Micro-vibration control method, device, equipment and product

Abstract

The invention provides a micro-vibration control method, device, equipment and product, and relates to the technical field of vibration control. The method comprises the following steps: performing spectral analysis on first vibration data acquired by a first sensor to obtain first vibration frequency division data and second vibration frequency division data; respectively inputting the first vibration frequency division data and the second vibration frequency division data into a cobweb simulation filter layer, respectively carrying out vibration control through a radial fiber structure and a spiral fiber structure, and respectively obtaining first vibration suppression data and second vibration suppression data; and injecting chaotic compensation data of which the vibration amplitude is the first vibration amplitude into the target vibration suppression data of which the vibration amplitude is greater than the safe vibration amplitude, and obtaining second vibration data of which the vibration amplitude is less than the safe vibration amplitude. Therefore, active suppression of vibration data is achieved, sudden vibration can be solved due to the fact that the radial fiber structure can respond quickly, and accumulated defects can be solved through the radial fiber structure and the spiral fiber structure.

Description

Technical Field

[0001] This invention relates to the field of vibration control technology, and in particular to a micro-vibration control method, device, equipment and product. Background Technology

[0002] As semiconductor processes advance to 3nm and below, lithography machines need to achieve exposure precision with linewidths ≤5nm, requiring vibration displacement control within <1nm RMS. This nanometer-level tolerance makes micro-vibration a core limiting factor for yield. Vibration sources are complex and diverse, including internal sources such as air conditioning units (50Hz low-frequency vibration), lithography machine stepper motors (5g acceleration causing 10-200Hz vibration), robotic arm movements, and airflow disturbances; external sources such as subway construction 50 meters away transmitting 2-5μm / s vibrations, and earthquakes causing mirror group displacement in lithography machines. However, traditional vibration control relies on single suppression methods (such as passive vibration isolation tables), which have significant limitations, failing to address both sudden vibrations and accumulated defects; current technologies require compensation to rely on simulation results, leading to response delays. Summary of the Invention

[0003] The purpose of this invention is to provide a micro-vibration control method, device, equipment and product to solve the problem that the passive suppression in existing vibration suppression technology cannot take into account both sudden vibration and cumulative defects.

[0004] To achieve the above objectives, embodiments of the present invention provide a micro-vibration control method, comprising:

[0005] Spectral analysis is performed on the first vibration data collected by the first sensor to obtain first vibration frequency division data and second vibration frequency division data; wherein, the frequency of the first vibration frequency division data is less than a first preset frequency, and the second vibration frequency division data is greater than or equal to the first preset frequency.

[0006] The first vibration frequency division data is input into the spider web simulation filter layer, and the first vibration frequency division data is controlled by the radial fiber structure in the spider web simulation filter layer to obtain the first vibration suppression data.

[0007] The second vibration frequency division data is input into the spider web simulation filter layer, and the second vibration frequency division data is controlled by the spiral fiber structure in the spider web simulation filter layer to obtain the second vibration suppression data.

[0008] Chaotic compensation data with a vibration amplitude of the first vibration amplitude is injected into the target vibration suppression data in the first vibration suppression data and the second vibration suppression data where the vibration amplitude is greater than the safe vibration amplitude. Then, second vibration data with a vibration amplitude of less than the safe vibration amplitude after the target vibration suppression data is injected with chaotic compensation data is obtained. The first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data.

[0009] Optionally, the method, wherein performing spectral analysis on the first vibration data collected by the first sensor to obtain first vibration frequency division data and second vibration frequency division data classified according to a first preset frequency, includes:

[0010] Vibration detection is performed using the first sensor at a sampling rate of a second preset frequency, and raw vibration data is collected.

[0011] The original vibration data is verified for electromagnetic interference by a second sensor, abnormal data points are removed, and the first vibration data is obtained.

[0012] Perform spectral analysis on the first vibration data to obtain the energy distribution results of the first vibration data;

[0013] Based on the energy distribution results, the first vibration data is classified according to the first preset frequency to obtain the first vibration frequency division data and the second vibration frequency division data.

[0014] Optionally, the method, wherein the first vibration frequency division data is input into the spider web simulation filter layer, and vibration control of the first vibration frequency division data is performed through the radial fiber structure in the spider web simulation filter layer to obtain first vibration suppression data, includes:

[0015] An electric current is applied to the shape memory alloy material of the radial fiber structure that makes up the spider web simulation filter layer, and the material is heated to the austenitic phase.

[0016] The vibration energy of the first vibration frequency division data is guided to the damping well through the radial fiber structure to obtain the first vibration suppression data.

[0017] Optionally, the method, wherein the second vibration frequency division data is input into the spider web simulation filter layer, and vibration control is performed on the second vibration frequency division data through the helical fiber structure in the spider web simulation filter layer to obtain second vibration suppression data, includes:

[0018] The spiral fiber structure in the spider web simulation filter layer is activated to absorb the vibration energy of the second vibration frequency division data through resonance.

[0019] A first current is input to the magnetorheological damper connected to the spiral fiber structure to adjust the magnetic field, the viscosity of the suspension is set, and the second vibration suppression data is obtained.

