An automated coating method and system

Abstract

The application discloses an automatic coating method and system, and relates to the technical field of industrial control, which comprises the following steps: collecting coating distribution characteristics and gap distribution characteristics of each coating section of a target substrate based on a multi-section coating control model, and obtaining a characteristic array with a set control length through characteristic similarity dimension reduction processing; after the pre-trained coating retrieval characteristics and gap retrieval characteristics are combined, the cohesion mechanism is adopted to extract the aggregation characteristics of the fusion core elements; the gating fusion characteristics of the aggregation characteristics and the coating and gap characteristic arrays are respectively extracted through a double-feature gating mechanism; and multi-section intermittent coating control instructions are generated based on the fusion characteristics. The application optimizes the data processing efficiency through dynamic characteristic dimension reduction, realizes the accurate association of core process elements through the cohesion mechanism and the double-feature gating fusion, effectively improves the coating uniformity and gap precision, is suitable for high-precision multi-section coating scenes of substrates such as bottom-coated foils, enhances the multi-section collaborative control capability, and improves the product yield.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of industrial control, in particular to an automatic coating method and system. BACKGROUND

[0002] Multi-section intermittent coating is a key process in the field of precision coating of lithium battery pole pieces, flexible electronic devices, etc., and the core is to realize the coordinated control of the uniformity of the coating section thickness and the precision of the gap section. In the prior art, the traditional coating control method relies on manual experience or simple parameter closed-loop adjustment, and it is difficult to dynamically associate the coating distribution characteristics and gap distribution characteristics in the multi-section coating process. At the same time, the original collected characteristic data has high dimension and much redundant information, and direct use for control will lead to a sharp increase in model complexity and response delay, which cannot meet the needs of high-precision coating. SUMMARY

[0003] The purpose of the present application is to provide an automatic coating method and system.

[0004] In a first aspect, an embodiment of the present application provides an automatic coating method, comprising:

[0005] Based on a multi-section coating control model, data acquisition is performed on each coating section in the target substrate to obtain corresponding coating distribution characteristics and gap distribution characteristics, and based on the characteristic similarity, dimensionality reduction processing is performed on the obtained coating distribution characteristic set and gap distribution characteristic set respectively to obtain coating distribution characteristic arrays and gap distribution characteristic arrays of a set control length;

[0006] After merging the pre-trained coating retrieval features and gap retrieval features, the corresponding aggregation characteristics are extracted using a cohesion mechanism; the coating retrieval features and the gap retrieval features are obtained by training the corresponding adjustable retrieval features based on typical coating samples; the aggregation characteristics fuse the core elements in the coating retrieval features and the gap retrieval features;

[0007] A double-feature gating mechanism is used to extract first gating fusion features between the aggregation characteristics and each coating distribution feature in the coating distribution characteristic array, and second gating fusion features between the aggregation characteristics and each gap distribution feature in the gap distribution characteristic array;

[0008] Based on the obtained first gating fusion features and second gating fusion features, the multi-section intermittent coating control of the target substrate is controlled.

[0009] In a possible implementation, the multi-section coating control model comprises a gated feature encoder; the first gated fusion features between the aggregated feature and each coating distribution feature in the coating distribution feature array and the second gated fusion features between the aggregated feature and each gap distribution feature in the gap distribution feature array are extracted respectively by adopting a double-feature gated mechanism, which comprises:

[0010] splitting the aggregated feature into a first aggregated feature corresponding to the coating distribution feature and a second aggregated feature corresponding to the gap distribution feature;

[0011] extracting the first gated fusion features between the first aggregated feature, each coating distribution feature in the coating distribution feature array and the second gated fusion features between the second aggregated feature, each gap distribution feature in the gap distribution feature array by a first gating unit in the gated feature encoder based on the double-feature gated mechanism;

[0012] extracting the first gated fusion features between the first aggregated feature, each coating distribution feature in the coating distribution feature array and the second gated fusion features between the second aggregated feature, each gap distribution feature in the gap distribution feature array by a second gating unit in the gated feature encoder based on the double-feature gated mechanism.

