A micro-nano grating line deviation rapid detection method and system

Abstract

The present application relates to the technical field of optical detection, and more particularly to a micro-nano grating line deviation rapid detection method and system; the present application adopts coding illumination technology to construct a structured complex light field, breaks through the single-channel detection limit through multi-channel parallel illumination; based on Fourier space multi-frame imaging, composite diffraction spot distribution data is obtained; the coherence and time correlation characteristics that cannot be obtained by traditional intensity measurement are obtained by using the second-order intensity correlation function; a joint forward physical model is constructed, and the grating complex amplitude distribution and line edge roughness parameters are reconstructed by using an iterative inversion algorithm to detect the micro-nano grating line deviation; the present application combines multi-channel coding illumination and second-order intensity correlation function analysis, reduces the measurement time through parallel detection, improves the detection reliability by using the redundant information between channels, and significantly improves the detection efficiency while ensuring the accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optical detection, and in particular to a micro-nano grating line deviation rapid detection method and system. BACKGROUND

[0002] Under the development trend of continuous shrinking of semiconductor manufacturing process technology nodes and continuous improvement of integrated circuit function density, the geometric size precision and topography quality of micro-nano gratings, as the core structural unit of lithography mask, optical element and microelectronic device, directly determine the optical performance, electrical characteristics and manufacturing yield of the device. With the evolution of process nodes to 7nm, 5nm or even more advanced processes, the tolerance requirements of key parameters such as line width, pitch and edge roughness of micro-nano gratings are becoming increasingly stringent. Traditional direct measurement technologies based on scanning electron microscopy, atomic force microscopy and the like face severe challenges in terms of detection speed, sample damage and large-area detection capability. First, the existing detection methods mostly use contact or destructive measurement mode, and the electron beam irradiation and probe contact in the detection process easily cause irreversible damage to the micro-nano structure. Moreover, the single-point measurement mode leads to low detection efficiency, which is difficult to meet the online real-time monitoring demand. Secondly, the traditional optical detection technology is limited by the diffraction limit, and has limited resolution capability for sub-wavelength scale micro-nano gratings. It cannot accurately extract fine topography features such as line edge roughness and sidewall angle, and lacks sensitive analysis of hidden geometric deviations. In addition, the existing detection systems are mostly limited to single-physical-quantity measurement, and cannot realize the collaborative extraction and global quality evaluation of multi-dimensional grating line deviation information.

[0003] For example, the Chinese patent with the authorization announcement number CN109031894B discloses a method for detecting the bottom topography of phase-type defects of an extreme ultraviolet lithography mask multilayer film. The method deposits a multilayer film structure on a mask substrate, uses the phase modulation effect of phase-type defects on reflected light, and combines optical microscope imaging technology to detect the bottom topography features of the defects. The patent mainly solves the problem of insufficient sensitivity of traditional detection methods to buried defects, and realizes effective identification and positioning of phase-type defects inside the multilayer film.

[0004] In addition, the Chinese patent with the authorization announcement number CN102222329B discloses a grating scanning method for depth detection. The technology scans light on an object by grating and detects the light reflected from the object, determines the distance information to the object based on the time-of-flight principle or spatial difference analysis, and generates a three-dimensional mapping of the object. The patent mainly solves the problem of insufficient ranging accuracy of traditional depth cameras in complex scenes, and realizes high-precision three-dimensional reconstruction based on grating scanning.

[0005] The prior art has the following disadvantages in micro-nano grating line deviation detection: limited detection parameter dimension, unable to realize multi-dimensional grating line deviation collaborative analysis; single information extraction means, unable to fully utilize the coherence characteristics and space-time correlation information of the light field; insufficient physical modeling capability, lack of accurate inversion mechanism from diffraction signal to geometric parameters. Therefore, the existing solutions still have significant defects such as limited detection accuracy, insufficient detection speed and lack of multi-parameter collaborative analysis capability in dealing with device performance degradation caused by advanced process micro-nano grating line deviation. To solve the above systematic technical problems, the present application proposes a micro-nano grating line deviation rapid detection method and system. SUMMARY

[0006] The present application adopts coded illumination technology to construct a structured composite light field, breaking through the limitation of traditional single-channel detection; obtains composite diffraction spot distribution data through multi-frame imaging in Fourier space, deeply excavates physical layer information by using second-order intensity correlation function; further constructs a joint forward physical model, reconstructs grating complex amplitude distribution and line edge roughness parameters from diffraction information through iterative inversion algorithm; finally establishes a two-dimensional deviation distribution map, and realizes global quality evaluation combined with abnormal area identification. The present application combines multi-channel coded illumination and second-order intensity correlation function analysis, significantly improves the detection speed and result reliability while ensuring the detection accuracy.

[0007] To achieve the above purpose, the present application provides the following technical solutions:

[0008] A micro-nano grating line deviation rapid detection method, the deviation rapid detection method comprising:

[0009] acquiring composite diffraction spot distribution data generated by the micro-nano grating to be measured in the Fourier plane through multi-frame continuous imaging, calculating the first-order intensity sequence of the composite diffraction spot distribution data and its time fluctuation data;

[0010] According to the first-order intensity sequence and its time fluctuation data, the actual second-order intensity correlation function characteristic quantity of each illumination channel is calculated through channel separation and registration processing;

[0011] inputting the actual second-order intensity correlation function characteristic quantity into a preset joint forward physical model, and obtaining the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters through iterative inversion;

[0012] According to the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters, the grating line deviation at the current coordinate position is calculated.

[0013] Before acquiring the composite diffraction spot distribution data, the method further comprises:

[0014] The structured composite light field is generated by spatially structured modulation of incident light through coded illumination technology, the spatially structured modulation includes phase coding modulation and amplitude coding modulation, wherein:

[0015] The phase coding modulation generates a first illumination channel by applying differential phase delay to different spatial regions of the incident light through loading a preset phase mask pattern, wherein a phase distribution of the first illumination channel is adjusted according to a phase delay distribution of the phase mask pattern;

[0016] The amplitude coding modulation generates a second illumination channel by selectively transmitting different spatial regions of the incident light through loading a preset amplitude mask pattern, wherein an amplitude distribution of the second illumination channel is adjusted by a transmittance distribution of the amplitude mask pattern;

[0017] According to the first illumination channel and the second illumination channel, a plurality of parallel illumination channels with a preset spatial spectrum distribution are generated, wherein the coherence characteristics of the parallel illumination channels are different from each other, and a structured composite light field is generated by spatial superposition according to the parallel illumination channels and the coherence characteristics.

[0018] The complex diffraction spot distribution data generated by the micro-nano grating to be measured in the Fourier plane is acquired through multi-frame continuous imaging, and the first-order intensity sequence of the complex diffraction spot distribution data and the time fluctuation data thereof are calculated, including:

[0019] According to the complex diffraction spot distribution data under the illumination of the structured composite light field, continuous multi-frame real-time acquisition is performed through a preset sampling frequency and total frame number requirement, and continuous multi-frame Fourier plane images of the complex diffraction spot distribution data are acquired;

[0020] According to the spatial distribution functions of the first illumination channel and the second illumination channel, the diffraction spot regions corresponding to each illumination channel are extracted in the multi-frame Fourier plane images through an image segmentation algorithm, and integral light intensity values of each diffraction spot region at each time are calculated through an integral operation method according to the diffraction spot regions;

[0021] According to the integral light intensity values, light intensity response sequences of each illumination channel are constructed in time sequence, and a first-order intensity sequence including at least two illumination channels is formed;

[0022] According to the first-order intensity sequence, intensity variances and time autocorrelation functions of each illumination channel are calculated through a statistical analysis method, and time fluctuation data of each illumination channel is obtained according to the intensity variances and the time autocorrelation functions.

[0023] The actual second-order intensity correlation function characteristic quantity of each illumination channel is calculated according to the first-order intensity sequence and the time fluctuation data thereof, through channel separation and registration processing, including:

[0024] According to the central position coordinates and the spatial distribution function of the first illumination channel and the second illumination channel, the theoretical diffraction positions of the illumination channels in the Fourier plane are determined through a spatial positioning algorithm.

[0025] The channel separation processing is performed on the multi-frame Fourier plane images, the diffraction spot regions corresponding to the illumination channels are identified through an image segmentation algorithm, and the actual central positions of the diffraction spot regions are located through a sub-pixel level registration algorithm.

[0026] According to the actual central positions and the theoretical diffraction positions, the first-order intensity sequence is corrected through a displacement correction algorithm to obtain a corrected first-order intensity sequence.

[0027] According to the corrected first-order intensity sequence, the second-order intensity correlation function between any two illumination channels is calculated through cross-correlation operation, and the peak value, full width at half maximum, decay time constant and coherence degree of the second-order intensity correlation function are extracted as characteristic parameters.

[0028] According to the characteristic parameters, the actual second-order intensity correlation function characteristic quantity is obtained through normalization processing in combination with the coherence characteristics and the relative light intensity weights of the illumination channels.

[0029] The actual second-order intensity correlation function characteristic quantity is input into a preset joint forward physical model, and the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters are obtained through iterative inversion, including:

[0030] According to the central position coordinates, the spatial distribution function, the coherence characteristics and the relative light intensity weights of the first illumination channel and the second illumination channel, a joint forward physical model of micro-nano grating diffraction under coded illumination is constructed based on scalar diffraction theory.

[0031] According to the joint forward physical model, a mathematical mapping relationship from the grating complex amplitude distribution and the line edge roughness power spectrum parameters to the theoretical second-order intensity correlation function characteristic quantity is constructed.

[0032] According to the mathematical mapping relationship and the actual second-order intensity correlation function characteristic quantity, the residual error between the theoretical second-order intensity correlation function characteristic quantity and the actual second-order intensity correlation function characteristic quantity is minimized through a nonlinear least squares iterative inversion algorithm, and the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters are obtained.