[0020] Optionally, the method, wherein injecting chaotic compensation data with a vibration amplitude of the first vibration amplitude into target vibration suppression data (where the vibration amplitude is greater than the safe vibration amplitude) from the first vibration suppression data and the second vibration suppression data, and obtaining second vibration data (where the vibration amplitude is less than the safe vibration amplitude after injecting the chaotic compensation data), includes:

[0021] If the vibration amplitude of the target vibration suppression data is greater than the safe vibration amplitude, the first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data and the original vibration amplitude; wherein, the original vibration amplitude is the vibration amplitude of the first vibration frequency division data corresponding to the target vibration suppression data before it is input into the spider web simulation filter layer and the vibration amplitude of the target vibration frequency division data in the first vibration frequency division data;

[0022] The vibration suppression data is injected with chaotic compensation data whose vibration amplitude is the first vibration amplitude to obtain the second vibration data.

[0023] Optionally, the method further includes:

[0024] The first vibration data is input into a graph neural network to predict the line width deviation value, and the predicted line width deviation value is obtained.

[0025] If the predicted linewidth deviation value is greater than the safe linewidth deviation value, the predicted compensation vibration amplitude is obtained based on the predicted linewidth deviation value and the first vibration data.

[0026] Inject chaotic compensation data with vibration amplitude equal to the predicted compensated vibration amplitude into the first vibration data to obtain the predicted residual vibration amplitude of the predicted third vibration data;

[0027] Based on the predicted residual vibration amplitude, the vibration amplitude of the first vibration data, and the predicted linewidth deviation value, the predicted corrected linewidth deviation value is obtained;

[0028] If the predicted correction linewidth deviation is less than or equal to the safe linewidth deviation, the predicted compensation vibration amplitude is recorded.

[0029] Optionally, the method includes:

[0030] Establish a first quantization relationship between the first vibration amplitude and the first vibration data;

[0031] Establish a second quantitative relationship between the predicted compensation vibration amplitude and the first vibration data;

[0032] A knowledge graph is obtained based on the first quantization relationship and the second quantization relationship.

[0033] To achieve the above objectives, embodiments of the present invention provide a micro-vibration control device, comprising:

[0034] The first acquisition module is used to perform spectrum analysis on the first vibration data collected by the first sensor to acquire first vibration frequency division data and second vibration frequency division data; wherein, the frequency of the first vibration frequency division data is less than a first preset frequency, and the second vibration frequency division data is greater than or equal to the first preset frequency.

[0035] The second acquisition module is used to input the first vibration frequency division data into the spider web simulation filter layer, and to perform vibration control on the first vibration frequency division data through the radial fiber structure in the spider web simulation filter layer to acquire the first vibration suppression data.

[0036] The third acquisition module is used to input the second vibration frequency division data into the spider web simulation filter layer, and to control the vibration of the second vibration frequency division data through the spiral fiber structure in the spider web simulation filter layer to acquire the second vibration suppression data.

[0037] The fourth acquisition module is used to inject chaotic compensation data with a vibration amplitude of the first vibration amplitude into the target vibration suppression data in the first vibration suppression data and the second vibration suppression data where the vibration amplitude is greater than the safe vibration amplitude, and to acquire second vibration data where the vibration amplitude is less than the safe vibration amplitude after the target vibration suppression data is injected with chaotic compensation data; wherein, the first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data.

[0038] To achieve the above objectives, embodiments of the present invention provide an electronic device, including: a transceiver, a processor, a memory, and a program or instructions stored in the memory and executable on the processor; wherein, when the processor executes the program or instructions, it implements the micro-vibration control method as described above.

[0039] To achieve the above objectives, embodiments of the present invention provide a readable storage medium having a program or instructions stored thereon, wherein the program or instructions, when executed by a processor, implement the steps in the micro-vibration control method as described above.

[0040] To achieve the above objectives, embodiments of the present invention provide a computer program product, which includes computer instructions that, when executed by a processor, implement the steps of the micro-vibration control method as described above.

[0041] The beneficial effects of the above-mentioned technical solution of the present invention are as follows:

[0042] In this embodiment of the invention, the first vibration data collected by the first sensor is divided into first vibration frequency division data and second vibration frequency division data, which are respectively input into the radial fiber structure and helical fiber structure of the spider web simulation filter layer. Chaotic compensation data with a vibration amplitude of the first vibration amplitude is injected into the target vibration suppression data in the first vibration suppression data and the second vibration suppression data where the vibration amplitude is greater than the safe vibration amplitude. Then, second vibration data with a vibration amplitude of less than the safe vibration amplitude after injecting the chaotic compensation data is obtained. This achieves active suppression of vibration data. Furthermore, since the radial fiber structure can respond quickly, it can solve sudden vibrations, and the radial fiber structure and the helical fiber structure can solve cumulative defects. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the micro-vibration control method according to an embodiment of the present invention;

[0044] Figure 2 This is a flowchart of the micro-vibration control method described in an embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of the micro-vibration control device according to an embodiment of the present invention. Detailed Implementation

[0046] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0047] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0048] In various embodiments of the present invention, it should be understood that the sequence number of each process described below does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0049] In addition, the terms "system" and "network" are often used interchangeably in this article.

[0050] In the embodiments provided by this invention, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information.

[0051] For ease of understanding, the following describes some aspects of the embodiments of the present invention:

[0052] like Figure 1 As shown, an embodiment of the present invention provides a micro-vibration control method, which includes:

[0053] S10, perform spectrum analysis on the first vibration data collected by the first sensor to obtain first vibration frequency division data and second vibration frequency division data; wherein, the frequency of the first vibration frequency division data is less than the first preset frequency, and the second vibration frequency division data is greater than or equal to the first preset frequency.

