Multi-source disturbance self-calibration method and system for rotary filtering type monochromator

Abstract

The invention relates to a multi-source disturbance self-calibration method and system for a rotary filtering type monochromator, and belongs to the technical field of optical instrument filtering calibration. The method comprises the following steps: converting a multi-source disturbance signal response characteristic into a disturbance energy spectrum basis vector, and extracting spectrum bandwidth distribution to obtain a spectrum self-disturbance calibration parameter; luminous flux preposition constraint is carried out, the convergence state of a real-time spectrum observation value under the disturbance change condition is detected, confidence coefficient calibration is carried out on the disturbance contribution weight on which the self-calibration operation depends, and a filtering disturbance compensation control strategy is generated; constructing a filtering performance evolution model, identifying spectral wavelength shift and resolution data after disturbance self-calibration, verifying the filtering degradation period of the monochromator, and outputting a filtering calibration feedback report; and setting a critical interference stable threshold, establishing a filtering virtual regulation and control mechanism, enhancing the compensation calibration weight of a filtering disturbance compensation control strategy, and generating a filtering disturbance self-calibration scheme.

Description

Technical Field

[0001] This invention belongs to the field of optical instrument filter calibration technology, specifically relating to a multi-source perturbation self-calibration method and system for a rotating filter monochromator. Background Technology

[0002] With the widespread application of high-precision spectral measurement technology in industrial testing, environmental monitoring, biomedical analysis, and precision manufacturing, the requirements for measurement stability and long-term reliability of spectrometer systems are constantly increasing. However, in actual operation, spectral acquisition systems are susceptible to multi-source disturbances, including changes in ambient temperature and humidity, mechanical vibration, fluctuations in light source output, and aging of filter devices. These factors can cause spectral center wavelength drift, bandwidth changes, resolution degradation, and distortion of spectral microstructure, leading to measurement results deviating from the true values. Existing technologies mostly employ periodic manual calibration or compensation methods based on a single error model, which are difficult to accurately characterize the spectral structure changes under the coupling effect of multi-source disturbances and cannot dynamically sense and predict filter degradation trends. Furthermore, traditional calibration methods typically focus on static error correction and lack the ability to model the evolution characteristics of spectral performance over time, failing to achieve adaptive adjustment under continuously changing disturbances. This results in decreased measurement consistency and increased maintenance costs during long-term operation. Therefore, there is an urgent need for a spectral measurement method that can sense multi-source disturbances, characterize the evolution of spectral performance, and achieve self-calibration and adaptive compensation control to improve the stability and intelligence level of spectral systems. Summary of the Invention

[0003] To address the aforementioned problems in the prior art, this invention provides a multi-source perturbation self-calibration method for a rotating filter monochromator. The objective of this invention can be achieved through the following technical solutions: S1: Acquire multi-source perturbation signal response data, construct a filter mapping perturbation response architecture, convert the multi-source perturbation signal response features into perturbation energy spectrum basis vectors, extract the spectral bandwidth distribution through a spectral network decomposition algorithm, store the spectral bandwidth distribution in an equivalent distortion reference spectral library with perturbation labels, and obtain spectral self-perturbation calibration parameters. S2: Based on the spectral self-perturbation calibration parameters, perform pre-constraint of light flux, input the perturbation sensing spectral shape inversion function, detect the convergence state of real-time spectral observations under perturbation change conditions, call the self-calibration operation dependency of multi-source perturbation signal response data, calibrate the confidence level of the perturbation contribution weight of the self-calibration operation dependency, and generate a filter perturbation compensation control strategy. S3: For the filter perturbation compensation control strategy, construct a filter performance evolution model, identify the spectral wavelength shift and resolution data after perturbation self-calibration, and perform similarity matching with historical degradation samples in the equivalent distortion reference spectrum library based on the time-series feature alignment algorithm to verify the monochromator filter degradation period and output a filter calibration feedback report. S4: Based on the filter calibration feedback report, analyze the filter stability margin of the spectral self-disturbance calibration parameters, set the critical interference stability threshold, establish a filter virtual control mechanism, and when the change in the disturbance projection approaches the critical interference stability threshold, enhance the compensation calibration weight of the filter disturbance compensation control strategy to generate a filter disturbance self-calibration scheme.

[0004] Specifically, the method for constructing the filter mapping perturbation response architecture is as follows: Based on multi-source perturbation signal response data, which includes environmental perturbation response data and light source operating state perturbation response data, the environment and filter perturbations are uniformly parameterized and encoded to generate a time-calibrated set of perturbation feature vectors. Then, combined with the spectral response sequence corresponding to the perturbation feature vector set, cross-band correction decomposition is performed to map the spectral response sequence to the spectral state tensor domain. Wavelength drift components and resolution attenuation components caused by different perturbation sources are extracted to construct the filter mapping perturbation response architecture.

[0005] Specifically, the process of extracting the spectral bandwidth distribution using the spectral network decomposition algorithm is as follows: multidimensional numerical field reconstruction is performed on the local spectral intensity, phase, and noise characteristics of the perturbation energy spectrum basis vector to simulate the multi-source perturbation vector field within the rotating filter monochromator. Based on the network decomposition control algorithm, the local harmonic intensity and spectral index within the perturbation vector field are calculated. Using the spectral resolution as the perturbation label, the bandwidth distribution is decomposed and adjusted to obtain the spectral bandwidth distribution of the multi-source perturbation signal response data.