[0013] In a possible implementation, the gated feature encoder comprises sequentially connected gated feature processors; for the first gated feature processor, the current processing feature comprises the coating distribution feature array and the gap distribution feature array; for other gated feature processors, the current processing feature comprises the first gated fusion features and the second gated fusion features output by the previous gated feature processor;

[0014] the multi-section intermittent coating control of the target substrate is controlled based on the obtained first gated fusion features and second gated fusion features, which comprises:

[0015] the multi-section intermittent coating control of the target substrate is controlled based on the first gated fusion features and the second gated fusion features output by the last gated feature processor.

[0016] In a possible implementation, the multi-section coating control model comprises a coating parameter generation model; the multi-section intermittent coating control of the target substrate is controlled based on the obtained first gated fusion features and second gated fusion features, which comprises:

[0017] the first gated fusion features and the second gated fusion features are combined and converted into coating data features in the same feature domain as the dimension of the coating process parameters of the target substrate;

[0018] The coating process description features and coating control parameters of the target substrate are taken as coating parameter guide information, and the coating data features are input into the coating parameter generation model to generate multi-section intermittent coating control instructions for the target substrate.

[0019] In a possible implementation, each gap distribution feature corresponds to one gap sub-section in a coating section; the acquired coating distribution feature set and gap distribution feature set are respectively subjected to dimension reduction processing based on feature similarity to obtain coating distribution feature arrays and gap distribution feature arrays of a set control length, including:

[0020] Each time a coating section is collected, the coating distribution features of the coating section are written into a coating global storage unit, and the gap distribution features of the coating section are written into a gap local storage unit;

[0021] When the number of coating distribution features stored in the coating global storage unit reaches the set control length, the coating distribution features of at least two coating sections in the coating global storage unit are subjected to feature aggregation based on feature similarity, and the coating distribution features to be written are written into the coating global storage unit;

[0022] In addition, for each gap sub-section, when the number of corresponding gap distribution features stored in the gap local storage unit reaches the set control length, the corresponding gap distribution features of at least two coating sections in the gap local storage unit are subjected to feature aggregation based on feature similarity, and the gap distribution features to be written are written into the gap local storage unit.

[0023] In a possible implementation, before the first gating fusion features between the aggregation characteristics and each coating distribution feature in the coating distribution feature array and the second gating fusion features between the aggregation characteristics and each gap distribution feature in the gap distribution feature array are extracted respectively by using the double-feature gating mechanism, the method further includes:

[0024] When a set control node is reached, the coating distribution feature array stored in the coating global storage unit is read, and the gap distribution feature array stored in the gap local storage unit is read.

[0025] In a possible implementation, the features to be aggregated of at least two coating sections are subjected to feature aggregation in any of the following manners:

[0026] Based on feature similarity, the features to be aggregated of the continuous coating sections closest in parameters in the storage unit are subjected to mean value calculation;

[0027] The to-be-aggregated features to be written are grouped and classified with the to-be-aggregated features already written in the storage unit based on feature similarity, and the to-be-aggregated features of each coated section after grouping and classification are subjected to mean value calculation.

[0028] The to-be-aggregated features are coating distribution features or gap distribution features, and the storage unit is a coating global storage unit or a gap local storage unit.

[0029] In a possible implementation, if there are multiple groups of continuous coated sections with the closest parameters, the to-be-aggregated features of the continuous coated sections with the closest parameters in the storage unit are subjected to mean value calculation, including:

[0030] The to-be-aggregated features of the continuous coated sections with the closest parameters and written first in the storage unit are subjected to mean value calculation; or

[0031] The to-be-aggregated features of each group of continuous coated sections with the closest parameters in the storage unit are subjected to mean value calculation respectively.

[0032] In a possible implementation, the target substrate is a foil substrate with a primer; and the data of each coated section in the target substrate is collected respectively to obtain corresponding coating distribution features and gap distribution features, including:

[0033] The primer data and white space data of the current coated section are collected by detecting the three different color regions of the foil substrate, i.e., the empty foil region, the primer region and the coating region, through a high-precision color scale color sensor.

[0034] Based on the primer data and white space data, the coating length feature of the coated section and the white space length feature of the gap sub-section are obtained, and the coating length feature is taken as the coating distribution feature and the white space length feature is taken as the gap distribution feature.

[0035] In the second aspect, an embodiment of the present application provides a server system, including a server, which is configured to execute the method in the first aspect.