[0054] It should be noted that semiconductor equipment vibration specifications require the suppression of low-frequency vibrations (0.1-100Hz, primarily transmitted from the ground) in equipment such as lithography machines; and the control of high-frequency vibrations (100-5000Hz) to the sub-nanometer level (sensitive frequency band for precision components). Therefore, the first vibration data was analyzed using... Figure 2 Step S10 classifies the vibration data according to the vibration characteristics into first vibration frequency division data and second vibration frequency division data. The first preset frequency is 100Hz, which is divided into low frequency (<100Hz) and high frequency (≥100Hz).

[0055] S20, the first vibration frequency division data is input into the spider web simulation filter layer, and the first vibration frequency division data is controlled by the radial fiber structure in the spider web simulation filter layer to obtain the first vibration suppression data;

[0056] It should be noted that, as Figure 2 As shown, in step S20, the radial fiber structure in the spider web simulation filter layer simulates the radial fibers of the outer ring of a spider web, which is used to suppress the vibration of the first vibration frequency division data and obtain the first vibration suppression data.

[0057] S30, input the second vibration frequency division data into the spider web simulation filter layer, and use the spiral fiber structure in the spider web simulation filter layer to control the vibration of the second vibration frequency division data to obtain the second vibration suppression data;

[0058] It should be noted that, as Figure 2 As shown, in step S30, the spiral fiber structure in the spider web simulation filter layer simulates the spiral fibers in the inner circle of a spider web, which is used to suppress the vibration of the second vibration frequency division data and obtain the second vibration suppression data.

[0059] S40, inject chaotic compensation data with a vibration amplitude of the first vibration amplitude into the target vibration suppression data in the first vibration suppression data and the second vibration suppression data where the vibration amplitude is greater than the safe vibration amplitude, and obtain second vibration data where the vibration amplitude is less than the safe vibration amplitude after injecting the chaotic compensation data into the target vibration suppression data; wherein, the first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data;

[0060] It should be noted that, as Figure 2 As shown, in step S40, for the target vibration suppression data that still fails to meet the standard after passing through the spider web simulation filter layer, a chaotic compensation emergency mode is used to further suppress the target vibration suppression data, thereby obtaining second vibration data with a vibration amplitude less than the safe vibration amplitude.

[0061] In this embodiment, the first vibration data collected by the first sensor is divided into first vibration frequency division data and second vibration frequency division data, which are respectively input into the radial fiber structure and the helical fiber structure of the spider web simulation filter layer. Chaotic compensation data with a vibration amplitude of the first vibration amplitude is injected into the target vibration suppression data with a vibration amplitude greater than the safe vibration amplitude in the first vibration suppression data and the second vibration suppression data. Then, the second vibration data with a vibration amplitude less than the safe vibration amplitude after injecting the chaotic compensation data is obtained, thereby realizing active suppression of vibration data and simultaneously solving sudden vibrations and cumulative defects.

[0062] Optionally, the method, wherein step S10 includes:

[0063] Vibration detection is performed using the first sensor at a sampling rate of a second preset frequency, and raw vibration data is collected.

[0064] The original vibration data is verified for electromagnetic interference by a second sensor, abnormal data points are removed, and the first vibration data is obtained.

[0065] Perform spectral analysis on the first vibration data to obtain the energy distribution results of the first vibration data;

[0066] Based on the energy distribution results, the first vibration data is classified according to the first preset frequency to obtain the first vibration frequency division data and the second vibration frequency division data.

[0067] In this embodiment, such as Figure 2As shown, in the magnetic domain wall sensing layer, vibration is detected by the first sensor at a sampling rate of the second preset frequency (which can be 10kHz), and the original vibration data is collected. For example, the original vibration data is 125Hz vibration with a vibration amplitude of 1.2nm (signal-to-noise ratio 42dB). The original vibration data is then checked for electromagnetic interference by a second sensor, which can be a diamond nitrogen-vacancy (NV) color center sensor. Nanoscale sensing technology based on the quantum effect of diamond NV color centers removes abnormal data points, eliminates electromagnetic interference and temperature drift noise in the magnetic domain wall sensor signal, retains the true vibration data, and obtains the first vibration data. For example, the vibration amplitude of the original vibration data is corrected to 1.18nm.

[0068] Next, we will adopt... Figure 2 The edge computing layer performs spectral analysis on the first vibration data to calculate the vibration energy distribution and obtain the energy distribution result. A first example is provided, where the energy distribution result is 15% for 63Hz, 65% for 400Hz, and 20% for 1000Hz. With the first preset frequency at 100Hz, 15% of the 63Hz vibration data is divided into the first vibration frequency division data, and 65% of the 400Hz vibration data and 20% of the 1000Hz vibration data are divided into the second vibration frequency division data.

[0069] Optionally, the method, wherein step S20 includes:

[0070] An electric current is applied to the shape memory alloy material of the radial fiber structure that makes up the spider web simulation filter layer, and the material is heated to the austenitic phase.

[0071] The vibration energy of the first vibration frequency division data is guided to the damping well through the radial fiber structure to obtain the first vibration suppression data.