[0033] The nonlinear least squares iterative inversion algorithm includes:

[0034] According to the preset initial grating complex amplitude distribution and initial line edge roughness power spectrum parameters, the theoretical second-order intensity correlation function characteristic quantity is calculated through the joint forward physical model;

[0035] The theoretical second-order intensity correlation function characteristic quantity and the actual second-order intensity correlation function characteristic quantity are compared to obtain a residual vector, and a numerical differentiation method is used to calculate a corresponding Jacobian matrix;

[0036] According to the residual vector and the Jacobian matrix, a parameter update amount is calculated through a least squares method, and the updated grating complex amplitude distribution and line edge roughness power spectrum parameters are obtained according to the parameter update amount;

[0037] According to the updated grating complex amplitude distribution and line edge roughness power spectrum parameters, the iteration is repeatedly executed, and when the relative change amount of the two norms of the residual vector is less than a preset threshold, the iteration is terminated and the final grating complex amplitude distribution and line edge roughness power spectrum parameters are output.

[0038] According to the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters, the grating line deviation at the current coordinate position is calculated, including:

[0039] According to the reconstructed grating complex amplitude distribution, the amplitude distribution and phase distribution of the to-be-measured micro-nano grating are extracted, the transmittance distribution and phase delay distribution of the to-be-measured micro-nano grating are calculated, and the geometric contour and boundary position information of the grating lines are extracted through threshold segmentation and edge detection algorithm;

[0040] According to the geometric contour and boundary position information of the grating lines, the position deviation amount of the micro-nano grating lines at the current coordinate position is calculated, and the grating line deviation statistical result is obtained through statistical analysis;

[0041] According to the line edge roughness power spectrum parameters, the line edge roughness characteristic parameters are calculated through function integration;

[0042] According to the grating line deviation statistical result and the line edge roughness characteristic parameters, the grating line deviation at the current coordinate position is obtained through a preset quality standard.

[0043] After the grating line deviation at the current coordinate position is calculated, the method further comprises:

[0044] The plurality of preset scanning positions of the to-be-measured micro-nano grating are scanned, the deviation distribution atlas is constructed according to the grating line deviation and the line edge roughness index of each scanning position, the grating line deviation detection result is obtained through abnormal area identification, and the corresponding adjustment instruction is generated, wherein the deviation distribution atlas is constructed according to the grating line deviation and the line edge roughness index of each scanning position, including:

[0045] According to the preset plurality of scanning positions, a two-dimensional regular grid covering the entire detection area is constructed;

[0046] According to the two-dimensional regular grid and the plurality of scanning positions, the obtained grating line deviation and line edge roughness characteristic parameters are obtained by a spatial interpolation algorithm to obtain an interpolated regular grid;

[0047] According to the interpolated regular grid, a line width deviation distribution map, a position deviation distribution map, a pitch deviation distribution map, and a line edge roughness distribution map are constructed, and a two-dimensional deviation distribution atlas is constructed by a weighted fusion algorithm.

[0048] The grating line deviation detection result obtained by the abnormal area identification generates corresponding adjustment instructions, including:

[0049] According to the two-dimensional deviation distribution atlas, an abnormal threshold is set, an abnormal grid node is located, and a connected abnormal area is obtained by spatial clustering;

[0050] According to the area and shape characteristics of the connected abnormal area, the abnormal type is determined by pattern recognition;

[0051] According to the abnormal type, a re-inspection strategy is formulated, a re-inspection result is obtained, and the re-inspection strategy includes performing local high-density scanning and prolonging exposure time;

[0052] According to the re-inspection result, the true abnormal area is determined by confidence test, and the grating line deviation detection result and the final adjustment instruction are generated.

[0053] A micro-nano grating line deviation rapid detection system is used to realize a micro-nano grating line deviation rapid detection method, and the deviation rapid detection system includes a light source illumination module, a coding modulation module, a scanning control module, an optical imaging module, and a data processing module, wherein:

[0054] The light source illumination module is used to provide a partially coherent light source;

[0055] The coding modulation module includes a phase modulator and an amplitude modulator, which are used to perform phase coding modulation and amplitude coding modulation on incident light to generate a first illumination channel and a second illumination channel;

[0056] The scanning control module includes a scanning driving unit and a position configuration unit, the scanning driving unit is used to control the structured composite light field to perform point scanning on a plurality of preset scanning positions of the micro-nano grating to be measured, and the position configuration unit is used to output a preset position coordinate to provide a position reference for point scanning;

[0057] The optical imaging module comprises a Fourier lens, a channel separation optical element and an imaging detector, the Fourier lens is used for focusing multi-stage diffraction light to a Fourier plane, the channel separation optical element is used for separating diffraction light of each parallel illumination channel, and the imaging detector is used for multi-frame acquisition of composite diffraction spot distribution data of the Fourier plane.

[0058] The data processing module is used for performing fast detection of grating line deviation, outputting a grating line deviation detection result and generating a final adjustment instruction.

[0059] Compared with the prior art, the present application has the following beneficial effects:

[0060] The present application can obtain the diffraction light field coherence characteristic information that cannot be obtained by traditional single-channel detection through multi-channel coded illumination technology and second-order intensity correlation function analysis combined with joint forward physical model inversion. The Fourier space imaging and iterative inversion algorithm are used to realize accurate reconstruction of the grating complex amplitude distribution and line edge roughness parameters. The two-dimensional deviation distribution map and abnormal area identification are used to significantly improve the detection reliability. Compared with the existing method, the present application breaks through the information acquisition limitation of traditional single-channel detection, avoids the time consumption of point-by-point scanning, realizes large-area parallel detection, and greatly improves the detection efficiency while ensuring high precision. BRIEF DESCRIPTION OF DRAWINGS

[0061] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments made with reference to the accompanying drawings:

[0062] Figure 1 FIG. 1 is a flowchart of a micro-nano grating line deviation fast detection method according to an embodiment of the present application;

[0063] Figure 2 FIG. 3 is an execution flowchart of system scanning on multiple preset scanning positions of a micro-nano grating to be measured according to an embodiment of the present application;

[0064] Figure 3 FIG. 5 is a schematic diagram of structured composite light field generation according to an embodiment of the present application;

[0065] Figure 4 FIG. 7 is a whole framework diagram of a micro-nano grating line deviation fast detection system according to an embodiment of the present application. DETAILED DESCRIPTION

[0066] The technical scheme of the present application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific features in the embodiments and the specific embodiments are detailed descriptions of the technical scheme of the present application, but not limitations of the technical scheme of the present application.

[0067] The term "and / or" in the following description is merely an association relationship between associated objects, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone. In addition, the character " / " generally represents an "or" relationship between the associated objects before and after it.

[0068] The embodiment of the present application provides an optical detection method, comprising:

[0069] The complex diffraction spot distribution data generated by the sample to be detected in the Fourier plane is collected through multi-frame continuous imaging, the first-order intensity sequence of the complex diffraction spot distribution data and time fluctuation data thereof are calculated;

[0070] According to the first-order intensity sequence and the time fluctuation data thereof, the actual second-order intensity correlation function characteristic quantity of each illumination channel is calculated through channel separation and registration processing;

[0071] The actual second-order intensity correlation function characteristic quantity is input into a preset joint forward physical model, and the reconstructed sample structure parameters and surface characteristic parameters are obtained through iterative inversion;

[0072] According to the reconstructed sample structure parameters and surface characteristic parameters, the sample feature deviation of the current coordinate position is calculated.

[0073] Exemplarily, the embodiment takes micro-nano grating line width deviation detection as an example to specifically illustrate the above optical detection method.

[0074] In the prior art, the detection of micro-nano grating line deviation usually adopts two ways, way one is that the system adopts a traditional single-channel optical imaging detection, the grating is vertically illuminated and imaged, and then the image processing algorithm is used to extract the grating line edge information to calculate the line width deviation. This method ensures that in the case of relatively regular grating structure, the basic geometric dimension measurement function can be realized while the system simplicity is maintained. However, this method has the problem of insufficient sensitivity to small structural changes, and the detection accuracy of the system is mainly affected by the imaging resolution and the edge extraction algorithm accuracy, and the insufficient sensitivity seriously affects the detection performance of the system under the requirement of nanometer accuracy.

[0075] Way two is to use a contact detection method based on a scanning probe microscope to measure the grating line deviation, which obtains the three-dimensional topographic information of the grating surface through point-by-point scanning of the probe. Although this method has extremely high spatial resolution in theory, in actual application, due to the need for point-by-point mechanical scanning and complex data processing, the detection speed is extremely slow, and in large-area detection, there are still problems of low efficiency and easy disturbance by environmental vibration.

[0076] Therefore, the prior art is often difficult to achieve a perfect balance among improving detection accuracy, reducing detection time and reducing system complexity, especially in micro-nano manufacturing and other application scenarios with extremely high requirements for detection efficiency and accuracy, the existing grating deviation detection method is difficult to meet the dual requirements of fast detection and high-precision measurement.

[0077] In order to realize efficient and accurate detection of grating line deviation in a micro-nano manufacturing line, improve the adaptability of the detection system to complex grating structures, and quickly respond to quality abnormalities to maximize product qualification rate, the embodiments of the present application provide a kind of micro-nano grating line deviation fast detection method based on the above optical detection method, as shown in the figure, the execution steps of the method include: Figure 1

[0078] S1: acquiring the composite diffraction spot distribution data generated by the micro-nano grating to be measured in the Fourier plane by multi-frame continuous imaging, calculating the first-order intensity sequence of the composite diffraction spot distribution data and its time fluctuation data;

[0079] In this step, the multi-stage diffraction light generated by the micro-nano grating to be measured is used as input by structured composite light field irradiation, the composite diffraction spot distribution data is obtained by Fourier space multi-frame continuous imaging technology, and the light intensity response sequence of each illumination channel is extracted by using image segmentation algorithm and integral operation method, and the first-order intensity sequence and time fluctuation data are calculated. The composite diffraction data acquisition method significantly improves the information acquisition ability and detection efficiency of the small structural changes of the grating by fusing multi-channel parallel detection and imaging technology, and provides a more abundant and reliable original data basis for subsequent accurate analysis.