[0006] Specifically, the equivalent distortion reference spectral library adopts a hierarchical spectral state coding structure, including a spectral energy distribution layer and a resolution attenuation feature layer. Perturbation source labels and time calibration information are introduced into the features of each layer to generate spectral distortion state vectors. Equivalent mapping and clustering are performed on the spectral responses under different perturbation combinations to form perturbation equivalent subspace clusters. A corresponding set of distortion basis vectors is established for each subspace cluster to store the spectral structure evolution mode under different filter degradation stages and perturbation intensity levels, thereby obtaining spectral self-perturbation calibration parameters.

[0007] Specifically, before the perturbation-sensing spectral shape inversion function is executed, the spectral self-perturbation calibration parameters are projected onto the spectral flux control space, a suppression weight is applied to the energy adjustment amplitude corresponding to the high-frequency spectral shape component, and a compensation weight is applied to the low-frequency energy drift component to correct the spectral shape information change caused by spectral flux adjustment, thereby obtaining the filter pre-constraint value.

[0008] Specifically, the perturbation-sensing spectral shape inversion function takes the filter pre-constraint value as the input variable and outputs an initial spectral shape state vector. The initial spectral shape state vector includes wavelength drift components and bandwidth compression components. Based on the spectral self-perturbation calibration parameters, the initial spectral shape state vector is mapped to local spectral shape response nodes, and the spectral shape changes in different wavelength segments are associated with the corresponding perturbation source contribution weights. The spectral shape inversion coefficients and non-steady-state sensing parameters are recursively corrected, and the convergence state is detected.

[0009] Specifically, the method for generating the filter perturbation compensation control strategy is as follows: calling the self-calibration operation dependency of the multi-source perturbation signal response data, reconstructing the weighted coupling relationship of the response of different perturbation sources in each frequency band of the spectrum, constructing the spectral shift contribution matrix, and combining the spectral domain orthogonal projection algorithm to analyze the spectral shift contribution matrix of the current filter state, calculating the perturbation energy spectrum basis vector in the calibration confidence interval, performing short-time perturbation pre-compensation on the temporal drift of the perturbation energy spectrum basis vector, and generating the filter perturbation compensation control strategy.

[0010] Specifically, the method for constructing the filter performance evolution model is as follows: The spectral observation sequence after spectral self-perturbation calibration is mapped to the multidimensional spectral state tensor domain, which includes: wavelength drift gradient dimension and spectral resolution attenuation dimension. Degraded samples in the equivalent distortion reference spectral library are time-calibrated to obtain the spectral filtering evolution trajectory. Based on the spectral filtering evolution trajectory, the perturbation evolution sequence of the current multidimensional spectral state tensor is calculated to construct a filtering performance evolution model containing the spectral domain evolution structure. The spectral domain evolution structure includes a frequency domain branch and a time domain branch; the frequency domain branch is used to control the spectral bandwidth compression and center wavelength shift range, and the time domain branch is used to characterize the resolution decay rate and recovery inertia parameter under disturbance triggering.

[0011] Specifically, the temporal feature alignment algorithm limits the local deformation amplitude of the perturbation alignment path based on the self-calibrated spectral wavelength shift and resolution data, and introduces a perturbation sensitivity penalty factor during the calibration path search process to adaptively adjust the matching cost of the perturbation mutation segment, and outputs the corresponding degradation stage matching result.

[0012] Specifically, the output process of the filter calibration feedback report is as follows: Based on historical degradation samples in the equivalent distortion reference spectrum library, a degradation risk density function is constructed; the risk density function is stratified into confidence intervals to generate a filter operating state level; the filter operating state level is inversely mapped to the filter compensation gain value in the filter disturbance compensation control strategy; the first-order rate of change and the second-order acceleration factor in the degradation evolution trajectory are extracted; and the filter is compressed, labeled, and time-stamped according to the disturbance sensitivity classification rules to output the filter calibration feedback report.

[0013] Specifically, the virtual control mechanism for filtering includes a spectral response layer and a filtering execution layer; The spectral response layer: using spectral shift and perturbation gradient as input variables, calling the spectral shift contribution matrix, performing gradient calculation on the convection diffusion process of the filter distribution within the monochromator, and obtaining the spectral response increment of each grid node; The filter execution layer: Based on the critical interference stability threshold, the compensation upper and lower limits corresponding to each time step and each grid node are used as dynamic boundary conditions for control variables. Based on the dynamic boundary conditions, the compensation change amplitude of adjacent time steps is gradient-limited to generate a filter perturbation self-calibration scheme.