[0036] Compared to existing technologies, the beneficial effects of this invention include: Utilizing an automated coating method and system disclosed herein, the coating distribution characteristics and gap distribution characteristics of each coating segment on the target substrate are collected based on a multi-segment coating control model. A feature array with a set controllable length is obtained through feature similarity dimensionality reduction. After merging the pre-trained coating retrieval features and gap retrieval features, a cohesion mechanism is used to extract the aggregation characteristics of the fusion core elements. A dual-feature gating mechanism is used to extract gating fusion features between the aggregation characteristics and the coating and gap feature arrays, respectively. Multi-segment intermittent coating control commands are generated based on the fusion features. This invention optimizes data processing efficiency through dynamic feature dimensionality reduction, and achieves precise correlation between core process elements through cohesion mechanism and dual-feature gating fusion, effectively improving coating uniformity and gap accuracy. It is suitable for high-precision multi-segment coating scenarios on substrates such as substrates with undercoated foils, enhancing multi-segment collaborative control capabilities and improving product yield. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as limiting the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram of the steps of the automated coating method provided in the embodiments of the present invention;

[0039] Figure 2 A schematic block diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0041] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0042] In order to solve the technical problems mentioned in the background art Figure 1 This is a schematic flowchart of an automated coating method provided in an embodiment of the present disclosure. The automated coating method will be described in detail below.

[0043] Step S201: Based on the multi-segment coating control model, data is collected from each coating segment in the target substrate to obtain the corresponding coating distribution features and gap distribution features. Based on feature similarity, the obtained coating distribution feature set and gap distribution feature set are dimensionality reduced to obtain a coating distribution feature array and a gap distribution feature array with a set control length.

[0044] Step S202: After merging the pre-trained coating retrieval features and gap retrieval features, a cohesive mechanism is used to extract the corresponding aggregation characteristics; the coating retrieval features and the gap retrieval features are obtained by training the corresponding adjustable retrieval features based on typical coating samples; the aggregation characteristics integrate the core elements of the coating retrieval features and the gap retrieval features.

[0045] Step S203: A dual-feature gating mechanism is adopted to extract the first gating fusion feature between the polymerization characteristic and each coating distribution feature in the coating distribution feature array, and to extract the second gating fusion feature between the polymerization characteristic and each gap distribution feature in the gap distribution feature array.

[0046] Step S204: Based on the acquired first gated fusion features and second gated fusion features, control the multi-segment intermittent coating of the target substrate.

[0047] In this embodiment of the invention, executively speaking, this embodiment uses a server as the execution subject and details the implementation process of an automated coating method for a multi-segment intermittent coating scenario of a substrate with a base-coated foil (aluminum foil for lithium-ion battery electrodes). The target substrate is an aluminum foil with a thickness of 12μm, pre-coated with a carbon black base coating (black, RGB values ​​30, 30, 30), and the coating material is lithium iron phosphate active material (dark gray, RGB values ​​100, 100, 100). The empty foil area is the natural color of the aluminum foil (silver white, RGB values ​​255, 255, 240). The three colors are significantly different, which facilitates data acquisition through a high-precision color mark sensor (resolution 0.01mm, sampling frequency 1kHz). The server preloads a multi-segment coating control model, which includes a data acquisition module, a feature storage unit (global coating storage unit and local gap storage unit), a feature dimensionality reduction processing module, a gated feature encoder (including the first / second gated unit and three sequentially connected gated feature processors GP1-GP3), and a coating parameter generation model. The control length is set to 5 (i.e., the final generated feature array length is 5).

[0048] The server controls the coating machine to transport the aluminum foil substrate at a speed of 0.5 m / s, and simultaneously activates the color mark sensor to scan the current coating segment in real time. The sensor identifies the boundary between the coating segment and the gap segment by color transitions: when the color changes from the base coat (30,30,30) to the coating layer (100,100,100), it is recorded as the start point of the coating segment; when it changes from the coating layer to the base coat, it is recorded as the end point of the coating segment. The distance between the two points is the "coating length characteristic" (coating distribution characteristic), for example, the length of the first coating segment is 150 mm, the second coating segment is 152 mm, the third coating segment is 149 mm, the fourth coating segment is 151 mm, and the fifth coating segment is 153 mm. For the gap distribution characteristics, the blank area (gap segment) following each coating segment is identified by the color change from the end of the coating to the start of the next coating segment. The spacing is the "blank length characteristic". For example, the first gap segment is 20mm, the second gap segment is 19.8mm, the third gap segment is 20.2mm, the fourth gap segment is 19.9mm, and the fifth gap segment is 20.1mm.