[0072] In this embodiment, such as Figure 2 As shown, in step S20, the shape memory alloy material of the radial fiber structure constituting the spiderweb simulation filter layer is energized. Heating to the austenitic phase requires only milliseconds, resulting in a rapid increase in stiffness. Thus, the radial fiber structure forms a rigid conduction network, and due to its rapid response, it can activate to cope with sudden vibrations. The first vibration frequency division data is guided to the damping well at the edge of the plant through the radial fiber structure, preventing overall equipment resonance. In the first example, the vibration amplitude of the 63Hz vibration data was reduced from 1.5nm to 0.4nm (a 73% attenuation), which is superior to the traditional spring damping system that typically attenuates by 40-50%.

[0073] The radial fiber structure requires only a momentary power-on activation, with no continuous energy consumption; its phase change cycle life exceeds 100,000 cycles, making it virtually maintenance-free; the rigid network suppresses low-frequency displacement, reducing alignment errors in the lithography machine. Through frequency division and synchronization processing, it achieves greater efficiency and avoids the limitations of single methods, realizing full-domain vibration suppression.

[0074] Optionally, the method, wherein step S30 includes:

[0075] The spiral fiber structure in the spider web simulation filter layer is activated to absorb the vibration energy of the second vibration frequency division data through resonance.

[0076] A first current is input to the magnetorheological damper connected to the spiral fiber structure to adjust the magnetic field, the viscosity of the suspension is set, and the second vibration suppression data is obtained.

[0077] In this embodiment, such as Figure 2 As shown, in step S20, the spiral fiber structure simulating the inner ring of a spider web is combined with a magnetorheological damper, and the viscosity of the suspension is changed by adjusting the current to achieve resonant absorption. In the first example, the vibration amplitude of the 400Hz vibration data was reduced from 0.8nm to 0.1nm (attenuation of 87.5%), which far exceeds the attenuation of approximately 50% in the passive rubber vibration isolation pads of the prior art.

[0078] The spiral fiber structure dissipates energy through resonance during high-frequency vibrations, providing localized protection for precision components. The viscosity of the magnetorheological fluid can be adjusted in real time (milliseconds) to adapt to different high-frequency vibration characteristics. High-frequency energy is locally absorbed, preventing its conduction to other sensitive areas. The overall response delay is <1ms, solving the latency problem caused by reliance on traditional simulations, and the magnetorheological fluid has low power consumption. Even in the event of local fiber damage, the overall function is not affected (similar to how a broken section of a spider web can still function).

[0079] The data for the first example is shown in Table 1:

[0080] Vibration type Handling method Input amplitude Output amplitude Decrease 63Hz (low frequency) Radial fiber structure + damping well 1.5nm 0.4nm 73％ 400Hz (high frequency) Helical fiber structure + magnetorheological damper 0.8nm 0.1nm 87.5％

[0081] Optionally, the method, wherein step S40 includes:

[0082] If the vibration amplitude of the target vibration suppression data is greater than the safe vibration amplitude, the first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data and the original vibration amplitude; wherein, the original vibration amplitude is the vibration amplitude of the first vibration frequency division data corresponding to the target vibration suppression data before it is input into the spider web simulation filter layer and the vibration amplitude of the target vibration frequency division data in the first vibration frequency division data;

[0083] The vibration suppression data is injected with chaotic compensation data whose vibration amplitude is the first vibration amplitude to obtain the second vibration data.

[0084] In this embodiment, if the vibration amplitude of the target vibration suppression data is still greater than the safe vibration amplitude after suppression by the spider web simulation filter layer, then proceeding to... Figure 2 The chaos compensation emergency mode in step S40. Chaos compensation is a control strategy designed for chaotic systems (quasi-random motion generated by deterministic nonlinear systems). It eliminates or suppresses chaotic behavior through specific techniques, stabilizing the system in the desired state.

[0085] In the second example, the safe vibration amplitude is set to 0.5 nm. The vibration amplitude of the target vibration suppression data at 63 Hz, transmitted through the radial fiber structure of the spiderweb simulation filter layer, is suppressed from the original amplitude of 1.5 nm to 0.8 nm in the spiderweb simulation filter layer. This is still greater than the safe vibration amplitude, therefore, it needs to be suppressed again through the emergency mode. The suppression target is to reduce the vibration amplitude of the target vibration suppression data to below 0.2 nm. The first vibration amplitude is obtained using the following formula:

[0086]

[0087] Among them, A 残余 A represents the vibration amplitude of the target vibration suppression data after passing through the radial fiber structure of the spiderweb-simulated filter layer. 原始 A represents the vibration amplitude of the target vibration suppression data before it passes through the radial fiber structure of the spiderweb simulation filter layer. 补偿 The first vibration amplitude is Δφ, which is the experimentally calibrated anti-phase interference value, usually set to 160°.

[0088] In the second example, the first vibration amplitude calculated using the above formula is 0.68 nm. A chaotic vibration of 0.68 nm is injected into the target vibration suppression data, resulting in a vibration amplitude of the second vibration data between 0.15 and 0.2 nm, which is less than the set safe vibration amplitude of 0.5 nm. The data from the second example are shown in Table 2.

[0089] stage vibration amplitude Decrease Target vibration frequency division data 1.5nm (63Hz) X Target vibration suppression data 0.8nm 46.7％ Second vibration data (injected at 0.68 nm) 0.18nm 88％

[0090] Where X represents the absence of a decrease in the target vibration frequency division data stage.