[0080] S2: according to the first-order intensity sequence and its time fluctuation data, the actual second-order intensity correlation function characteristic quantity of each illumination channel is calculated by channel separation and registration processing;

[0081] In this step, the first-order intensity sequence and time fluctuation data are used as input, the actual center position of each diffraction spot is located by channel separation and sub-pixel registration processing technology, and the actual second-order intensity correlation function characteristic quantity of each illumination channel is calculated by using cross-correlation operation and feature parameter extraction method. The second-order intensity correlation function analysis method significantly improves the sensitivity and feature extraction accuracy of the coherent information of the light field by fusing spatial positioning correction and coherent characteristic deep mining technology, and provides more accurate and stable characteristic quantity input for physical model inversion.

[0082] S3: inputting the actual second-order intensity correlation function characteristic quantity into a preset joint forward physical model, and obtaining the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters by iterative inversion;

[0083] ​In this step, the actual second-order intensity correlation function characteristic quantity is taken as input, the mapping relationship between diffraction information and grating structure parameters is established through joint forward physical model, and the grating complex amplitude distribution and line edge roughness power spectrum parameters are further reconstructed by using nonlinear least squares iterative inversion algorithm. The joint physical model inversion method significantly improves the reconstruction accuracy and calculation stability of the grating structure parameters by fusing multi-physical field coupling modeling and intelligent optimization algorithm, and provides more reliable and complete structure information basis for accurate deviation calculation.

[0084] S4: According to the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters, the grating line deviation at the current coordinate position is calculated;

[0085] In this step, the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters are taken as input, the spatial distribution characteristics of grating lines are quantified by geometric profile extraction algorithm, and the grating line deviation at the current coordinate position is further obtained by using multi-dimensional deviation calculation method. The grating line deviation calculation method significantly improves the detection accuracy and parameter characterization ability of nanoscale structure deviation by fusing complex amplitude information analysis and multi-dimensional geometric feature quantization technology, and provides more accurate and comprehensive deviation evaluation results for micro-nano manufacturing quality control.

[0086] In one aspect, according to the reconstructed grating complex amplitude distribution, the amplitude distribution and phase distribution of the measured micro-nano grating are extracted, the transmittance distribution and phase delay distribution of the measured micro-nano grating are calculated, and the geometric profile and boundary position information of the grating lines are extracted by threshold segmentation and edge detection algorithm. Specifically, the amplitude component and phase component are separated from the reconstructed grating complex amplitude distribution, the amplitude distribution is obtained by modulus calculation according to the amplitude component; the continuous phase distribution is obtained by amplitude angle calculation combined with phase unwrapping algorithm according to the phase component; the amplitude distribution is converted into transmittance distribution based on the optical transmission characteristics of the grating, and the phase delay distribution is directly obtained; the transmittance distribution is binarized by threshold segmentation algorithm, and the transmittance image is obtained. The transmittance image is segmented into high-transmittance grating line area and low-transmittance background area by adaptive threshold, and the segmentation noise is eliminated by morphological filtering; the segmented transmittance image is processed by edge detection algorithm, the boundaries of grating lines and background are detected by gradient operator, the single-pixel-width edge line is obtained by non-maximum suppression and double-threshold processing, and the boundary profile of grating lines is formed; the single-pixel-width edge line is connected and arranged by contour tracking algorithm, the complete geometric profile of each grating line is extracted, and the left and right boundary positions of each line are accurately determined on the transmittance profile by sub-pixel positioning technology, and finally the boundary position information including line center profile, boundary coordinates and geometric parameters is obtained.

[0087] In another aspect, according to the geometric profile and boundary position information of the grating lines, the position deviation of the micro-nano grating lines at the current coordinate position is calculated, and the grating line deviation statistical result is obtained through statistical analysis. Specifically, according to the geometric profile and boundary position information of the grating lines, the least square fitting method is used to fit the center axis of each grating line to obtain the center position coordinates of the grating lines; the actual line width value of each grating line is calculated according to the boundary position information, and the actual spacing value is calculated by the distance between the center axes of adjacent grating lines; the center position coordinates, actual line width value and actual spacing value are compared with the preset design nominal value respectively, and the values of line width deviation, position deviation and spacing deviation are calculated, and through statistical analysis, the mean value, standard deviation, maximum value and minimum value of each type of deviation are calculated to obtain the grating line deviation statistical result.

[0088] Further, according to the line edge roughness power spectrum parameters, the line edge roughness characteristic parameters are calculated through function integration. Specifically, according to the reconstructed line edge roughness power spectrum parameters, the power spectrum density function is constructed; the power spectrum density function is integrated in the full frequency domain range by using the numerical integration method, and the root mean square value of the line edge roughness is calculated; the spatial correlation length is extracted from the full width at half maximum feature of the power spectrum by analyzing the frequency characteristics of the power spectrum density function; the spectral index is calculated according to the slope of the power spectrum density function in the logarithmic coordinates; and the root mean square value, spatial correlation length and spectral index are combined to form the line edge roughness characteristic parameters.

[0089] Finally, according to the grating line deviation statistical result and the line edge roughness characteristic parameters, the grating line deviation at the current coordinate position is obtained through the preset quality standard. Specifically, according to the grating line deviation statistical result and the line edge roughness characteristic parameters, the preset quality standard including line width deviation tolerance, position deviation tolerance, spacing deviation tolerance and line edge roughness tolerance is constructed; each deviation statistical value is compared with the corresponding quality tolerance to calculate the quality score of each index; the line width deviation score, position deviation score, spacing deviation score and roughness score are integrated by using the weighted average algorithm to calculate the comprehensive quality score; according to the comparison between the comprehensive quality score and the preset quality level threshold, the grating line deviation level at the current coordinate position is determined, and the grating line deviation at the current coordinate position including specific deviation value, quality score and level determination is generated.

[0090] As Figure 2As shown, in the present embodiment, a plurality of preset scanning positions of the micro-nano grating to be tested are systematically scanned, and the grating line deviation and line edge roughness indicators are obtained according to the procedures of steps S1-S4 for each scanning position to construct a two-dimensional deviation distribution map. Specifically, first, according to the preset scanning position distribution, a two-dimensional regular grid covering the entire detection area is constructed, and the grid node spacing in the two-dimensional regular grid is set according to the grating feature size and the detection accuracy requirement to ensure sufficient spatial resolution. Then, according to the grating line deviation and line edge roughness characteristic parameters obtained at each scanning position, a bicubic spline interpolation algorithm is used for numerical interpolation of the grid nodes to obtain an interpolated regular grid, which contains the continuous distribution information of each deviation parameter in the entire detection area. Then, according to the interpolated regular grid, a line width deviation distribution map, a position deviation distribution map, a pitch deviation distribution map and a line edge roughness distribution map are constructed respectively, and each distribution map uses pseudo-color coding to display the deviation size. Finally, a weighted fusion algorithm is used to synthesize the four distribution maps, and the weight coefficients of the synthesis are set according to the influence degree of different deviation parameters on the grating performance to construct a two-dimensional deviation distribution map.

[0091] Further, according to the two-dimensional deviation distribution map, the grating line deviation detection result is obtained by abnormal area identification, and the corresponding adjustment instruction is generated. Specifically, according to the numerical distribution characteristics in the two-dimensional deviation distribution map, a statistical method is used to set an abnormal threshold, and the grid nodes with deviation values exceeding the abnormal threshold range are marked as abnormal nodes; a density-based spatial clustering algorithm is used for clustering analysis of the abnormal nodes, and adjacent abnormal nodes are merged into connected abnormal areas, and the parameters of the clustering are optimized and set according to the grid resolution and the minimum detection size of the abnormal area; for the identified connected abnormal areas, the geometric feature parameters such as area, aspect ratio and shape complexity are calculated, and pattern recognition is performed through a machine learning classifier, and the abnormal type is classified according to the multi-dimensional features such as the geometric shape characteristics, spatial distribution law, deviation intensity and position relationship in the grating structure of the abnormal area; according to the abnormal type classification result, a corresponding re-inspection strategy is formulated, including adjusting the scanning resolution, optimizing the exposure parameters, using multi-angle scanning and the like; after re-inspection, the re-inspection result is compared and analyzed with the initial inspection result, a statistical test method is used for confidence verification, and the true abnormal area is determined through statistical significance judgment; finally, the grating line deviation detection result containing the abnormal area position coordinates, abnormal type, severity and the like is generated, and the corresponding process adjustment instruction is generated according to the grating line deviation detection result, including at least photolithography parameter optimization and process condition correction.

[0092] The specific steps of S1 are as follows:

[0093] S1.1: spatially structured modulation of the incident light by coded illumination technique, generating a structured composite light field comprising multiple coherent channels, the spatial structured modulation comprising phase coding modulation and amplitude coding modulation;

[0094] In the present embodiment, before collecting the composite diffraction spot distribution data, the incident light is first spatially structured modulated by coded illumination technique, generating a structured composite light field comprising multiple coherent channels. As shown in FIG. 1, the spatial structured modulation comprises two basic modulation modes, phase coding modulation and amplitude coding modulation. Figure 3

[0095] In one aspect, the phase coding modulation applies differential phase delay to different spatial regions of the incident light by loading a preset phase mask pattern, generating a first illumination channel. The phase mask pattern is designed according to the structural features of the micro / nano grating to be detected and the detection requirements, containing spatially distributed phase delay information. The phase mask pattern is implemented by a liquid crystal spatial light modulator, each pixel unit of which applies corresponding phase delay to the incident light according to the requirements of the phase mask pattern. The liquid crystal material changes the molecular orientation under the driving of electric field, thereby changing the effective refractive index of the region and producing phase modulation to the passing light wave. The phase distribution of the first illumination channel is adjusted according to the phase delay distribution of the phase mask pattern, forming a modulated light beam with a specific wavefront distribution. In another embodiment, the phase mask pattern is implemented by a diffractive optical element, which uses a surface relief structure to make a preset depth distribution on a transparent substrate, different depth regions produce different optical path differences to the incident light, realizing the function of phase modulation.