[0049] The server writes the coating distribution features into the global coating storage unit in real time, and writes the gap distribution features into the corresponding gap local storage units according to the gap sub-segments (e.g., the first gap sub-segment feature is written into storage unit 1, the second gap sub-segment into storage unit 2, etc.). When the number of features in the global coating storage unit reaches the set control length of 5, the server performs dimensionality reduction on the feature set based on feature similarity (Euclidean distance): the distance between the second coating segment (152mm) and the fourth coating segment (151mm) is calculated to be the minimum (1.0), the average of the two is aggregated (151.5mm), the original features are deleted and the aggregated result is written. At this time, the storage unit features are [150,151.5,149,153]; after writing the sixth coating segment feature 150.5mm, the storage unit restores 5 features and generates the coating distribution feature array [150,151.5,149,153,150.5]. The gap local storage unit is processed in the same way. For example, when the number of features in the first gap segment storage unit reaches 5, the first (20mm) and the fifth (20.1mm) features (average 20.05mm) are aggregated, and the sixth feature of 19.7mm is written to generate the gap distribution feature array [20.05, 19.8, 20.2, 19.9, 19.7].

[0050] The server loads coating retrieval features (512 dimensions, including core dimensions such as thickness deviation coefficient and uniformity index) and gap retrieval features (512 dimensions, including core dimensions such as width standard deviation and die-cutting adaptation coefficient) trained based on 1000 sets of typical coating samples, and merges them into a 1024-dimensional joint feature vector. The information entropy of each element in the vector is calculated through a cohesion mechanism, and the top 256 core elements with the highest entropy values ​​(such as "thickness deviation coefficient ≤ 0.5μm" for coating retrieval features and "width standard deviation ≤ 0.1mm" for gap retrieval features) are selected and fused to form a 256-dimensional aggregated characteristic. This characteristic simultaneously carries the key constraints of coating uniformity and gap accuracy. The feature mapping layer compresses the 640-dimensional joint feature to 128 dimensions using a 128×640-dimensional weight matrix. The cohesion mechanism calculates the information entropy H=-Σp(x)logp(x) of each element in the joint feature vector, and selects the 256 core elements with entropy values ​​below a threshold (such as H<0.1), retaining the key control dimensions of coating and gap features.

[0051] The server invokes a gated feature encoder to split the 256-dimensional aggregated features into 128-dimensional first aggregated features (coating-related) and 128-dimensional second aggregated features (gap-related). The first gating unit receives the first aggregated features and the coating distribution feature array, and calculates the fusion weight between each coating feature and the first aggregated feature using a dual-feature gating mechanism (G1=σ(W1·[P1;Ci]+b1), where σ is the sigmoid function), outputting five 128-dimensional first-gated fusion features. The second gating unit similarly outputs five 128-dimensional second-gated fusion features. These features are input into a sequentially connected gating feature processor: GP1 receives the original feature array and outputs the first-round fusion features; GP2 receives the output from GP1 and optimizes feature interaction information (such as the correlation between coating and gap features); GP3, as the last processor, outputs five 64-dimensional first-gated fusion features (G1final1~G1). final5 ) and 5 64-dimensional second-gated fusion features (G2) final1 ~G2 final5 ).

[0052] The server maps the fused features (5×64×2=640 dimensions) output by GP3 to 128-dimensional coating data features through a linear transformation. This data is then concatenated with the target substrate process description features ("lithium iron phosphate cathode coating, target thickness 80μm", 128 dimensions) and current control parameters (pressure 0.3MPa, flow rate 50ml / min, velocity 0.5m / s, 128 dimensions) to form a 256-dimensional input vector. This vector is then input into the coating parameter generation model (Transformer sequence generation model). The model uses a self-attention mechanism to capture feature associations and generate multiple control commands, such as those based on G1. final1 (For a 150mm coating section) Adjust the pressure to 0.32MPa to compensate for thickness; based on G2final1 (Corresponding to a 20.05mm gap) Adjust the speed to 0.49m / s to control the blanking deviation; optimize the pressure and flow parameters for subsequent coating sections. The server sends instructions to the coating machine PLC via the Profinet bus to achieve closed-loop control. The final coating thickness standard deviation is ≤0.8μm, and the gap deviation is ≤0.1mm, meeting the electrode production accuracy requirements.