[0091] Optionally, the method further includes:

[0092] The first vibration data is input into a graph neural network to predict the line width deviation value, and the predicted line width deviation value is obtained.

[0093] If the predicted linewidth deviation value is greater than the safe linewidth deviation value, the predicted compensation vibration amplitude is obtained based on the predicted linewidth deviation value and the first vibration data.

[0094] Inject chaotic compensation data with vibration amplitude equal to the predicted compensated vibration amplitude into the first vibration data to obtain the predicted residual vibration amplitude of the predicted third vibration data;

[0095] Based on the predicted residual vibration amplitude, the vibration amplitude of the first vibration data, and the predicted linewidth deviation value, the predicted corrected linewidth deviation value is obtained;

[0096] If the predicted correction linewidth deviation is less than or equal to the safe linewidth deviation, the predicted compensation vibration amplitude is recorded.

[0097] In this embodiment, such as Figure 2 As shown, the first vibration data is used to predict the linewidth deviation value through a graph neural network (GNN) in the digital twin prediction layer, thus obtaining the predicted linewidth deviation value. The formula for the graph neural network is as follows:

[0098]

[0099] in, Let N(i) be the feature vector of node i at layer l+1 (containing vibration spectrum, equipment parameters, etc.), and let N(i) be the set of neighboring nodes of node i (representing associated sensors or process links in digital twins). ij W is the normalization coefficient for nodes i and j (the reciprocal of the sensor / device node distance). (l) σ is the trainable weight matrix of the l-th layer (learning the mapping relationship between vibration and defects through historical data), and σ is a non-linear activation function (such as ReLU) that can enhance the expressive power of the model.

[0100] In the third example, the first vibration data has a vibration frequency of 1000 Hz and a vibration amplitude of 0.5 nm. The lithography machine parameters are: exposure dose 30 mJ / cm². 2 Numerical Aperture (NA) = 0.33.

[0101] Step 1: Initialize the features. (The data in the formula are vibration amplitude, exposure dose, and NA, respectively).

[0102] Step 2: Perform neighbor aggregation calculation (taking the first layer as an example):

[0103] Topology:

[0104] The current node i has 3 neighboring nodes j1, j2, j3, and their feature vectors are the same (simplifying the calculation).

[0105] Normalization coefficient c ij =1.2 (based on spacing calibration).

[0106] Weight matrix W (0) (3×3, post-training parameters):

[0107]

[0108] Aggregate calculation:

[0109]

[0110] Among them, c ij W is the normalization coefficient. (0) This is the weight matrix. To initialize the feature vector.

[0111] Step-by-step results:

[0112] 1. Matrix multiplication:

[0113]

[0114] 2. Normalize and aggregate neighbors:

[0115]

[0116] Step 3: Nonlinear activation (ReLU function):

[0117]

[0118] Step 4, Iterative Propagation (L=3 layers):

[0119] Repeat steps two through three above to finally output the layer features. Mapped to the predicted linewidth deviation: Predicted linewidth deviation = 1.2nm.

[0120] If the predicted linewidth deviation is greater than the safe linewidth deviation value when the safe linewidth deviation value is 1nm, i.e. the prediction deviation exceeds the threshold, then the chaotic compensation prediction mode is entered.

[0121] In the chaotic compensation prediction mode, chaotic compensation data with a vibration amplitude equal to the predicted compensation vibration amplitude is injected into the first vibration data to obtain the predicted residual vibration amplitude of the predicted third vibration data. The formula for the predicted compensation vibration amplitude is as follows:

[0122] A 补偿 = 0.3 × predicted line width deviation + 0.1 × historical error

[0123] Among them, A 补偿 The predicted compensation vibration amplitude is defined as follows: the historical error is the deviation between the model prediction and the actual measurement obtained from historical data (this can be set as needed). In the third example, the historical error is 0.1 nm.

[0124] The calculation result is 0.3×1.2nm+0.1×0.1nm=0.36nm+0.01nm=0.4nm.

[0125] Next, the predicted residual vibration amplitude is obtained using the following formula:

[0126]

[0127] Among them, A 残余 For the predicted residual vibration amplitude, A 原始 The vibration amplitude of the first vibration data is 0.5 nm. 补偿 The predicted compensation vibration amplitude is 0.4 nm, and Δφ is the experimentally calibrated anti-phase interference value, which is usually set to 160°.

[0128] The calculated predicted residual vibration amplitude is 0.2 nm.

[0129] Vibration suppression ratio: (A) 原始 -A 残余 ) / A 原始 =(0.5nm-0.2nm) / 0.5nm=60%.

[0130] The formula for predicting and correcting the linewidth deviation value is as follows:

[0131] ΔCD correction = ΔCD original × (1 - inhibition rate × process coefficient) = 1.2nm × (1 - 60% × 0.7) = 0.7nm.

[0132] The process transfer coefficient is the sensitivity coefficient of the photoresist to vibration, which is 0.7 in the third example.

[0133] The data for the third example is shown in Table 3:

[0134] stage vibration amplitude Line width deviation initial vibration 0.5nm (1000Hz) Predicted linewidth deviation: 1.2 nm After chaos compensation (0.4nm injection) 0.2nm X After process effect correction X The predicted linewidth deviation is 0.7 nm.

[0135] Here, X represents data that is not recorded or calculated.