[0096] In another aspect, the amplitude coding modulation selectively transmits different spatial regions of the incident light by loading a preset amplitude mask pattern, generating a second illumination channel. The amplitude mask pattern design contains spatially distributed transmittance information, which is used to control the transmission intensity of the incident light in different regions. The amplitude mask pattern is implemented by a digital micromirror device, which contains a large number of independently controllable micro-mirror units, each micro-mirror can be switched between two angular positions, and the spatial selective reflection is realized by controlling the state of each micro-mirror. When the micro-mirror is in the working angle, the incident light is reflected to the effective light path, and when it is in the non-working angle, the incident light is guided to the light trap. The amplitude distribution of the second illumination channel is adjusted by the transmittance distribution of the amplitude mask pattern, forming a modulated light beam with a specific intensity distribution. In other embodiments, the amplitude coding modulation is implemented by a liquid crystal light valve combined with a polarization device, which uses the electrically controllable birefringence characteristics of liquid crystal to adjust the polarization state of the transmitted light, and then realizes the amplitude modulation by a polarizer.

[0097] ​Further, according to the first and second illumination channels, a plurality of parallel illumination channels with preset spatial spectrum distribution are generated. The parallel illumination channels are realized by an optical beam splitting system, which splits the input light beam into a plurality of independent sub-beams, each of which is subjected to independent modulation processing. Each parallel illumination channel adopts a different combination of phase mask pattern and amplitude mask pattern to form an illumination light beam with differentiated spatial spectrum characteristics. The coherence characteristics of the parallel illumination channels are different from each other, which are realized by optical path control and coherence adjustment. The optical path control adopts an adjustable delay line to accurately adjust the optical path length of each illumination channel and control the phase relationship between the illumination channels. The coherence characteristic adjustment is realized by a partially coherent light source or a coherence modulation device, so that each parallel illumination channel has different coherence characteristics.

[0098] According to the parallel illumination channels and the coherence characteristics, a structured composite light field is generated by spatial superposition. The spatial superposition is realized by a beam combiner or an interferometer configuration, which spatially overlaps each parallel illumination channel in a target irradiation region. The light field of each parallel illumination channel is coherently or incoherently superimposed in the superposition region according to its coherence characteristics, forming a structured composite light field with complex spatial structure. The structured composite light field contains information of multiple coherent channels and has rich spatial spectrum content and adjustable illumination characteristics, providing optimized illumination conditions for multi-dimensional illumination of micro / nano gratings.

[0099] S1.2: According to the composite diffraction spot distribution data under the structured composite light field irradiation, continuous multi-frame real-time acquisition is performed through a preset sampling frequency and total frame number requirement, to obtain continuous multi-frame Fourier plane images of the composite diffraction spot distribution data;

[0100] In this embodiment, the continuous multi-frame real-time acquisition is a process of acquiring time-series diffraction information through imaging technology. Specifically, the sampling frequency is determined by analyzing the time spectrum characteristics of the composite diffraction spot distribution data. The total frame number is determined according to the sample size requirement and the measurement accuracy index of statistical analysis, and the sample size requirement is obtained through theoretical analysis and numerical simulation calculation. In some optional embodiments, the determination of the total frame number also needs to consider the actual constraints such as system stability, storage capacity and processing efficiency.

[0101] Further, according to the sampling frequency and the total number of frames, a Fourier plane image is continuously acquired by an imaging device such as a high-speed CCD camera, a CMOS image sensor, or a photodiode array, at a predetermined time interval. The imaging device receives the diffracted light generated after the structured complex light field illuminates the micro-nano grating under test and converts the diffracted light into digital image data. During the acquisition process, the time synchronization of each frame image acquisition is ensured by precise timing control. A master clock generator provides a stable reference frequency as the time reference, a trigger signal distributor generates acquisition trigger signals for each frame according to the sampling frequency, and a frame synchronization monitor records the time stamp of each frame image in real time.

[0102] Finally, the digital image data is transmitted in real time through a data transmission interface, and the continuous acquisition and storage are realized by using a ring buffer and pipeline processing technology, so as to obtain the continuous multiple Fourier plane images containing time sequence information.

[0103] S1.3: According to the spatial distribution functions of the first and second illumination channels, the diffracted light spot regions corresponding to each illumination channel are extracted from the multiple Fourier plane images by an image segmentation algorithm. According to the diffracted light spot regions, the integral light intensity values of each diffracted light spot region at each time are calculated by an integral operation method.

[0104] In this embodiment, the spatial distribution functions of each illumination channel are established according to the encoding modulation process. For the first illumination channel, the spatial distribution function of the first illumination channel is constructed according to the phase delay distribution of the phase mask pattern, which includes the spatial intensity distribution and phase distribution characteristics of the light field after phase encoding modulation. For the second illumination channel, the spatial distribution function of the second illumination channel is constructed according to the transmittance distribution of the amplitude mask pattern, which includes the spatial intensity distribution and phase distribution characteristics of the light field after amplitude encoding modulation. According to the spatial distribution functions, the diffraction transform results of the spatial distribution functions of each illumination channel in the Fourier plane are calculated by Fraunhofer diffraction theory, the corresponding spatial spectrum distribution of each illumination channel in the Fourier plane is obtained, and the theoretical center position coordinates, spot shape characteristics, size range, and intensity distribution law of the diffracted light spot generated by each illumination channel in the Fourier plane are determined, thereby providing a theoretical prediction benchmark and segmentation region positioning basis for the subsequent image segmentation algorithm.

[0105] Further, according to the spatial distribution function, a theoretical diffraction model of each illumination channel is constructed by diffraction optics theory, and a mapping relationship from illumination channel parameters to Fourier plane diffraction spot characteristics is established. The theoretical diffraction model includes prediction of position distribution of diffraction spots of each illumination channel in the Fourier plane, calculation of spot boundary profile, simulation of light intensity distribution, and analysis of interference effect between adjacent spots. For each frame of the plurality of Fourier plane images, a region-based image segmentation algorithm is used to extract the diffraction spot region corresponding to each illumination channel. Specifically, according to the theoretical position coordinates determined by the spatial distribution function, a search window for each diffraction spot is established in the Fourier plane image to determine the approximate positioning area of each diffraction spot. In each search window, an adaptive threshold segmentation method is used to automatically determine the segmentation threshold by analyzing the local gray scale distribution characteristics, and the diffraction spot region is effectively separated from the background noise. At the same time, a region growing algorithm based on seed points is used, and the theoretical center position of each diffraction spot is taken as the seed point. According to the pixel gray scale similarity criterion and spatial connectivity constraint, the region is gradually expanded and grown to accurately determine the actual boundary profile of each diffraction spot region.

[0106] Further, according to the diffraction spot region, an integral light intensity value of each diffraction spot region at each time is calculated by an integral operation method. Specifically, for each identified diffraction spot region, the gray scale values of all pixels in the region are subjected to spatial integral operation to obtain the total light intensity value of the diffraction spot region. According to the pixel response characteristics and quantum efficiency of the imaging device, the total light intensity value of the diffraction spot region is calibrated and corrected to obtain a corrected integral light intensity value. The integral operation is repeatedly performed on each diffraction spot region, and the integral light intensity values of each diffraction spot region at each time are recorded in time sequence to obtain the integral light intensity values of each diffraction spot region at each time.

[0107] S1.4: According to the integral light intensity value, a light intensity response sequence of each illumination channel is constructed in time sequence to form a first-order intensity sequence including at least two illumination channels;

[0108] In this embodiment, according to the sampling time stamp, the integral light intensity values of each illumination channel at different times are arranged in time sequence to establish a corresponding relationship between time index and light intensity data. The light intensity data corresponding to the same physical time point under the same time index of each illumination channel ensures the time synchronization between the multi-channel data.

[0109] For each illumination channel, a complete light intensity response sequence is constructed according to the order of sampling time. The light intensity response sequence is a one-dimensional array including all integral light intensity values of the illumination channel within the entire measurement time period, and the elements of the array are arranged in time sequence. For a system including multiple illumination channels, the light intensity response sequences of each illumination channel are constructed respectively, and each sequence maintains the same time length and sampling interval.

[0110] Further, the light intensity response sequence of each illumination channel is organized into a matrix-form multi-channel time series data structure, where each row corresponds to the light intensity response sequence of one illumination channel, and each column corresponds to the integrated light intensity value of all illumination channels at a certain time instant. The number of rows of the multi-channel time series data structure is equal to the total number of illumination channels, and the number of columns is equal to the total number of sampling time instants.

[0111] Finally, the light intensity response sequence of each illumination channel is defined as a first-order intensity sequence, which directly reflects the variation law of the integrated light intensity value over time. In the above manner, the first-order intensity sequence of at least two illumination channels is constructed, providing a time series data basis for subsequent light intensity deviation analysis.

[0112] S1.5: According to the first-order intensity sequence, the intensity variance and the time autocorrelation function of each illumination channel are calculated by statistical analysis method, and the time fluctuation data of each illumination channel are obtained according to the intensity variance and the time autocorrelation function.

[0113] In this embodiment, first, according to the first-order intensity sequence, the intensity variance is calculated by variance statistics, which is used to obtain the intensity characteristic information of time fluctuation. Specifically, according to the first-order intensity sequence of each illumination channel, the average value in the time dimension is calculated as the light intensity reference level of each illumination channel; the deviation of each light intensity sampling value in the first-order intensity sequence from the light intensity reference level is compared to obtain the deviation value, and the square value of the deviation is obtained by squaring; the intensity variance of each illumination channel is calculated by averaging the square value of the deviation.