[0014] Specifically, a multi-source perturbation self-calibration system for a rotating filter monochromator includes: Spectral perturbation sensing modeling module: acquires multi-source perturbation signal response data, constructs a filter mapping perturbation response architecture, converts the multi-source perturbation signal response features into perturbation energy spectrum basis vectors, extracts the spectral bandwidth distribution through a spectral network decomposition algorithm, stores the spectral bandwidth distribution in an equivalent distortion reference spectrum library with perturbation labels, and obtains spectral self-perturbation calibration parameters. Filter self-perturbation inversion supplement module: Based on the spectral self-perturbation calibration parameters, perform luminous flux pre-constraint, input the perturbation sensing spectral shape inversion function, detect the convergence state of real-time spectral observations under perturbation change conditions, call the self-calibration operation dependency of multi-source perturbation signal response data, perform confidence calibration on the perturbation contribution weight of the self-calibration operation dependency, and generate a filter perturbation compensation control strategy; Filter evolution verification module: For the filter perturbation compensation control strategy, a filter performance evolution model is constructed, the spectral wavelength shift and resolution data after perturbation self-calibration are identified, and based on the time-series feature alignment algorithm, the similarity matching is performed with the historical degradation samples in the equivalent distortion reference spectrum library to verify the monochromator filter degradation period and output a filter calibration feedback report. Virtual feedback analysis and control module: Based on the filter calibration feedback report, analyze the filter stability margin of the spectral self-disturbance calibration parameters, set the critical interference stability threshold, establish a filter virtual control mechanism, and when the change in the disturbance projection approaches the critical interference stability threshold, enhance the compensation calibration weight of the filter disturbance compensation control strategy to generate a filter disturbance self-calibration scheme.

[0015] The beneficial effects of this invention are as follows: This invention constructs a spectral mapping perturbation response architecture to model the coupling relationship between multi-source perturbation factors and spectral structure changes. This transforms spectral distortion from an independent error into a characterizable and decomposable perturbation energy spectrum feature, thereby improving the accuracy of perturbation identification and the physical consistency of calibration parameters. The invention introduces a pre-constraint of optical flux and a perturbation-aware spectral shape inversion mechanism. Before spectral inversion, the energy input boundary is adaptively modulated to avoid the influence of detector nonlinear response on the spectral shape structure, while preserving high-spectral shape microstructure information, thus improving inversion stability and convergence speed. By constructing a spectral performance evolution model and combining it with a time-series feature alignment algorithm, dynamic identification and periodic verification of filter degradation stages are achieved, enabling the system to predict degradation trends and reduce the risk of sudden inaccuracies. Furthermore, this invention establishes a virtual filter control mechanism and a compensation weight enhancement strategy, which can adjust the compensation gain in advance when the perturbation approaches the critical interference stability threshold, realizing a shift from passive error correction to feedforward self-calibration control. This significantly improves the long-term stability, measurement consistency, and intelligent self-maintenance capabilities of the spectral measurement system, while reducing maintenance frequency and operating costs. Attached Figure Description

[0016] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0017] Figure 1 This is a schematic diagram of the framework of a multi-source perturbation self-calibration method and system for a rotating filter monochromator according to the present invention.

[0018] Figure 2 This is a schematic diagram illustrating the execution of the multi-source perturbation self-calibration method and the virtual control mechanism of the filter in the system of the rotating filter monochromator of the present invention. Detailed Implementation

[0019] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0020] Please see Figure 1 A self-calibration method for multi-source perturbation of a rotating filter monochromator: S1: Acquire multi-source perturbation signal response data, construct a filter mapping perturbation response architecture, convert the multi-source perturbation signal response features into perturbation energy spectrum basis vectors, extract the spectral bandwidth distribution through a spectral network decomposition algorithm, store the spectral bandwidth distribution in an equivalent distortion reference spectral library with perturbation labels, and obtain spectral self-perturbation calibration parameters. S2: Based on the spectral self-perturbation calibration parameters, perform pre-constraint of light flux, input the perturbation sensing spectral shape inversion function, detect the convergence state of real-time spectral observations under perturbation change conditions, call the self-calibration operation dependency of multi-source perturbation signal response data, calibrate the confidence level of the perturbation contribution weight of the self-calibration operation dependency, and generate a filter perturbation compensation control strategy. S3: For the filter perturbation compensation control strategy, construct a filter performance evolution model, identify the spectral wavelength shift and resolution data after perturbation self-calibration, and perform similarity matching with historical degradation samples in the equivalent distortion reference spectrum library based on the time-series feature alignment algorithm to verify the monochromator filter degradation period and output a filter calibration feedback report. S4: Based on the filter calibration feedback report, analyze the filter stability margin of the spectral self-disturbance calibration parameters, set the critical interference stability threshold, establish a filter virtual control mechanism, and when the change in the disturbance projection approaches the critical interference stability threshold, enhance the compensation calibration weight of the filter disturbance compensation control strategy to generate a filter disturbance self-calibration scheme.

[0021] In this embodiment, the method for constructing the filter mapping perturbation response architecture is as follows: based on multi-source perturbation signal response data, which includes environmental perturbation response data and light source operating state perturbation response data, the environment and filter perturbation are uniformly parameterized and encoded to generate a time-calibrated set of perturbation feature vectors. Combined with the spectral response sequence corresponding to the perturbation feature vector set, cross-band correction decomposition is performed to map the spectral response sequence to the spectral state tensor domain. Wavelength drift components and resolution attenuation components caused by different perturbation sources are extracted to construct the filter mapping perturbation response architecture.

[0022] In this embodiment, the process of extracting the spectral bandwidth distribution by the spectral network decomposition algorithm is as follows: multi-dimensional numerical field reconstruction is performed on the local spectral intensity, phase and noise characteristics of the perturbation energy spectrum basis vector to simulate the multi-source perturbation vector field in the rotating filter monochromator, and based on the network decomposition control algorithm, the local harmonic intensity and spectral index in the perturbation vector field are calculated. Using the spectral resolution as the perturbation label, the bandwidth distribution is decomposed and adjusted to obtain the spectral bandwidth distribution of the multi-source perturbation signal response data.