[0136] Optionally, the method includes:

[0137] Establish a first quantization relationship between the first vibration amplitude and the first vibration data;

[0138] Establish a second quantitative relationship between the predicted compensation vibration amplitude and the first vibration data;

[0139] A knowledge graph is obtained based on the first quantization relationship and the second quantization relationship.

[0140] In this embodiment, such as Figure 2 As shown, by establishing the first quantization relationship between the first vibration amplitude and the first vibration data, and the second quantization relationship between the predicted compensation vibration amplitude and the first vibration data, a quantization relationship between vibration characteristics and chip defects is established, thus forming the knowledge graph.

[0141] It should be noted that in this embodiment of the invention, the first sensor installed on the production equipment collects vibration data caused by various factors such as the environment or equipment during the production process. On one hand, a biological combination technology using a spider web simulation filter layer is employed to suppress vibration data of different frequencies. If the suppression by the spider web simulation filter layer fails to bring the vibration amplitude to a preset safe vibration amplitude, chaotic compensation data is injected for further suppression, resulting in vibration data that better meets production requirements. On the other hand, based on the collected vibration data, linewidth deviation values ​​are predicted. The simulation then injects the predicted chaotic compensation data, calculated from the vibration data and the predicted linewidth deviation, into the vibration data to achieve vibration data that meets production requirements. This prevents vibration data that does not meet production requirements from affecting the production process. By combining vibration suppression during the production process with preventative vibration suppression, and optimizing through a genetic algorithm to update the parameters of the spider web simulation filter layer and chaotic compensation data, continuous real-time dynamic learning is performed, thereby achieving a closed-loop quality feedback mechanism. Figure 3 As shown, to achieve the above objectives, embodiments of the present invention provide a micro-vibration control device, comprising:

[0142] The first acquisition module 301 is used to perform spectrum analysis on the first vibration data collected by the first sensor to acquire first vibration frequency division data and second vibration frequency division data; wherein, the frequency of the first vibration frequency division data is less than a first preset frequency, and the second vibration frequency division data is greater than or equal to the first preset frequency.

[0143] The second acquisition module 302 is used to input the first vibration frequency division data into the spider web simulation filter layer, and to perform vibration control on the first vibration frequency division data through the radial fiber structure in the spider web simulation filter layer to acquire the first vibration suppression data.

[0144] The third acquisition module 303 is used to input the second vibration frequency division data into the spider web simulation filter layer, and to control the vibration of the second vibration frequency division data through the spiral fiber structure in the spider web simulation filter layer to acquire the second vibration suppression data.

[0145] The fourth acquisition module 304 is used to inject chaotic compensation data with a vibration amplitude of the first vibration amplitude into the target vibration suppression data in the first vibration suppression data and the second vibration suppression data where the vibration amplitude is greater than the safe vibration amplitude, and to acquire second vibration data where the vibration amplitude is less than the safe vibration amplitude after the target vibration suppression data is injected with chaotic compensation data; wherein, the first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data.

[0146] Optionally, in the aforementioned apparatus, the first acquisition module 301 includes:

[0147] The first processing unit is used to perform vibration detection using the first sensor at a sampling rate of a second preset frequency and to collect raw vibration data.

[0148] The first acquisition unit is used to verify electromagnetic interference by using a second sensor to remove abnormal data points and acquire the first vibration data.

[0149] The second acquisition unit is used to perform spectrum analysis on the first vibration data to obtain the energy distribution result of the first vibration data;

[0150] The third acquisition unit is used to classify the first vibration data according to the first preset frequency based on the energy distribution result, and acquire the first vibration frequency division data and the second vibration frequency division data.

[0151] Optionally, in the aforementioned apparatus, the second acquisition module 302 includes:

[0152] The second processing unit is used to energize and heat the shape memory alloy material of the radial fiber structure that makes up the spider web simulation filter layer to the austenitic phase.

[0153] The fourth acquisition unit is used to guide the vibration energy of the first vibration frequency division data to the damping well through the radial fiber structure, and acquire the first vibration suppression data.

[0154] Optionally, in the aforementioned apparatus, the third acquisition module 303 includes:

[0155] The third processing unit is used to activate the spiral fiber structure in the spider web simulation filter layer and absorb the vibration energy of the second vibration frequency division data through resonance.

[0156] The fifth acquisition unit is used to input a first current to the magnetorheological damper connected to the spiral fiber structure to adjust the magnetic field, set the viscosity of the suspension, and acquire the second vibration suppression data.

[0157] Optionally, in the aforementioned apparatus, the fourth acquisition module 304 includes:

[0158] The sixth acquisition unit is used to acquire the first vibration amplitude based on the vibration amplitude of the target vibration suppression data and the original vibration amplitude when the vibration amplitude of the target vibration suppression data is greater than the safe vibration amplitude; wherein, the original vibration amplitude is the vibration amplitude of the first vibration frequency division data corresponding to the target vibration suppression data before it is input into the spider web simulation filter layer and the vibration amplitude of the target vibration frequency division data in the first vibration frequency division data;

[0159] The seventh acquisition unit is used to inject chaotic compensation data with a vibration amplitude of the first vibration amplitude into the vibration suppression data to acquire the second vibration data.

[0160] Optionally, the device further includes:

[0161] The fifth acquisition module is used to input the first vibration data into a graph neural network to predict the line width deviation value and obtain the predicted line width deviation value.

[0162] The sixth acquisition module is used to acquire the predicted compensation vibration amplitude based on the predicted line width deviation value and the first vibration data when the predicted line width deviation value is greater than the safe line width deviation value.