[0053] In this embodiment of the invention, the multi-segment coating control model includes a gated feature encoder; the adoption of a dual-feature gating mechanism to extract the first gated fusion feature between the aggregation characteristic and each coating distribution feature in the coating distribution feature array, and to extract the second gated fusion feature between the aggregation characteristic and each gap distribution feature in the gap distribution feature array, can be implemented through the following example.

[0054] The polymerization characteristics are broken down into a first polymerization characteristic corresponding to the coating distribution characteristics and a second polymerization characteristic corresponding to the gap distribution characteristics;

[0055] The first gating unit in the gated feature encoder extracts the first aggregated feature and the first gating fusion feature between each coating distribution feature in the coating distribution feature array based on the dual feature gating mechanism.

[0056] The second gating unit in the gated feature encoder extracts the second aggregated feature and the second gating fusion feature between each gap distribution feature in the gap distribution feature array based on the dual feature gating mechanism.

[0057] In an embodiment of the present invention, exemplarily, in the multi-segment coating control model loaded by the server, the gated feature encoder includes a first gated unit (coating feature processing) and a second gated unit (gap feature processing), which operate in parallel to improve processing efficiency. The target substrate is aluminum foil with a base coating, which has been pre-processed to generate 256-dimensional polymerization characteristics (integrating coating and gap core control elements), a 5-dimensional coating distribution feature array ([150,151.5,149,153,150.5] mm, coating length feature), and a 5-dimensional gap distribution feature array ([20.05,19.8,20.2,19.9,19.7] mm, blank length feature). The server performs the following steps based on the above data:

[0058] The server uses a feature domain partitioning algorithm to split the 256-dimensional aggregation characteristics into 128-dimensional first aggregation characteristics (related to coating control) and 128-dimensional second aggregation characteristics (related to gap control). The first aggregation characteristics include coating thickness deviation coefficients (dimensions 1-32, characterizing the correlation between coating length and thickness, such as the thickness compensation amount corresponding to each 1mm increase in length), coating uniformity indices (dimensions 33-64, such as the thickness standard deviation threshold at different lengths), and slurry flowability adaptation parameters (dimensions 65-128, such as the matching coefficient between coating length and slurry flow rate). The second aggregation characteristics include gap width standard deviation (dimensions 1-32, controlling the fluctuation range of the blank length), die-cutting adaptation coefficients (dimensions 33-64, such as the die-cutting blade position parameters corresponding to a 20mm standard gap), and substrate tension correlation parameters (dimensions 65-128, the adjustment relationship between blank length and substrate tension). After splitting, the two types of aggregation characteristics are respectively input into the first and second gating units.

[0059] The first gating unit loads pre-trained gating parameters (weight matrix W1 is 128×129 dimensions, bias vector b1 is 128 dimensions), receives a 128-dimensional first aggregation feature (denoted as P1) and a 5-dimensional coating distribution feature array (denoted as C=[C1,C2,C3,C4,C5]=[150,151.5,149,153,150.5]mm), and performs dual-feature gating fusion on each coating distribution feature Ci:

[0060] Feature concatenation: P1 (128-dimensional) is concatenated with the current coating distribution feature Ci (1-dimensional, normalized to a value in the 0-1 range, e.g., C1=150mm is normalized to 0.32) to form a 129-dimensional vector [P1;Ci];

[0061] Weight calculation: The fusion weight Gi = σ(W1·[P1;Ci] + b1) is calculated using the sigmoid activation function, resulting in a 128-dimensional weight vector (element values ​​0-1, representing the fusion ratio of P1 and Ci).

[0062] Feature fusion: The first gated fusion feature Fi = Gi⊙P1 + (1-Gi)⊙(Ci mapping vector), where “⊙” means element-wise multiplication, and the Ci mapping vector is converted into a 128-dimensional feature by passing through a 1×128-dimensional linear layer after normalization.

[0063] For example, for C1 = 150 mm (normalized to 0.32):

[0064] The concatenated vector [P1;0.32] is calculated by W1·[P1;0.32]+b1 and then outputs Gi=[0.85,0.82,...,0.79] (128 dimensions, mean 0.81) via sigmoid.

[0065] The Ci mapping vector is [0.32, 0.32, ..., 0.32] (128 dimensions);

[0066] The fusion feature F1 = 0.85 × P11 + 0.15 × 0.32, ..., 0.79 × P1 128 +0.21×0.32, and we get the 128-dimensional first-gated fusion feature F1.