[0114] Subsequently, the time autocorrelation function is calculated according to the first-order intensity sequence, which is used to reveal the time characteristic information of the light intensity signal. Specifically, first, the mean value of the first-order intensity sequence is removed. Then a series of time interval parameters are set, starting from zero time interval, with the sampling time interval of the first-order intensity sequence as the step, gradually increasing to a certain proportion of the total length of the sequence as the maximum analysis time interval. For each set time interval, the correlation coefficient between the first-order intensity sequence after mean value removal and the sequence translated by the time interval is calculated. By systematically traversing all set time intervals, a complete time autocorrelation function curve is constructed, and the calculation result of the time autocorrelation function reflects the correlation degree of the first-order intensity sequence of the same illumination channel under different time intervals, and the decay characteristic reveals the time correlation scale of the light intensity fluctuation.

[0115] Further, the time fluctuation data of each illumination channel is extracted by combining the intensity variance and the time autocorrelation function. Specifically, the intensity feature information of the time fluctuation is obtained from the intensity variance, and the time feature information of the time fluctuation is extracted from the time autocorrelation function. The consistency of the calculation is verified by analyzing the relationship between the value of the time autocorrelation function at zero time interval and the intensity variance. The intensity feature information and the time feature information are combined to form complete time fluctuation data of each illumination channel, providing the necessary fluctuation characteristic information basis for subsequent light intensity deviation analysis and micro-nano grating parameter measurement.

[0116] The specific steps of S2 are as follows:

[0117] S2.1: According to the center position coordinates and the spatial distribution function of the first illumination channel and the second illumination channel, the theoretical diffraction position of each illumination channel in the Fourier plane is determined by a spatial positioning algorithm.

[0118] In this embodiment, the spatial geometric characteristics of the illumination channels are extracted according to the center position coordinates and the spatial distribution function of the first illumination channel and the second illumination channel, and the theoretical diffraction position of each illumination channel in the Fourier plane is determined by theoretical calculation, providing a theoretical benchmark for subsequent diffraction spot positioning and displacement correction.

[0119] Specifically, first, the center position coordinates of each illumination channel are extracted according to the preset phase mask pattern and amplitude mask pattern. For the first illumination channel, the phase delay distribution data is extracted from the preset phase mask pattern, the geometric center of the effective area with phase modulation effect is determined as the center position coordinates of the first illumination channel by using the geometric barycenter calculation method. For the second illumination channel, the transmittance distribution data is extracted from the preset amplitude mask pattern, the effective transmission area is identified by setting a transmittance threshold, and the centroid position of the transmission area is determined as the center position coordinates of the second illumination channel by using the centroid calculation method.

[0120] Further, according to the equipment specification and technical specification document of the optical detection system, a geometric parameter model of the optical system is constructed. Key system parameters are obtained from the technical specification document provided by the optical system manufacturer, including the nominal wavelength of the incident light source, the focal length parameter of the objective lens, the numerical aperture specification and the working distance range. By referring to the equipment specification, the design distance of the objective lens to the Fourier plane, the relative position relationship of the optical elements and the geometric configuration of the optical path are obtained. According to the system parameters and the geometric configuration, a geometric parameter model of the optical system is constructed, and the coordinate transformation relationship and the optical propagation characteristics from the object plane to the Fourier plane are defined.

[0121] Finally, according to the center position coordinates of each illumination channel and the spatial distribution function, combined with the geometric parameter model of the optical system, the theoretical diffraction position of each illumination channel in the Fourier plane is calculated. In one aspect, for the first illumination channel, the center position coordinates and the corresponding spatial distribution function of the channel are taken as input, and the focal length parameter and the numerical aperture specification in the geometric parameter model of the optical system are used to establish the correspondence between the object plane coordinates and the Fourier plane coordinates by Fourier optical theory. By using the diffraction integral calculation method, combined with the wavelength parameter of the incident light source, the light field distribution of the illumination channel after diffraction through the micro-nano grating to be tested in the Fourier plane is calculated. According to the design distance parameter from the objective lens to the Fourier plane, the spatial frequency distribution obtained by calculation is converted into the physical coordinate position of the Fourier plane. By analyzing the complex amplitude distribution characteristics of the light field distribution, the intensity barycenter positioning algorithm is used to extract the intensity barycenter position of the diffraction light field as the theoretical diffraction position of the first illumination channel.

[0122] In another aspect, for the second illumination channel, the center position coordinates and the corresponding spatial distribution function of the channel are taken as input, and the focal length parameter, the numerical aperture specification and the design distance from the objective lens to the Fourier plane in the geometric parameter model of the optical system are also used for coordinate transformation and physical position mapping. By using the same diffraction integral calculation method, combined with the nominal wavelength of the incident light source and the relative position relationship parameter of the optical elements, the optical path propagation path is corrected, and the theoretical diffraction position of the second illumination channel in the Fourier plane is determined.

[0123] The above calculation process uses the geometric parameter model of the optical system to accurately convert from the theoretical calculation result to the actual physical coordinates, obtains the theoretical diffraction position coordinates of each illumination channel in the Fourier plane, and forms a theoretical diffraction position data set, which provides an accurate theoretical reference benchmark for the actual diffraction spot positioning and displacement correction algorithm in the subsequent steps.

[0124] S2.2: Channel separation processing is performed on the multiple frames of Fourier plane images, each illumination channel corresponding diffraction spot area is identified by an image segmentation algorithm, and the actual center position of each diffraction spot area is located by using a sub-pixel level registration algorithm.

[0125] In this embodiment, the multi-algorithm fusion sub-pixel level registration algorithm is used to realize high-precision center position recognition of complex diffraction spots, and the precision limitation of traditional pixel-level positioning is broken through. The multi-algorithm fusion includes the cooperative application of the centroid calculation method, the Gaussian fitting method and the ellipse fitting method.

[0126] Firstly, the multi-frame Fourier plane images are pre-processed, including adaptive noise filtering, dynamic background correction and gray scale normalization. The adaptive noise filtering dynamically selects filter type and parameters according to local signal-to-noise ratio characteristics of the image, removes noise while maintaining the clarity of the diffraction spot edge. The dynamic background correction uses a time domain filtering combined with a spatial polynomial fitting method to eliminate background inhomogeneity caused by environmental interference and device drift.

[0127] Then, the multi-frame Fourier plane images are intelligently processed by channel separation. According to the theoretical diffraction positions of the first and second illumination channels, an adaptive region of interest is defined in each image. A hierarchical threshold segmentation method is used to achieve accurate separation of diffraction spots of different intensities through global threshold coarse segmentation and local adaptive threshold fine segmentation.

[0128] Further, a multi-feature fusion image segmentation algorithm is used to identify the diffraction spot regions corresponding to each illumination channel. An improved connected component analysis algorithm is used to identify continuous high-intensity pixel regions as candidate regions. By calculating multi-dimensional geometric feature parameters of each connected component, including area, length-to-short-axis ratio, direction angle, compactness, etc., a multi-feature weighted matching criterion is established to filter out the real diffraction spot regions. For multiple illumination channels, parallel identification processes are performed to achieve efficient and independent separation of diffraction spots for each illumination channel.

[0129] Finally, a multi-algorithm fusion sub-pixel registration algorithm is used to locate the actual center position of each diffraction spot region. First, a weighted centroid calculation method is used to determine the rough center position. In the adaptive neighborhood, multiple mathematical fitting models are constructed in parallel, including a two-dimensional Gaussian function model and an elliptical Gaussian model. By adjusting the parameters of each model through a multi-objective optimization algorithm, a hybrid optimization strategy of least squares method and gradient descent method is used to minimize the fitting error. Finally, through a weighted fusion algorithm, the weighted average center position coordinates are calculated according to the accuracy and reliability of each fitting model, and the sub-pixel precision actual center position of each diffraction spot region is obtained.

[0130] S2.3: According to the actual center position and the theoretical diffraction position, the first-order intensity sequence is corrected by a displacement correction algorithm to obtain a corrected first-order intensity sequence;

[0131] In this embodiment, according to the actual center position and the theoretical diffraction position, a multi-level displacement correction algorithm is used to systematically correct the first-order intensity sequence, eliminating measurement errors caused by spot position offset and achieving sub-pixel level high-precision correction.

[0132] Firstly, the two-dimensional spatial displacement vector of each illumination channel diffraction spot is calculated. Specifically, for each illumination channel, the actual center position is compared with the corresponding theoretical diffraction position, the horizontal displacement component and the vertical displacement component are calculated to form a two-dimensional spatial displacement vector. Then, the light intensity integration region is dynamically corrected according to the two-dimensional spatial displacement vector. Based on the preset light intensity integration region, for each frame of Fourier plane image in the time sequence, the center coordinates of the original light intensity integration region are translated according to the corresponding two-dimensional spatial displacement vector to obtain the corrected light intensity integration region of each illumination channel at each time.

[0133] Further, the light intensity integration value at each time is recalculated according to the corrected light intensity integration region. For each illumination channel at each time, the pixel gray value is accumulated in the corresponding corrected light intensity integration region to obtain the corrected light intensity integration value sequence of each illumination channel. This correction process eliminates the systematic error caused by the position offset of the diffraction spot on the light intensity measurement, and ensures that the light intensity integration value accurately reflects the actual intensity change of the diffraction spot.

[0134] Finally, the corrected first-order intensity sequence of each illumination channel is reconstructed in time sequence. The corrected light intensity integration value sequence of each illumination channel is arranged in time sequence to finally obtain the corrected first-order intensity sequence, and the time sequence eliminates the influence of the spot displacement error, providing accurate and reliable input data for the subsequent second-order intensity correlation function calculation.

[0135] S2.4: According to the corrected first-order intensity sequence, the second-order intensity correlation function between any two illumination channels is calculated by cross-correlation operation, and the peak value, full width at half maximum, decay time constant and coherence degree of the second-order intensity correlation function are extracted as characteristic parameters.

[0136] In this embodiment, the second-order intensity correlation characteristics between multiple channels are calculated by a systematic cross-correlation analysis method according to the corrected first-order intensity sequence, and key characteristic parameters are extracted for subsequent physical parameter inversion and system performance evaluation.