[0023] In this embodiment, the equivalent distortion reference spectral library adopts a hierarchical spectral state coding structure, including a spectral energy distribution layer and a resolution attenuation feature layer. Perturbation source labels and time calibration information are introduced into the features of each layer to generate spectral distortion state vectors. Equivalent mapping and clustering are performed on the spectral responses under different perturbation combinations to form perturbation equivalent subspace clusters. A corresponding set of distortion basis vectors is established for each subspace cluster to store the spectral structure evolution mode under different filter degradation stages and perturbation intensity levels, thereby obtaining spectral self-perturbation calibration parameters.

[0024] In this embodiment, a rotating filter monochromator from an optical laboratory is used in a spectrometer to process the rotating filter monochromator (Monochromator M1, wavelength range 200-800 nm, resolution 0.1 nm) facing multi-source disturbances such as ambient temperature fluctuations (±5°C), light source power drift (±10%), and mechanical vibration (frequency 10-50 Hz). The laboratory is equipped with a sensor array: a spectrometer (Spectrometer S2, for acquiring real-time spectral data), temperature / vibration sensors (SensorT3 / V4), and a control unit (PLC controller). The disturbance signal response data comes from experimental simulations; for example, under a standard deuterium lamp light source, introducing a temperature disturbance causes a wavelength shift of 0.5 nm. The simulation is implemented using a Matlab-based algorithm run on a central server, integrating the SymPy library for symbolic computation, the SciPy library for signal decomposition, and the NetworkX library for spectral network modeling. The equivalent distortion reference spectral library is stored using an SQLite database. Digital twin simulations are performed using COMSOL Multiphysics software to simulate the spectral response under disturbances. Example data: Multi-source perturbation signal (temperature sequence [25,30,28]°C, power [100,95,105]W), spectral observations (peak wavelength shift [500.0,500.3,499.8] nm).

[0025] The specific technical solution is as follows: Acquire multi-source perturbation signal response data, construct perturbation response architecture, extract spectral bandwidth distribution, and obtain self-perturbation calibration parameters; Construction of the filter mapping perturbation response architecture: A unified parameterized encoding is used to generate a set of perturbation feature vectors F = [Tnorm, Vnorm, Pnorm] (normalized to [0,1]). Combined with the spectral response sequence R = [intensity I(λ), phase φ(λ)], cross-band correction decomposition is performed: using SciPy fft to decompose the wavelength drift component Δλ = 0.3 nm and the resolution attenuation component ΔR = 0.05 nm. This is then mapped to the state tensor domain Tensor = [Δλ, ΔR, t] to construct the architecture.

[0026] Spectral network decomposition algorithm: Multidimensional reconstruction of the perturbation energy spectrum basis vector E = [local intensity S=0.8, phase φ=π / 4, noise N=0.02]. Simulate the perturbation vector field and use network decomposition: harmonic intensity H=∑S*cos(φ), spectral exponent α=2.1. Using a resolution of 0.1 nm as the label, the decomposition bandwidth distribution B=[narrowband: 0.2 nm, wideband: 0.5 nm].

[0027] Equivalent distortion reference spectrum library: hierarchical structure: energy distribution layer (vector [I(200nm)=0.9, ...]), resolution attenuation layer ([ΔR=0.05, ...]). Labels are introduced (perturbation source: temperature, time t=0 min), generating the distortion vector Vdist = [0.3,0.05]. A clustering subspace (K-means, k=3) stores the set of basis vectors. Calibration parameters are obtained: {Δλ_max:0.5 nm, ΔR_th:0.1 nm}.

[0028] Based on the calibration parameters, pre-constraints on luminous flux are performed, convergence status is detected, and compensation control strategies are generated. Luminous flux pre-constraint: Projected onto the control space, a suppression weight wsup=0.7 is applied to the high-frequency components (>50 Hz), and a low-frequency drift compensation wcomp=1.2 is applied. After correcting the spectral shape change, a constraint value C=0.85 (energy adjustment amplitude) is obtained.

[0029] Perturbation-sensing spectral inversion function: Input C, output initial vector V_init = [Δλ=0.3, bandwidth compression=0.04]. Mapped to response nodes, associated wavelength range (400-600 nm: weight 0.6). Recursively corrected inversion coefficients k=0.9, unsteady-state parameter p=0.2. Convergence detection: if ||Vinit - Vtarget||<0.01, state convergence is achieved.

[0030] Filter perturbation compensation control strategy: Invoke self-calibration dependency to reconstruct weights for perturbation sources (temperature: 0.4, vibration: 0.3, power: 0.3). Construct offset matrix M = [[0.2, 0.1], [0.15, 0.05]]. Perform orthogonal projection analysis to calculate confidence interval [0.1, 0.4]. Short-time pre-compensation ΔE = 0.05, generating strategy: {compensation gain: 0.25, confidence level: 0.8}.