[0163] The seventh acquisition module is used to inject chaotic compensation data with vibration amplitude as the predicted compensation vibration amplitude into the first vibration data to obtain the predicted residual vibration amplitude of the predicted third vibration data.

[0164] The eighth acquisition module is used to acquire a predicted correction linewidth deviation value based on the predicted residual vibration amplitude, the vibration amplitude of the first vibration data, and the predicted linewidth deviation value.

[0165] The first processing module is used to record the predicted compensation vibration amplitude when the predicted correction linewidth deviation value is less than or equal to the safe linewidth deviation value.

[0166] Optionally, the device includes:

[0167] The second processing module is used to establish a first quantization relationship between the first vibration amplitude and the first vibration data;

[0168] The third processing module is used to establish a second quantitative relationship between the predicted compensation vibration amplitude and the first vibration data;

[0169] The ninth acquisition module is used to acquire the knowledge graph based on the first quantization relationship and the second quantization relationship.

[0170] It should be noted that the apparatus provided in this embodiment of the invention can implement all the method steps implemented in the above method embodiment and can achieve the same technical effect. Therefore, the parts and beneficial effects that are the same as those in the method embodiment will not be described in detail here.

[0171] To achieve the above objectives, embodiments of the present invention provide an electronic device, including: a transceiver, a processor, a memory, and a program or instructions stored in the memory and executable on the processor; wherein, when the processor executes the program or instructions, it implements the micro-vibration control method as described above.

[0172] To achieve the above objectives, embodiments of the present invention provide a readable storage medium having a program or instructions stored thereon, wherein the program or instructions, when executed by a processor, implement the steps in the micro-vibration control method as described above.

[0173] To achieve the above objectives, embodiments of the present invention provide a computer program product, which includes computer instructions that, when executed by a processor, implement the steps of the micro-vibration control method as described above.

[0174] It should be further noted that the terminals described in this specification include, but are not limited to, smartphones, tablets, etc., and many of the functional components described are referred to as modules in order to emphasize the independence of their implementation.

[0175] In this embodiment of the invention, the module can be implemented in software so that it can be executed by various types of processors. For example, an identified executable code module may include one or more physical or logical blocks of computer instructions, which may be constructed as objects, procedures, or functions. Nevertheless, the executable code of the identified module does not need to be physically located together, but may include different instructions stored in different bits, which, when logically combined, constitute the module and achieve the module's intended purpose.

[0176] In practice, an executable code module can be a single instruction or many instructions, and can even be distributed across multiple different code segments, different programs, and across multiple memory devices. Similarly, operational data can be identified within the module and can be implemented in any suitable form and organized within any suitable type of data structure. This operational data can be collected as a single dataset or distributed across different locations (including different storage devices), and can exist, at least in part, solely as electronic signals within the system or network.

[0177] When a module can be implemented using software, considering the current level of hardware technology, modules that can be implemented in software can be implemented using hardware circuits by those skilled in the art to achieve the corresponding functions, without considering cost. These hardware circuits include conventional very-large-scale integrated circuits (VLSI) or gate arrays, as well as existing semiconductors such as logic chips and transistors, or other discrete components. Modules can also be implemented using programmable hardware devices, such as field-programmable gate arrays, programmable array logic, and programmable logic devices.

[0178] The exemplary embodiments described above are with reference to the accompanying drawings. Many different forms and embodiments are feasible without departing from the spirit and teachings of the invention. Therefore, the invention should not be construed as limiting the exemplary embodiments set forth herein. Rather, these exemplary embodiments are provided to make the invention complete and convey its scope to those skilled in the art. In these drawings, component dimensions and relative dimensions may be exaggerated for clarity. The terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. As used herein, unless clearly indicated otherwise, the singular forms “a,” “an,” and “the” are intended to include all such forms. It will be further understood that the terms “comprising” and / or “including”, when used in this specification, indicate the presence of the stated features, integers, steps, operations, components, and / or elements, but do not exclude the presence or addition of one or more other features, integers, steps, operations, components, and / or groups thereof. Unless otherwise indicated, when stated, a range of values ​​includes the upper and lower limits of the range and any subranges in between.

[0179] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A micro-vibration control method, characterized in that, include: Spectral analysis is performed on the first vibration data collected by the first sensor to obtain first vibration frequency division data and second vibration frequency division data; wherein, the frequency of the first vibration frequency division data is less than a first preset frequency, and the second vibration frequency division data is greater than or equal to the first preset frequency. The first vibration frequency division data is input into the spider web simulation filter layer, and the first vibration frequency division data is controlled by the radial fiber structure in the spider web simulation filter layer to obtain the first vibration suppression data. The second vibration frequency division data is input into the spider web simulation filter layer, and the second vibration frequency division data is controlled by the spiral fiber structure in the spider web simulation filter layer to obtain the second vibration suppression data. Chaotic compensation data with a vibration amplitude of the first vibration amplitude is injected into the target vibration suppression data in the first vibration suppression data and the second vibration suppression data where the vibration amplitude is greater than the safe vibration amplitude. Then, second vibration data with a vibration amplitude of less than the safe vibration amplitude after the target vibration suppression data is injected with chaotic compensation data is obtained. The first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data.