[0137] Firstly, the corrected first-order intensity sequence is standardized and pretreated. Specifically, for each illumination channel, the average value of the corrected first-order intensity sequence in the time dimension is calculated, and the average value is subtracted to obtain an intensity fluctuation sequence, so as to eliminate the common mode noise caused by factors such as light source power fluctuation, system gain change, environmental light interference and the like. Further, the intensity fluctuation sequence is normalized, and the standard deviation of the intensity fluctuation of each illumination channel is taken as the normalization factor to obtain a standardized intensity fluctuation sequence. The normalization process eliminates the amplitude difference and gain inconsistency between different channels.

[0138] Then, the second-order intensity correlation function between any two illumination channels is calculated according to the normalized intensity fluctuation sequence by using the cross-correlation operation method. According to the quantum optics theory, the normalized second-order intensity correlation function between any two illumination channels is defined, which contains the time delay parameter to represent the correlation characteristics at different times. The cross-correlation coefficient is calculated step by step in a preset time delay range by using the sliding time window method, and the cross-correlation operation is accelerated by using the fast Fourier transform algorithm, so as to obtain the second-order intensity correlation function containing the time delay information.

[0139] In another aspect, the decay time constant is extracted according to the decay characteristics of the second-order intensity correlation function by using an exponential fitting algorithm. A suitable fitting window is selected near the peak of the correlation function, and an exponential function fitting is performed by using a nonlinear least square method, so as to obtain the decay time constant, which reflects the time decay rate of the correlation between channels.

[0140] In another aspect, the normalized correlation value under the condition of zero time delay is calculated as the degree of coherence. The degree of coherence is calculated by the correlation between the intensities of different illumination channels at the same time, and reflects the instantaneous coherence degree between channels and the synchronization of the light field distribution.

[0141] Finally, the above calculation process is systematically performed on all possible combinations of illumination channel pairs, and the peak value, full width at half maximum, decay time constant and degree of coherence are combined to construct a complete feature matrix containing four types of feature parameters as feature parameters.

[0142] S2.5: According to the feature parameters, the coherence characteristics and the relative light intensity weight of the illumination channels are combined, and the actual second-order intensity correlation function feature quantity is obtained by normalization processing.

[0143] In this embodiment, first, a quantitative evaluation system of the coherence characteristics of the illumination channels is established. Specifically, for each illumination channel, the coherence length, coherence time and spatial coherence degree parameters are determined according to the light source characteristics, transmission path and modulation mode. By analyzing the spectral characteristics and time stability of each illumination channel, the time coherence function and the spatial coherence function are calculated, and the coherence characteristic matrix between the illumination channels is obtained by combining the coherence properties of the light source and the geometric configuration of the system, which describes the strength of the coherent coupling between different channel pairs.

[0144] Further, the relative light intensity weight coefficients of the illumination channels are determined. Specifically, a characteristic parameter correction model is constructed according to the coherence characteristic matrix and the relative light intensity weight coefficients. Through the characteristic parameter correction model, the peak characteristic parameter is weighted and corrected according to the relative coherence intensity between the illumination channels, so as to eliminate systematic deviation caused by coherence difference; the full width at half maximum is normalized in combination with the coherence time characteristics of the illumination channels, so as to reflect the actual coherence time scale of the system; the decay time constant is weighted and averaged according to the relative light intensity weight coefficients; and the coherence degree is transformed through the coherence characteristic matrix to obtain a corrected characteristic parameter set.

[0145] Further, a multi-level normalization processing strategy is used to standardize the corrected characteristic parameter set. In the first level of normalization, each type of corrected characteristic parameter is standardized to eliminate dimensional influence and numerical range difference. In the second level of normalization, main characteristic components are extracted through principal component analysis to eliminate the influence of correlation between characteristic parameters. In the third level of normalization, environmental adaptability correction is performed in combination with actual working conditions of the system to obtain a standardized characteristic parameter set.

[0146] Finally, the actual second-order intensity correlation function characteristic quantity is constructed according to the standardized characteristic parameter set. Specifically, a weight distribution method driven by a physical model is used to determine weight coefficients according to the contribution degree of each characteristic parameter to system performance, to weightedly combine the standardized characteristic parameter set, and to obtain the actual second-order intensity correlation function characteristic quantity.

[0147] The specific steps of S3 are as follows:

[0148] S3.1: According to the center position coordinates, spatial distribution functions, coherence characteristics and relative light intensity weights of the first illumination channel and the second illumination channel, a joint forward physical model of micro-nano grating diffraction under coded illumination is constructed based on scalar diffraction theory.

[0149] In this embodiment, first, a spatial field distribution model of multi-channel coded illumination is established. Specifically, according to the center position coordinates of the illumination channels, a spatial coordinate system is established on the grating plane to determine the illumination area of each illumination channel. For each illumination channel, a total illumination field distribution function of multi-channel superposition is constructed by using a Gaussian function according to the spatial intensity distribution and phase distribution characteristics in the spatial distribution function.

[0150] Then, a coherence matrix of the illumination field is established based on the coherence characteristics. Specifically, according to the coherence characteristics of each illumination channel, an inter-coherence function between channels is constructed by using the theory of partially coherent light. According to the inter-coherence function, a complex coherence function between each pair of illumination channels is obtained by normalization processing; and a complex coherence function matrix is formed by collecting the complex coherence functions between each pair of illumination channels.

[0151] Next, a diffraction calculation model of the micro-nano grating is constructed based on the scalar diffraction theory. Specifically, according to the geometric topographic information and material optical characteristic parameters of the grating, a complex amplitude transmission function is constructed by using the optical propagation theory, which comprehensively reflects the geometric structural characteristics and material optical response characteristics of the grating. For a periodic grating structure, the transmission function is expressed by a Fourier series based on the periodic characteristics of the grating, and the Fourier coefficients corresponding to each diffraction order are obtained to describe the transmission characteristics of the grating. For a non-periodic or defective grating, the transmission function is directly solved by using a numerical method such as finite difference or finite element method according to the irregular characteristics of the structure, and a spatially discrete transmission function distribution is obtained. In combination with the line edge roughness characteristics in the grating manufacturing process, a random phase modulation term is introduced into the transmission function based on the statistical distribution law of the roughness, and a phase disturbance conforming to the actual roughness characteristics is generated by a statistical modeling method, so as to finally obtain a complete transmission function model considering the influence of manufacturing defects.

[0152] Finally, a joint forward physical model under coded illumination is constructed. Specifically, according to the relative light intensity weights of each illumination channel, the multi-channel illumination field is coupled and calculated with the complex amplitude transmission function by using the linear superposition principle. Based on the scalar diffraction theory, the diffraction light field distribution generated by each illumination channel under the action of the micro-nano grating is calculated in combination with the complex coherence function matrix. By using a mixed processing method of coherent superposition and incoherent superposition, the diffraction contributions of each illumination channel are weighted and synthesized according to the coherence characteristics between channels, so as to form the total diffraction light field distribution under coded illumination. According to the total diffraction light field distribution, a complete forward mapping relationship from the illumination parameters and the geometric topographic information of the grating to the diffraction light field distribution is established, so as to construct a joint forward physical model of the micro-nano grating diffraction under coded illumination.

[0153] S3.2: According to the joint forward physical model, a mathematical mapping relationship from the grating complex amplitude distribution and the line edge roughness power spectrum parameters to the characteristic quantities of the theoretical second-order intensity correlation function is constructed.

[0154] In the present embodiment, firstly, a statistical property description of the diffracted light field is established according to the joint forward physical model. Specifically, the total diffracted light field distribution under the coded illumination condition is expressed in the form of superposition of deterministic component and random component, wherein the deterministic component is determined by the grating complex amplitude distribution, and the random component is controlled by the line edge roughness power spectrum parameter; through statistical optics theory, a complex amplitude statistical distribution model of the diffracted light field at different spatial positions is established, and first-order and second-order statistical moment characteristics of the light field are obtained.

[0155] Then, a theoretical calculation model of the second-order intensity correlation function is constructed. Specifically, according to the definition of the second-order intensity correlation function, a statistical average expression of the intensity product of the diffracted light field at two spatial positions is established; based on the statistical distribution characteristics of the light field complex amplitude, the second-order intensity correlation function is expanded into a fourth-order moment function of the light field complex amplitude through probability theory and random process theory; combined with the Gaussian random process characteristics of the line edge roughness, the fourth-order moment is decomposed into the product combination of the second-order moment by using the Isserlis theorem, and an analytical relationship between the second-order intensity correlation function and the second-order statistical moment of the light field is established.

[0156] Next, characteristic quantity parameters of the second-order intensity correlation function are extracted. Specifically, spatial frequency domain analysis is performed on the theoretical second-order intensity correlation function, and its power spectrum distribution characteristics are obtained through Fourier transform; key characteristic parameters such as peak position, full width at half maximum, and decay constant of the correlation function are extracted as characteristic quantities for describing the grating structure and roughness characteristics; a functional relationship between the characteristic quantity parameters and the periodicity, modulation depth and other structural parameters in the grating complex amplitude distribution, and the correlation length, root mean square roughness and other statistical parameters in the line edge roughness power spectrum parameter is established.

[0157] Finally, a complete mathematical mapping relationship model is constructed. Specifically, based on the foregoing theoretical analysis, an explicit mathematical expression is established from the grating complex amplitude distribution and the line edge roughness power spectrum parameter as input variables to the characteristic quantity of the theoretical second-order intensity correlation function as output variable; through parameter sensitivity analysis, the influence weight and coupling relationship of each input parameter on the characteristic quantity are determined; a parameterized mapping function model is established, realizing quantitative prediction from the grating physical structure parameters to the observable optical characteristic quantity, and forming a complete mathematical mapping relationship from the grating complex amplitude distribution, the line edge roughness power spectrum parameter to the characteristic quantity of the theoretical second-order intensity correlation function.

[0158] S3.3: According to the mathematical mapping relationship and the actual second-order intensity correlation function characteristic quantity, through a nonlinear least squares iterative inversion algorithm, the residual between the theoretical second-order intensity correlation function characteristic quantity and the actual second-order intensity correlation function characteristic quantity is minimized, and the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameter are obtained.