[0031] Construct a filter performance evolution model, identify offset and resolution data, perform similarity matching, and output a calibration feedback report; Filter performance evolution model: Mapping the observation sequence to the tensor domain Tensor = [Δλgradient=0.01 nm / min, ΔRdecay=0.02]. Time-calibrated degraded samples, calculating the evolution sequence Seq=[t=0: normal, t=10: offset 0.2]. Spectral domain structure: frequency domain branch (bandwidth compression 0.04, offset range ±0.3 nm), time domain branch (attenuation rate 0.015 / min, recovery inertia 0.9).

[0032] Temporal feature alignment algorithm: Deformation amplitude is limited to <0.05, and a penalty factor β=0.1 (perturbation mutation region) is introduced. The matching cost Cmatch=β*Δperturbation is adaptively adjusted. Output matching result: Degradation stage = intermediate (similarity 0.85).

[0033] Filter calibration feedback report: Construct risk density function f(r) = exp(-r / 0.2), confidence interval [0.1, 0.3]. State level = Level 2 (compensation gain 0.25). Extract first-order rate 0.01, second-order acceleration 0.001. Compressed annotation (sensitivity: high), timestamp bound t = 15 min. Output report: JSON {'Level': 'Level 2', 'Verification period': 20 min}.

[0034] Based on the feedback report, stability margin analysis is performed, critical thresholds are set, a virtual control mechanism is established, and a self-calibration scheme is generated. Filter stability margin analysis: Margin Mstab = 1 - (offset / threshold) = 0.7. Set the critical interference threshold Th = 0.4nm (strengthen the weight by 1.5 times when approaching).

[0035] Virtual control mechanism for light filtering: Spectral response layer: Input Δλ=0.3, perturbation gradient 0.02, call offset matrix M, gradient calculation of response increment Δresp=0.05 (grid nodes 10x10).

[0036] Filtering execution layer: dynamic boundary [lower limit 0.1, upper limit 0.5], gradient constraint ||Δcompensation|| < 0.03. Generation scheme: {weight enhancement: 1.5, scheme: pre-compensation 0.2 nm}.

[0037] In this embodiment, before the perturbation-sensing spectral shape inversion function is executed, the spectral self-perturbation calibration parameters are projected onto the spectral flux control space, a suppression weight is applied to the energy adjustment amplitude corresponding to the high-frequency spectral shape component, and a compensation weight is applied to the low-frequency energy drift component to correct the spectral shape information change caused by spectral flux adjustment, thereby obtaining the filter pre-constraint value.

[0038] In this embodiment, the perturbation-sensing spectral shape inversion function takes the filter pre-constraint value as the input variable and outputs an initial spectral shape state vector. The initial spectral shape state vector includes wavelength drift components and bandwidth compression components. Based on the spectral self-perturbation calibration parameters, the initial spectral shape state vector is mapped to local spectral shape response nodes, and the spectral shape changes in different wavelength segments are associated with the corresponding perturbation source contribution weights. The spectral shape inversion coefficients and non-steady-state sensing parameters are recursively corrected, and the convergence state is detected.

[0039] In this embodiment, the method for generating the filter perturbation compensation control strategy is as follows: calling the self-calibration operation dependency of the multi-source perturbation signal response data, performing weighted reconstruction of the response coupling relationship of different perturbation sources in each frequency band of the spectrum, constructing a spectral shift contribution matrix, and combining the spectral domain orthogonal projection algorithm to analyze the spectral shift contribution matrix of the current filter state, calculating the perturbation energy spectrum basis vector of the calibration confidence interval, performing short-time perturbation pre-compensation on the temporal drift of the perturbation energy spectrum basis vector, and generating the filter perturbation compensation control strategy.

[0040] In this embodiment, the method for constructing the filter performance evolution model is as follows: The spectral observation sequence after spectral self-perturbation calibration is mapped to the multidimensional spectral state tensor domain, which includes: wavelength drift gradient dimension and spectral resolution attenuation dimension. Degraded samples in the equivalent distortion reference spectral library are time-calibrated to obtain the spectral filtering evolution trajectory. Based on the spectral filtering evolution trajectory, the perturbation evolution sequence of the current multidimensional spectral state tensor is calculated to construct a filtering performance evolution model containing the spectral domain evolution structure. The spectral domain evolution structure includes a frequency domain branch and a time domain branch; the frequency domain branch is used to control the spectral bandwidth compression and center wavelength shift range, and the time domain branch is used to characterize the resolution decay rate and recovery inertia parameter under disturbance triggering.

[0041] In this embodiment, the temporal feature alignment algorithm limits the local deformation amplitude of the perturbation alignment path based on the self-calibrated spectral wavelength shift and resolution data, and introduces a perturbation sensitivity penalty factor during the calibration path search process to adaptively adjust the matching cost of the perturbation mutation segment, and outputs the corresponding degradation stage matching result.

