding degree of freedom based on the evaluation function value at the corresponding position in the corresponding degree of freedom includes: The position of the linear graded filter in the second direction corresponding to the minimum value of the spectral centroid offset evaluation function is taken as the focus position in the second direction.

5. The focusing method of the spectrometer according to claim 1 or 2, characterized in that, When the linear gradient filter is controlled to change position relative to the linear detector along a third direction, the third direction is perpendicular to the first direction, the third direction is perpendicular to the second direction, the second direction is perpendicular to the first direction, and the second direction is parallel to the horizontal direction; The process of determining the focus position in the corresponding degree of freedom based on the evaluation function value at the corresponding position in the corresponding degree of freedom includes: The position of the linear gradient filter corresponding to the maximum value of the total energy evaluation function in the third direction is taken as the focus position in the third direction.

6. The focusing method of the spectrometer according to claim 1 or 2, characterized in that, When the linear gradient filter is controlled to change position relative to the linear detector within a range of pitch angles... The process of determining the focus position in the corresponding degree of freedom based on the evaluation function value at the corresponding position in the corresponding degree of freedom includes: The position of the linear gradient filter corresponding to the minimum value of the evaluation function or the minimum value of the ratio of the edge to the center channel signal within the pitch angle range is taken as the focus position of the pitch angle.

7. The focusing method for a spectrometer according to claim 1 or 2, characterized in that, When the linear graded filter is controlled to change position relative to the linear detector within a yaw angle range, The process of determining the focus position in the corresponding degree of freedom based on the evaluation function value at the corresponding position in the corresponding degree of freedom includes: The position of the linear graded filter corresponding to the maximum value of the wavelength-channel fitting residual within the yaw angle range is taken as the focus position of the yaw angle.

8. The focusing method of the spectrometer according to claim 1 or 2, characterized in that, When the linear gradient filter is controlled to change position relative to the linear detector within a roll angle range, The process of determining the focus position in the corresponding degree of freedom based on the evaluation function value at the corresponding position in the corresponding degree of freedom includes: The position of the linear gradient filter corresponding to the minimum leakage rate of adjacent channels within the roll angle range is taken as the focus position of the roll angle.

9. The focusing method of the spectrometer according to claim 1 or 2, characterized in that, After determining the focus position for each degree of freedom based on the evaluation function value at the corresponding position, and taking the focus position corresponding to the last degree of freedom as the first focus position, the process also includes: The focusing method of the spectrometer as described in claim 1 is continued to be executed in a loop. When the difference between the evaluation function value corresponding to the j-th focusing position and the evaluation function value corresponding to the (j-1)-th focusing position is less than a preset value, the j-th focusing position is confirmed as the final focusing position, where j is greater than or equal to 2 and j is a positive integer.

10. The focusing method for a spectrometer according to claim 9, characterized in that, In the continued cyclic execution of the focusing method of the spectrometer as described in claim 1, the linear gradient filter is controlled to change position relative to the linear detector in the corresponding degrees of freedom according to the focusing sequence of translation in the first direction, rotation in the pitch angle, and rotation in the yaw angle.

11. The focusing method of the spectrometer according to claim 1, characterized in that, The spectrometer further includes a light source module, a light-transmitting slit, a collimating lens, and a focusing lens arranged sequentially along the beam transmission direction. The linear graded filter is located between the collimating lens and the focusing lens. Before controlling the linear graded filter to change position relative to the linear detector in the corresponding degrees of freedom, the focusing method further includes: The light-transmitting slit, the collimating lens, the focusing lens, and the linear detector are focused; The light source module includes a visible light source module, a near-infrared light source module, and a module switching unit. The light-transmitting slit is a pattern projection template. The module switching unit is connected to both the visible light source module and the near-infrared light source module. When the visible light source module is located in the optical path to enable focusing of the light-transmitting slit, the collimating lens, the focusing lens, and the linear detector, the linear gradient filter is placed in the bypass, or the linear gradient filter has a visible light extended band. The module switching unit is used to switch the visible light source module to the near-infrared light source module after focusing.

12. A focusing auxiliary system for a spectrometer, characterized in that, include: The spectrometer comprises a light source module, a light-transmitting slit, a collimating lens, a linear graded filter, a focusing lens, a linear detector, and a controller, arranged sequentially along the beam transmission direction. The linear graded filter is located on a six-axis free stage, and the six-axis free stage and the linear detector are located on a base. The controller is electrically connected to the six-axis free stage, the linear detector, and the light source module, respectively. The controller is used to execute the focusing method of the spectrometer as described in any one of claims 1-11.

Images