[0159] In the embodiment, by establishing a complete nonlinear least squares iterative inversion algorithm framework, based on the preset initial parameters, the second-order intensity correlation function characteristic quantity is calculated by using the joint forward physical model calculation theory, which is compared with the actual measured characteristic quantity to construct a standardized residual error vector, the numerical differentiation method is used to calculate the Jacobian matrix, the parameter update amount is solved by the least squares method and the grating complex amplitude distribution and the line edge roughness power spectrum parameters are iteratively updated under the physical constraint condition, and when the relative change amount of the two norm of the residual error vector reaches the convergence threshold, the final reconstruction parameters are output. The technical scheme accurately describes the composite physical effects of grating diffraction and line edge roughness by accurate modeling of the joint forward physical model, improves the robustness and accuracy of parameter estimation by using characteristic quantity matching and weighted residual error analysis, ensures the numerical stability and convergence of the iteration process by combining numerical differentiation Jacobian matrix calculation and adaptive parameter update strategy, and guarantees the physical rationality and measurement consistency of the reconstruction result by the physical constraint verification and quality evaluation mechanism, realizes the high-precision inversion reconstruction from the second-order intensity correlation function measurement data to the grating microstructure parameters, and provides a reliable technical means for grating manufacturing process optimization and performance evaluation.

[0160] The specific steps of S3.3 are as follows:

[0161] S3.3.1: According to the preset initial grating complex amplitude distribution and initial line edge roughness power spectrum parameters, the theoretical second-order intensity correlation function characteristic quantity is calculated by the joint forward physical model;

[0162] In the embodiment, first, according to the grating design parameters and manufacturing process information, the initial grating complex amplitude distribution is set, wherein the amplitude distribution uses an ideal rectangular wave function as an initial approximation, the duty cycle is determined according to the design value, and the phase distribution is calculated according to the initial estimation value of the phase delay according to the refractive index and thickness information of the grating material, to obtain the initial grating complex amplitude distribution containing amplitude and phase information; according to the typical characteristics of the manufacturing process, the initial values of the correlation length, roughness amplitude and spectral index are set to obtain the initial line edge roughness power spectrum parameters.

[0163] Then, according to the minimum feature size of the grating and the required accuracy requirement, the initial grating complex amplitude distribution is discretized into a numerical array on a spatial grid, the amplitude and phase information of each grid point is stored in a complex form, and a numerical grating complex amplitude distribution matrix is obtained; according to the physical meaning of the line edge roughness power spectrum parameters, a vector representation including key parameters such as correlation length, roughness amplitude, and spectral index is established, and normalized line edge roughness power spectrum parameters are obtained. According to the numerical grating complex amplitude distribution matrix and the normalized line edge roughness power spectrum parameters, a grating transmission function matrix is constructed. According to the grating transmission function matrix, a random phase distribution conforming to the power spectrum characteristics is generated by using the Monte Carlo method, a random phase modulation term is introduced into the transmission function, and a grating transmission function including the line edge roughness effect is obtained.

[0164] Further, according to the spatial distribution and coherence characteristic parameters of the multi-channel coded illumination, the incident field distribution of each illumination channel on the grating plane is calculated, and a multi-channel incident field distribution matrix is obtained according to the amplitude, phase, and polarization characteristics of each illumination channel in the incident field distribution.

[0165] Still further, according to the incident field distribution and the grating transmission function, a transmitted field distribution is obtained; according to the transmitted field distribution, the distribution of the diffraction field on the observation plane is calculated by using the angular spectrum propagation method or the Fresnel diffraction integral, and the total diffraction field distribution is obtained through the coherence superposition effect among the multi-channels in combination with the coherence matrix. According to the spatial sampling distribution of the Fourier plane, the complex amplitude signals of each spatial position are extracted from the total diffraction field distribution; according to the response characteristics and noise characteristics of the imaging process, the complex amplitude is converted into an intensity signal, and the intensity time sequence data of each spatial position is obtained by combining the time modulation sequence of the coded illumination and simulating the time variation process.

[0166] Finally, according to the intensity time sequence data, the second-order intensity correlation functions among the detection channels are calculated by using the cross-correlation algorithm, the correction model is constructed to correct the second-order intensity correlation functions through the finite sampling effect and statistical fluctuation characteristics, and the theoretical second-order intensity correlation functions with reduced statistical error are obtained through multiple independent calculations in combination with the multi-realization average method.

[0167] According to the theoretical second-order intensity correlation function data, the peak value, full width at half maximum, decay time constant, and coherence degree are extracted by using a feature extraction algorithm; the reliability of the peak value, full width at half maximum, decay time constant, and coherence degree is monitored and calculated according to the quality evaluation index. The extracted peak value, full width at half maximum, decay time constant, and coherence degree are organized into a feature vector form, and the theoretical second-order intensity correlation function feature quantity is obtained, which is used for subsequent comparison and optimization calculation.

[0168] S3.3.2: comparing the theoretical second-order intensity correlation function characteristic quantity and the actual second-order intensity correlation function characteristic quantity to obtain a residual vector, and using a numerical differentiation method to calculate a corresponding Jacobian matrix;

[0169] In this embodiment, first, a correspondence between the theoretical characteristic quantity and the actual characteristic quantity is established, the theoretical second-order intensity correlation function characteristic quantity is matched with the actual second-order intensity correlation function characteristic quantity one by one to obtain a characteristic quantity correspondence table; according to the correspondence between the theoretical second-order intensity correlation function characteristic quantity and the actual second-order intensity correlation function characteristic quantity, the difference between the theoretical characteristic quantity and the actual characteristic quantity is calculated one by one to obtain an original residual vector; according to the measurement uncertainty and the physical importance information of each characteristic quantity, a weight coefficient corresponding to the original residual vector is calculated, and the original residual vector is weighted and normalized to obtain a standardized residual vector.

[0170] Then, statistical characteristic parameters of the residual are calculated according to the standardized residual vector, the statistical characteristic parameters including a two-norm, a mean and a variance; a quantitative index of fitting error is obtained according to the statistical characteristic parameters; a data quality evaluation result is obtained by using an anomaly detection algorithm to identify data points that significantly deviate, in combination with the quantitative index of fitting error, according to the distribution characteristics of the residual vector; a time series analysis model is established according to historical residual data in the iteration process and the data quality evaluation result, and a residual change trend and a convergence evaluation result are obtained.

[0171] Further, a numerical differentiation calculation scheme of the Jacobian matrix is established according to the physical characteristics and numerical range of the to-be-optimized parameters, and a perturbation strategy corresponding to each parameter is determined; a perturbation step is determined by using an adaptive method according to the numerical size and physical dimension of the parameter, a small perturbation is applied to each parameter, the theoretical second-order intensity correlation function characteristic quantity is recalculated through a joint forward physical model, and characteristic quantity change data after parameter perturbation are obtained; partial derivatives are calculated by using a forward difference formula according to the characteristic quantity change data and the perturbation step information, and each element of the Jacobian matrix is obtained. The corresponding Jacobian matrix is calculated according to each element of the Jacobian matrix.

[0172] S3.3.3: according to the residual vector and the Jacobian matrix, a parameter update amount is calculated by a least square method, and an updated grating complex amplitude distribution and line edge roughness power spectrum parameter are obtained according to the parameter update amount;

[0173] In this embodiment, first, a linearized least square problem is established according to the standardized residual vector and the Jacobian matrix, and a normal equation set is constructed; a suitable solving algorithm is selected according to the condition number and numerical stability characteristics of the Jacobian matrix, and a QR decomposition or singular value decomposition method is used to solve the normal equation set to obtain a parameter update amount vector.

[0174] Then, according to the numerical size and physical rationality of the parameter update vector, the constraint condition of the update is set to prevent the parameter update from being too large to cause the physical model to fail; the trust region method or line search strategy is used to adjust the parameter update to ensure that the updated parameter is still within the physical allowable range; and the updated grating complex amplitude distribution and line edge roughness power spectrum parameters are calculated according to the adjusted parameter update.

[0175] Further, according to the updated grating complex amplitude distribution and line edge roughness power spectrum parameters, the physical consistency and numerical rationality are verified; for the grating complex amplitude distribution parameters, the value range of the amplitude and phase is checked to ensure that the optical physical constraints are met; for the line edge roughness power spectrum parameters, the positive value and dimensional consistency of the correlation length and roughness amplitude parameters are verified; and according to the verification result, the unreasonable parameter values are corrected to obtain the final updated parameters.

[0176] S3.3.4: According to the updated grating complex amplitude distribution and line edge roughness power spectrum parameters, the iteration is repeatedly performed, and when the relative change of the two-norm of the residual vector is less than a preset threshold, the iteration is terminated and the final grating complex amplitude distribution and line edge roughness power spectrum parameters are output.

[0177] In this embodiment, first, the updated grating complex amplitude distribution and line edge roughness power spectrum parameters are taken as new initial parameters, and the iteration calculation process of steps S3.3.1 to S3.3.3 is repeatedly performed; in each iteration, the two-norm value of the current residual vector is recorded to establish an iteration history data sequence.

[0178] Then, according to the residual vector two-norm of the current iteration and the previous iteration result, the relative change is calculated; according to the relative change and the preset convergence threshold, it is judged whether the iteration reaches the convergence condition, and the convergence judgment result is obtained; when the relative change is less than the preset threshold, it is considered that the optimization process has converged, and the iteration calculation is terminated.

[0179] Further, according to the convergence judgment result, the grating complex amplitude distribution and line edge roughness power spectrum parameters obtained by the final optimization are output; the quality of the grating complex amplitude distribution and line edge roughness power spectrum parameters obtained by the final optimization is evaluated, the fitting precision index and parameter uncertainty information are calculated; and according to the fitting precision index and parameter uncertainty information, the joint forward physical model is verified to obtain the final grating complex amplitude distribution and line edge roughness power spectrum parameters.