[0042] In this embodiment, the perturbation sensitivity classification rule is as follows: the perturbation state is classified and managed according to the rate of change of the perturbation energy spectrum amplitude over time and the degree of influence on the stability of the spectral structure. First, the changing trend of the perturbation response is monitored. When the perturbation amplitude changes slowly and its impact on key parameters such as the spectral center wavelength, bandwidth, and resolution is within a compensable range, it is classified as a low sensitivity level. When the perturbation amplitude shows continuous change within a certain time window and has caused a significant shift in the spectral structure parameters but has not yet disrupted the continuity of the spectral microstructure, it is classified as a medium sensitivity level. When the perturbation amplitude fluctuates rapidly or abruptly in a short period, leading to instability in the high-frequency spectral structure, a significant increase in the inversion residual, or an oscillating trend in the convergence process, it is classified as a high sensitivity level. Different sensitivity levels correspond to different compensation adjustment strategies. The low sensitivity level allows for a larger range of compensation gain adjustment, the medium sensitivity level limits the rate of compensation change and enhances stability constraints, and the high sensitivity level prioritizes suppressing rapid compensation actions and activates a protection mechanism to avoid loss of spectral microstructure information or system instability due to over-adjustment.

[0043] The perturbation sensitivity penalty factor is used to adjust the constraint parameter on the intensity of rapid perturbation response during spectral inversion and compensation control. It is determined comprehensively based on the rate of change of the perturbation energy spectrum amplitude in the time dimension, the degree of influence of the perturbation on the stability of the high-frequency structure of the spectrum, and the transient fluctuation of the inversion residual. When the perturbation changes slowly and the spectral structure remains stable, the penalty factor takes a small value to allow the compensation algorithm to fully adjust the low-frequency drift. When the rate of change of the perturbation increases or affects the continuity of the spectral microstructure, the penalty factor increases with the increase of perturbation sensitivity to increase the spectral matching cost or limit the change amplitude of the compensation gain, thereby suppressing the over-response to transient perturbations.

[0044] The spectral domain orthogonal projection algorithm decomposes the weighted constraint subspace of the spectral domain and calculates the weighted constraint minimization problem, with the specific formula as follows: , in, This is a data consistency term, indicating that the current spectral shift should be explainable by the perturbation model. The spectral smoothing term indicates that the energy spectrum of the perturbation should be smoothed as the perturbation source number changes. This is a time continuity term, indicating that the disturbance state changes continuously over time.

[0045] The process of analyzing the filter stability margin is as follows: After obtaining the filter calibration feedback report, key evolution parameters characterizing the filter state, such as the center wavelength shift, spectral bandwidth compression ratio, and resolution attenuation rate after perturbation self-calibration, are first extracted and mapped to the physical constraint space of the filter structure to characterize the stability range of the current filter system in the optical transfer function dimension. Then, combining the filter evolution trajectory of historical degradation samples, the slope, fluctuation amplitude, and trend consistency of these parameters over time are analyzed to identify the critical transition characteristics of the filter performance from a gradually changing stage to an accelerated degradation stage. Based on this, the real-time parameter shift is compared with... The difference between the nominal optical performance thresholds is designed as a characterization of the remaining stability space of the filter. A perturbation projection change is introduced to reduce and correct this stability space, reflecting the compression effect of external perturbations on the filter stability boundary. Furthermore, multiple stability influencing factors are normalized and fused to form a filter stability margin index, used to quantify the safety margin of the current filter system from the interference stability failure boundary. When this index shows a continuous downward trend or approaches a preset critical stability threshold, the filter system is deemed to have insufficient stability margin, providing a basis for subsequent virtual control and compensation weight enhancement. Let the center wavelength shift be one of the key filter evolution parameters after perturbation self-calibration. The spectral bandwidth compression ratio is The resolution attenuation rate is Where B(t) is the actual effective spectral bandwidth at time t, B0 is the nominal spectral bandwidth, R(t) is the spectral resolution at time t, and R0 is the nominal resolution.

[0046] The specific calculation formula is as follows: , Among them, C T C is the temperature drift coupling factor. θ is the coupling factor for the change in incident angle, and k1 and k2 are the structural sensitivity coefficients.

[0047] In this embodiment, the output process of the filter calibration feedback report is as follows: Based on historical degradation samples in the equivalent distortion reference spectrum library, a degradation risk density function is constructed; the risk density function is stratified into confidence intervals to generate a filter operating state level; the filter operating state level is inversely mapped to the filter compensation gain value in the filter disturbance compensation control strategy; the first-order rate of change and the second-order acceleration factor in the degradation evolution trajectory are extracted; and the filter is compressed, labeled, and time-stamped according to the disturbance sensitivity classification rules to output the filter calibration feedback report.

[0048] In this embodiment, as Figure 2 The virtual control mechanism for filtering shown includes a spectral response layer and a filtering execution layer; The spectral response layer: using spectral shift and perturbation gradient as input variables, calling the spectral shift contribution matrix, performing gradient calculation on the convection diffusion process of the filter distribution within the monochromator, and obtaining the spectral response increment of each grid node; The filter execution layer: Based on the critical interference stability threshold, the compensation upper and lower limits corresponding to each time step and each grid node are used as dynamic boundary conditions for control variables. Based on the dynamic boundary conditions, the compensation change amplitude of adjacent time steps is gradient-limited to generate a filter perturbation self-calibration scheme.