2. The method according to claim 1, characterized in that, Spectral analysis is performed on the first vibration data collected by the first sensor to obtain first vibration frequency division data and second vibration frequency division data classified according to a first preset frequency, including: Vibration detection is performed using the first sensor at a sampling rate of a second preset frequency, and raw vibration data is collected. The original vibration data is verified for electromagnetic interference by a second sensor, abnormal data points are removed, and the first vibration data is obtained. Perform spectral analysis on the first vibration data to obtain the energy distribution results of the first vibration data; Based on the energy distribution results, the first vibration data is classified according to the first preset frequency to obtain the first vibration frequency division data and the second vibration frequency division data.

3. The method according to claim 1, characterized in that, The first vibration frequency division data is input into the spider web simulation filter layer. Vibration control is applied to the first vibration frequency division data through the radial fiber structure in the spider web simulation filter layer to obtain first vibration suppression data, including: An electric current is applied to the shape memory alloy material of the radial fiber structure that makes up the spider web simulation filter layer, and the material is heated to the austenitic phase. The vibration energy of the first vibration frequency division data is guided to the damping well through the radial fiber structure to obtain the first vibration suppression data.

4. The method according to claim 1, characterized in that, The second vibration frequency division data is input into the spider web simulation filter layer. Vibration control is then performed on the second vibration frequency division data through the helical fiber structure in the spider web simulation filter layer to obtain second vibration suppression data, including: The spiral fiber structure in the spider web simulation filter layer is activated to absorb the vibration energy of the second vibration frequency division data through resonance. A first current is input to the magnetorheological damper connected to the spiral fiber structure to adjust the magnetic field, the viscosity of the suspension is set, and the second vibration suppression data is obtained.

5. The method according to claim 1, characterized in that, Injecting chaotic compensation data with a vibration amplitude greater than the safe vibration amplitude into the target vibration suppression data (both the first and second vibration suppression data), and then obtaining second vibration data (where the vibration amplitude is less than the safe vibration amplitude after injecting the chaotic compensation data), including: If the vibration amplitude of the target vibration suppression data is greater than the safe vibration amplitude, the first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data and the original vibration amplitude; wherein, the original vibration amplitude is the vibration amplitude of the first vibration frequency division data corresponding to the target vibration suppression data before it is input into the spider web simulation filter layer and the vibration amplitude of the target vibration frequency division data in the first vibration frequency division data; The vibration suppression data is injected with chaotic compensation data whose vibration amplitude is the first vibration amplitude to obtain the second vibration data.

6. The method according to claim 1, characterized in that, The method further includes: The first vibration data is input into a graph neural network to predict the line width deviation value, and the predicted line width deviation value is obtained. If the predicted linewidth deviation value is greater than the safe linewidth deviation value, the predicted compensation vibration amplitude is obtained based on the predicted linewidth deviation value and the first vibration data. Inject chaotic compensation data with vibration amplitude equal to the predicted compensated vibration amplitude into the first vibration data to obtain the predicted residual vibration amplitude of the predicted third vibration data; Based on the predicted residual vibration amplitude, the vibration amplitude of the first vibration data, and the predicted linewidth deviation value, the predicted corrected linewidth deviation value is obtained; If the predicted correction linewidth deviation is less than or equal to the safe linewidth deviation, the predicted compensation vibration amplitude is recorded.

7. The method according to claims 1 and 6, characterized in that, The method includes: Establish a first quantization relationship between the first vibration amplitude and the first vibration data; Establish a second quantitative relationship between the predicted compensation vibration amplitude and the first vibration data; A knowledge graph is obtained based on the first quantization relationship and the second quantization relationship.

8. A micro-vibration control device, characterized in that, include: The first acquisition module is used to perform spectrum analysis on the first vibration data collected by the first sensor to acquire first vibration frequency division data and second vibration frequency division data; wherein, the frequency of the first vibration frequency division data is less than a first preset frequency, and the second vibration frequency division data is greater than or equal to the first preset frequency. The second acquisition module is used to input the first vibration frequency division data into the spider web simulation filter layer, and to perform vibration control on the first vibration frequency division data through the radial fiber structure in the spider web simulation filter layer to acquire the first vibration suppression data. The third acquisition module is used to input the second vibration frequency division data into the spider web simulation filter layer, and to control the vibration of the second vibration frequency division data through the spiral fiber structure in the spider web simulation filter layer to acquire the second vibration suppression data. The fourth acquisition module is used to inject chaotic compensation data with a vibration amplitude of the first vibration amplitude into the target vibration suppression data in the first vibration suppression data and the second vibration suppression data where the vibration amplitude is greater than the safe vibration amplitude, and to acquire second vibration data where the vibration amplitude is less than the safe vibration amplitude after the target vibration suppression data is injected with chaotic compensation data; wherein, the first vibration amplitude is obtained based on the vibration amplitude of the target vibration suppression data.

9. An electronic device, comprising: A transceiver, a processor, a memory, and a program or instructions stored in the memory and executable on the processor; characterized in that, when the processor executes the program or instructions, it implements the micro-vibration control method as described in any one of claims 1-7.

10. A readable storage medium having a program or instructions stored thereon, characterized in that, When the program or instructions are executed by the processor, they implement the steps in the micro-vibration control method as described in any one of claims 1-7.

11. A computer program product, characterized in that, It includes computer instructions, which, when executed by a processor, implement the steps of the micro-vibration control method as described in any one of claims 1-7.

Images