[0180] Finally, according to the convergence characteristics of the iterative process and the final fitting error, a diagnostic report of the optimization process is generated; key intermediate results and calculation statistics are recorded to provide reference data for subsequent result analysis and method improvement; the final grating complex amplitude distribution and line edge roughness power spectrum parameters are output in a standardized format for grating performance evaluation and manufacturing process optimization.

[0181] Referring to Figure 4 A micro-nano grating line deviation rapid detection system is provided for the embodiment, and the deviation rapid detection system comprises a light source illumination module, an encoding modulation module, a scanning control module, an optical imaging module and a data processing module.

[0182] The light source illumination module is configured to provide a partially coherent light source.

[0183] The encoding modulation module comprises a phase modulator and an amplitude modulator, and is configured to perform phase encoding modulation and amplitude encoding modulation on incident light to generate a first illumination channel and a second illumination channel.

[0184] The scanning control module comprises a scanning driving unit and a position configuration unit, the scanning driving unit is configured to control the structured complex light field to perform point scanning on a plurality of preset scanning positions of the micro-nano grating to be measured, and the position configuration unit is configured to output preset position coordinates to provide a position reference for the point scanning.

[0185] The optical imaging module comprises a Fourier lens, a channel separation optical element and an imaging detector, the Fourier lens is configured to focus multi-stage diffraction light to a Fourier plane, the channel separation optical element is configured to separate diffraction light of each parallel illumination channel, and the imaging detector is configured to acquire multi-frame composite diffraction spot distribution data of the Fourier plane.

[0186] The data processing module is configured to perform grating line deviation rapid detection, output grating line deviation detection results and generate final adjustment instructions.

[0187] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0188] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0189] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0190] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for fast detection of line deviation of micro-nano gratings, characterized in that, Rapid deviation detection methods include: The distribution data of the composite diffraction spot generated by the micro / nano grating under test in the Fourier plane was acquired by multi-frame continuous imaging. The first-order intensity sequence and its temporal fluctuation data of the composite diffraction spot distribution data were calculated, including: Acquire consecutive multi-frame Fourier plane images of the composite diffraction spot distribution data; Based on the spatial distribution functions of the first and second illumination channels, the diffraction spot regions corresponding to each illumination channel are extracted from the multi-frame Fourier plane images using an image segmentation algorithm, and the integrated light intensity value of each diffraction spot region at each time step is calculated. Based on the integrated light intensity value, construct the light intensity response sequence of each illumination channel in chronological order to form a first-order intensity sequence that includes at least two illumination channels; Based on the first-order intensity sequence and its temporal fluctuation data, the actual second-order intensity correlation function characteristics of each lighting channel are calculated through channel separation and registration, including: Based on the center position coordinates and spatial distribution function of the first and second lighting channels, the theoretical diffraction position of each lighting channel in the Fourier plane is determined by a spatial positioning algorithm. Channel separation processing is performed on multiple frames of Fourier plane images. The diffraction spot regions corresponding to each illumination channel are identified by the image segmentation algorithm. The actual center position of each diffraction spot region is located by the sub-pixel level registration algorithm. Based on the actual center position and the theoretical diffraction position, the first-order intensity sequence is corrected using a displacement correction algorithm to obtain the corrected first-order intensity sequence. Based on the corrected first-order intensity sequence, the second-order intensity correlation function between any two illumination channels is calculated through cross-correlation operation. The peak value, full width at half maximum (FWHM), decay time constant, and coherence of the second-order intensity correlation function are extracted as feature parameters. Based on the characteristic parameters, combined with the coherence characteristics and light intensity weights of the illumination channel, the actual second-order intensity correlation function characteristic quantities are obtained through normalization. The actual second-order intensity correlation function features are input into a pre-defined joint forward physical model. Through iterative inversion, the reconstructed grating complex amplitude distribution and line-edge roughness power spectrum parameters are obtained, including: Based on the center position coordinates, spatial distribution function, coherence characteristics, and intensity weights of the first and second illumination channels, a joint forward physical model of micro / nano grating diffraction under coded illumination is constructed based on scalar diffraction theory. Based on the joint forward physical model, a mathematical mapping relationship is constructed from the complex amplitude distribution of the grating, the power spectrum parameters of the line edge roughness to the characteristic quantities of the theoretical second-order intensity correlation function. Based on the mathematical mapping relationship and the actual second-order intensity correlation function characteristics, the residual between the theoretical and actual second-order intensity correlation function characteristics is minimized by using a nonlinear least squares iterative inversion algorithm, thereby obtaining the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters. Based on the reconstructed complex amplitude distribution of the grating and the power spectrum parameters of the line edge roughness, the grating line deviation at the current coordinate position is calculated. Before acquiring composite diffraction spot distribution data, the method also includes: performing phase-encoding modulation, generating a first illumination channel by loading a preset phase mask pattern to impose differential phase delay on different spatial regions of the incident light, wherein a phase distribution of the first illumination channel is adjusted according to a phase delay distribution of the phase mask pattern; performing amplitude-encoding modulation, generating a second illumination channel by loading a preset amplitude mask pattern to selectively transmit different spatial regions of the incident light, wherein an amplitude distribution of the second illumination channel is adjusted by a transmittance distribution of the amplitude mask pattern; generating a plurality of parallel illumination channels with preset spatial spectral distributions according to the first illumination channel and the second illumination channel, wherein coherence characteristics of the parallel illumination channels are different from each other, and generating a structured complex light field by spatial superposition according to the parallel illumination channels and the coherence characteristics.

2. The method for fast detection of line deviation of micro-nano grating according to claim 1, characterized in that, calculating a first-order intensity sequence of the complex diffraction spot distribution data and time fluctuation data thereof, further comprising: calculating intensity variance and a time autocorrelation function of each illumination channel by a statistical analysis method according to the first-order intensity sequence, and obtaining time fluctuation data of each illumination channel according to the intensity variance and the time autocorrelation function.

3. The method according to claim 1, wherein, a nonlinear least squares iterative inversion algorithm, comprising: calculating a theoretical second-order intensity correlation function characteristic quantity according to a preset initial grating complex amplitude distribution and initial line edge roughness power spectrum parameters by a joint forward physical model; comparing the theoretical second-order intensity correlation function characteristic quantity with an actual second-order intensity correlation function characteristic quantity to obtain a residual vector, and calculating a corresponding Jacobian matrix by a numerical differentiation method; calculating a parameter update amount by a least squares method according to the residual vector and the Jacobian matrix, and obtaining an updated grating complex amplitude distribution and line edge roughness power spectrum parameters according to the parameter update amount; repeating the iteration according to the updated grating complex amplitude distribution and line edge roughness power spectrum parameters, and terminating the iteration and outputting the final grating complex amplitude distribution and line edge roughness power spectrum parameters when a two-norm change amount of the residual vector is less than a preset threshold.

4. The method for fast detection of line deviation of micro-nano grating according to claim 3, characterized in that, calculating the grating line deviation at the current coordinate position according to the reconstructed grating complex amplitude distribution and line edge roughness power spectrum parameters, comprising: extracting an amplitude distribution and a phase distribution of the to-be-tested micro-nano grating according to the reconstructed grating complex amplitude distribution, calculating a transmittance distribution and a phase delay distribution of the to-be-tested micro-nano grating, and extracting geometric contours and boundary position information of grating lines by a threshold segmentation and edge detection algorithm; calculating a position deviation amount of the micro-nano grating line at the current coordinate position according to the geometric contours and boundary position information of the grating lines, and obtaining grating line deviation statistical results by statistical analysis; calculating line edge roughness characteristic parameters by function integration according to the line edge roughness power spectrum parameters; obtaining the grating line deviation at the current coordinate position by a preset quality standard according to the grating line deviation statistical results and the line edge roughness characteristic parameters.

5. The method for fast detection of line deviation of micro-nano grating according to claim 4, characterized in that, After the grating line deviation of the current coordinate position is calculated, the method further comprises: scanning a plurality of preset scanning positions of the micro-nano grating to be measured, constructing a two-dimensional deviation distribution map according to the grating line deviation and the line edge roughness index of each scanning position, obtaining a grating line deviation detection result through abnormal area identification, and generating a corresponding adjustment instruction.

6. The method for fast detection of line deviation of micro-nano grating according to claim 5, characterized in that, The grating line deviation detection result is obtained through abnormal area identification, and a corresponding adjustment instruction is generated, including: An abnormal threshold is set according to the two-dimensional deviation distribution map, abnormal nodes are located, and connected abnormal areas are merged through spatial clustering; According to the area and shape characteristics of the connected abnormal areas, the mode recognition is used to determine the abnormal type; According to the abnormal type, a re-inspection strategy is formulated, a re-inspection result is obtained, and the re-inspection strategy includes local high-density scanning and prolonged exposure time; According to the re-inspection result, the true abnormal area is determined through confidence test, and the grating line deviation detection result and the final adjustment instruction are generated.

7. A system for fast detection of line deviation of micro-nano gratings, for implementing the method for fast detection of line deviation of micro-nano gratings according to any one of claims 1-6, characterized in that, The deviation rapid detection system comprises a light source illumination module, a coding modulation module, a scanning control module, an optical imaging module and a data processing module, wherein: The light source illumination module is used to provide a partially coherent light source; The coding modulation module comprises a phase modulator and an amplitude modulator, which are used to perform phase coding modulation and amplitude coding modulation on incident light to generate a first illumination channel and a second illumination channel; The scanning control module comprises a scanning driving unit and a position configuration unit, the scanning driving unit is used to control the structured complex light field to perform point scanning on a plurality of preset scanning positions of the micro-nano grating to be measured, and the position configuration unit is used to output preset position coordinates to provide a position reference for point scanning; The optical imaging module comprises a Fourier lens, a channel separation optical element and an imaging detector, the Fourier lens is used to focus multi-stage diffraction light to a Fourier plane, the channel separation optical element is used to separate the diffraction light of each parallel illumination channel, and the imaging detector is used to acquire a plurality of frames of complex diffraction spot distribution data of the Fourier plane; The data processing module is used to perform grating line deviation rapid detection, output a grating line deviation detection result and generate a final adjustment instruction.

Images