[0049] This invention also provides a multi-source perturbation self-calibration system for a rotating filter monochromator, specifically including: Spectral perturbation sensing modeling module: acquires multi-source perturbation signal response data, constructs a filter mapping perturbation response architecture, converts the multi-source perturbation signal response features into perturbation energy spectrum basis vectors, extracts the spectral bandwidth distribution through a spectral network decomposition algorithm, stores the spectral bandwidth distribution in an equivalent distortion reference spectrum library with perturbation labels, and obtains spectral self-perturbation calibration parameters. Filter self-perturbation inversion supplement module: Based on the spectral self-perturbation calibration parameters, perform luminous flux pre-constraint, input the perturbation sensing spectral shape inversion function, detect the convergence state of real-time spectral observations under perturbation change conditions, call the self-calibration operation dependency of multi-source perturbation signal response data, perform confidence calibration on the perturbation contribution weight of the self-calibration operation dependency, and generate a filter perturbation compensation control strategy; Filter evolution verification module: For the filter perturbation compensation control strategy, a filter performance evolution model is constructed, the spectral wavelength shift and resolution data after perturbation self-calibration are identified, and based on the time-series feature alignment algorithm, the similarity matching is performed with the historical degradation samples in the equivalent distortion reference spectrum library to verify the monochromator filter degradation period and output a filter calibration feedback report. Virtual feedback analysis and control module: Based on the filter calibration feedback report, analyze the filter stability margin of the spectral self-disturbance calibration parameters, set the critical interference stability threshold, establish a filter virtual control mechanism, and when the change in the disturbance projection approaches the critical interference stability threshold, enhance the compensation calibration weight of the filter disturbance compensation control strategy to generate a filter disturbance self-calibration scheme.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A multi-source perturbation self-calibration method for a rotating filter monochromator, characterized in that, include: S1: Acquire multi-source perturbation signal response data, construct a filter mapping perturbation response architecture, convert the multi-source perturbation signal response features into perturbation energy spectrum basis vectors, extract the spectral bandwidth distribution through a spectral network decomposition algorithm, store the spectral bandwidth distribution in an equivalent distortion reference spectral library with perturbation labels, and obtain spectral self-perturbation calibration parameters. S2: Based on the spectral self-perturbation calibration parameters, perform pre-constraint of light flux, input the perturbation sensing spectral shape inversion function, detect the convergence state of real-time spectral observations under perturbation change conditions, call the self-calibration operation dependency of multi-source perturbation signal response data, calibrate the confidence level of the perturbation contribution weight of the self-calibration operation dependency, and generate a filter perturbation compensation control strategy. S3: For the filter perturbation compensation control strategy, construct a filter performance evolution model, identify the spectral wavelength shift and resolution data after perturbation self-calibration, and perform similarity matching with historical degradation samples in the equivalent distortion reference spectrum library based on the time-series feature alignment algorithm to verify the monochromator filter degradation period and output a filter calibration feedback report. S4: Based on the filter calibration feedback report, analyze the filter stability margin of the spectral self-disturbance calibration parameters, set the critical interference stability threshold, establish a filter virtual control mechanism, and when the change in the disturbance projection approaches the critical interference stability threshold, enhance the compensation calibration weight of the filter disturbance compensation control strategy to generate a filter disturbance self-calibration scheme.

2. The method according to claim 1, characterized in that, The method for constructing the filter mapping perturbation response architecture is as follows: Based on multi-source perturbation signal response data, which includes environmental perturbation response data and light source operating state perturbation response data, the environment and filter perturbations are uniformly parameterized and encoded to generate a time-calibrated set of perturbation feature vectors. Combined with the spectral response sequence corresponding to the perturbation feature vector set, cross-band correction decomposition is performed to map the spectral response sequence to the spectral state tensor domain. Wavelength drift components and resolution attenuation components caused by different perturbation sources are extracted to construct the filter mapping perturbation response architecture.

3. The method according to claim 1, characterized in that, The process of extracting the spectral bandwidth distribution using the spectral network decomposition algorithm is as follows: multidimensional numerical field reconstruction is performed on the local spectral intensity, phase, and noise characteristics of the perturbation energy spectrum basis vector to simulate the multi-source perturbation vector field in the rotating filter monochromator. Based on the network decomposition control algorithm, the local harmonic intensity and spectral index in the perturbation vector field are calculated. Using the spectral resolution as the perturbation label, the bandwidth distribution is decomposed and adjusted to obtain the spectral bandwidth distribution of the multi-source perturbation signal response data.

4. The method according to claim 1, characterized in that, The equivalent distortion reference spectral library adopts a layered spectral state coding structure, including a spectral energy distribution layer and a resolution attenuation feature layer; Each layer of features introduces perturbation source labels and time calibration information to generate spectral distortion state vectors. Equivalent mapping and clustering are performed on the spectral responses under different perturbation combinations to form perturbation equivalent subspace clusters. A corresponding set of distortion basis vectors is established for each subspace cluster to store the spectral structure evolution mode under different filter degradation stages and perturbation intensity levels, thereby obtaining spectral self-perturbation calibration parameters.

5. The method according to claim 1, characterized in that, Before the perturbation-sensing spectral shape inversion function is executed, the spectral self-perturbation calibration parameters are projected onto the spectral flux control space. Suppression weights are applied to the energy adjustment amplitude corresponding to the high-frequency spectral shape components, and compensation weights are applied to the low-frequency energy drift components to correct the spectral shape information changes caused by spectral flux adjustment, thereby obtaining the filter pre-constraint value.

6. The method according to claim 5, characterized in that, The perturbation-sensing spectral shape inversion function takes the filter pre-constraint value as the input variable and outputs an initial spectral shape state vector. The initial spectral shape state vector includes wavelength drift components and bandwidth compression components. Based on the spectral self-perturbation calibration parameters, the initial spectral shape state vector is mapped to local spectral shape response nodes, and the spectral shape changes in different wavelength bands are associated with the corresponding perturbation source contribution weights. The spectral shape inversion coefficients and non-steady-state sensing parameters are recursively corrected, and the convergence state is detected.

7. The method according to claim 1, characterized in that, The method for generating the filter perturbation compensation control strategy is as follows: calling the self-calibration operation dependency of the multi-source perturbation signal response data, reconstructing the weighted coupling relationship of the response of different perturbation sources in each frequency band of the spectrum, constructing the spectral shift contribution matrix, and combining the spectral domain orthogonal projection algorithm to analyze the spectral shift contribution matrix of the current filter state, calculating the perturbation energy spectrum basis vector in the calibration confidence interval, performing short-time perturbation pre-compensation on the temporal drift of the perturbation energy spectrum basis vector, and generating the filter perturbation compensation control strategy.

8. The method according to claim 1, characterized in that, The method for constructing the filter performance evolution model is as follows: The spectral observation sequence after spectral self-perturbation calibration is mapped to the multidimensional spectral state tensor domain, which includes: wavelength drift gradient dimension and spectral resolution attenuation dimension. Degraded samples in the equivalent distortion reference spectral library are time-calibrated to obtain the spectral filtering evolution trajectory. Based on the spectral filtering evolution trajectory, the perturbation evolution sequence of the current multidimensional spectral state tensor is calculated to construct a filtering performance evolution model containing the spectral domain evolution structure. The spectral domain evolution structure includes a frequency domain branch and a time domain branch; the frequency domain branch is used to control the spectral bandwidth compression and center wavelength shift range, and the time domain branch is used to characterize the resolution decay rate and recovery inertia parameter under disturbance triggering.

9. The method according to claim 8, characterized in that, The temporal feature alignment algorithm limits the local deformation amplitude of the perturbation alignment path based on the self-calibrated spectral wavelength shift and resolution data, and introduces a perturbation sensitivity penalty factor during the calibration path search process to adaptively adjust the matching cost of the perturbation mutation section, and outputs the corresponding degradation stage matching result.

10. The method according to claim 1, characterized in that, The output process of the filter calibration feedback report is as follows: Based on historical degradation samples in the equivalent distortion reference spectrum library, a degradation risk density function is constructed. The risk density function is then stratified into confidence intervals to generate a filter operating state level. The filter operating state level is then inversely mapped to the filter compensation gain value in the filter disturbance compensation control strategy. The first-order rate of change and the second-order acceleration factor in the degradation evolution trajectory are extracted and compressed, labeled, and time-stamped according to the disturbance sensitivity classification rules. Finally, the filter calibration feedback report is output.

11. The method according to claim 1, characterized in that, The virtual control mechanism for filtering includes a spectral response layer and a filtering execution layer; The spectral response layer: using spectral shift and perturbation gradient as input variables, calling the spectral shift contribution matrix, performing gradient calculation on the convection diffusion process of the filter distribution within the monochromator, and obtaining the spectral response increment of each grid node; The filter execution layer: Based on the critical interference stability threshold, the compensation upper and lower limits corresponding to each time step and each grid node are used as dynamic boundary conditions for control variables. Based on the dynamic boundary conditions, the compensation change amplitude of adjacent time steps is gradient-limited to generate a filter perturbation self-calibration scheme.

12. A multi-source perturbation self-calibration system for a rotating filter monochromator, used to perform the method as described in any one of claims 1-11, characterized in that, include: Spectral perturbation sensing modeling module: acquires multi-source perturbation signal response data, constructs a filter mapping perturbation response architecture, converts the multi-source perturbation signal response features into perturbation energy spectrum basis vectors, extracts the spectral bandwidth distribution through a spectral network decomposition algorithm, stores the spectral bandwidth distribution in an equivalent distortion reference spectrum library with perturbation labels, and obtains spectral self-perturbation calibration parameters. Filter self-perturbation inversion supplement module: Based on the spectral self-perturbation calibration parameters, perform luminous flux pre-constraint, input the perturbation sensing spectral shape inversion function, detect the convergence state of real-time spectral observations under perturbation change conditions, call the self-calibration operation dependency of multi-source perturbation signal response data, perform confidence calibration on the perturbation contribution weight of the self-calibration operation dependency, and generate a filter perturbation compensation control strategy; Filter evolution verification module: For the filter perturbation compensation control strategy, a filter performance evolution model is constructed, the spectral wavelength shift and resolution data after perturbation self-calibration are identified, and based on the time-series feature alignment algorithm, the similarity matching is performed with the historical degradation samples in the equivalent distortion reference spectrum library to verify the monochromator filter degradation period and output a filter calibration feedback report. Virtual feedback analysis and control module: Based on the filter calibration feedback report, analyze the filter stability margin of the spectral self-disturbance calibration parameters, set the critical interference stability threshold, establish a filter virtual control mechanism, and when the change in the disturbance projection approaches the critical interference stability threshold, enhance the compensation calibration weight of the filter disturbance compensation control strategy to generate a filter disturbance self-calibration scheme.

Images