[0067] Repeat the above process, processing C2-C5 in sequence, and finally output five 128-dimensional first-gated fusion features {F1,F2,F3,F4,F5}.

[0068] The second gating unit is structurally symmetrical to the first gating unit. It loads gating parameters (weight matrix W2 is 128×129 dimensions, bias vector b2 is 128 dimensions), receives a 128-dimensional second aggregation feature (denoted as P2) and a 5-dimensional gap distribution feature array (denoted as G=[G1,G2,G3,G4,G5]=[20.05,19.8,20.2,19.9,19.7]mm), and performs dual-feature gating fusion on each gap distribution feature Gj.

[0069] Feature concatenation: P2 (128-dimensional) is concatenated with the current gap distribution feature Gj (1-dimensional, normalized, e.g., G1=20.05mm is normalized to 0.45) to form a 129-dimensional vector [P2;Gj];

[0070] Weight calculation: The fusion weight Hj=σ(W2·[P2;Gj]+b2) is calculated using the sigmoid function to obtain a 128-dimensional weight vector;

[0071] Feature fusion: The second gated fusion feature Hj = Hj⊙P2 + (1-Hj)⊙(Gj mapping vector), and the Gj mapping vector is converted into a 128-dimensional feature through a 1×128-dimensional linear layer.

[0072] For example, for G1 = 20.05 mm (normalized to 0.45):

[0073] After concatenating the vector [P2;0.45] and calculating it using W2・[P2;0.45]+b2, the sigmoid output is Hj=[0.78,0.80,...,0.83] (128 dimensions, mean 0.80).

[0074] The Gj mapping vector is [0.45, 0.45, ..., 0.45] (128 dimensions);

[0075] The fusion feature H1 = 0.78 × P21 + 0.22 × 0.45, ..., 0.83 × P2 128 +0.17×0.45, and we get the 128-dimensional second-gated fusion feature H1.

[0076] After processing, the second gating unit outputs five 128-dimensional second gating fusion features {H1,H2,H3,H4,H5}, which, together with {F1-F5} output by the first gating unit, serve as inputs to the subsequent gating feature processor for further optimization of coating control instructions.

[0077] In this embodiment of the invention, the gated feature encoder includes gated feature processors connected in sequence; for the first gated feature processor, the currently processed features include the coating distribution feature array and the gap distribution feature array; for the other gated feature processors, the currently processed features include each first gated fusion feature and each second gated fusion feature output by the previous gated feature processor.

[0078] The multi-segment intermittent coating control of the target substrate based on the acquired first gating fusion features and second gating fusion features can be implemented through the following example.

[0079] Based on the first gated fusion features and the second gated fusion features output by the last gated feature processor, the multi-segment intermittent coating control of the target substrate is controlled.

[0080] In this embodiment of the invention, for example, the gated feature encoder configured on the server includes three sequentially connected gated feature processors (GP1, GP2, GP3). Each processor has a built-in feature interaction module and a dimension optimization layer, which progressively enhance the gated fusion features in sequence, and finally achieve coating control based on the output of the last processor. The target substrate is aluminum foil with a base coating. After preprocessing, a 5-dimensional coating distribution feature array (C=[150,151.5,149,153,150.5]mm), a 5-dimensional gap distribution feature array (G=[20.05,19.8,20.2,19.9,19.7]mm) have been generated, as well as five 128-dimensional first gated fusion features (F1-F5) and five 128-dimensional second gated fusion features (H1-H5) that have been initially fused by the dual-feature gated mechanism.

[0081] The first gated feature processor (GP1): The current features being processed are the original coating distribution feature array C and the gap distribution feature array G. GP1 loads the feature interaction matrix (128×5-dimensional, pre-trained to capture the positional correlation between coating and gap features), aligns the 5 F1-F5 (coating-related) features with the 5 H1-H5 (gap-related) features according to their positions (the i-th F corresponds to the i-th H), and calculates the interaction weights of F and H at the same position (such as the correlation between F1 and H1, representing the synergistic adjustment relationship between the length of the first coating segment and the length of the first gap) through a multi-head attention mechanism (2 heads, 64 dimensions per head). Then, it outputs 5 first-gated fusion features of 96 dimensions (F1'=96-dimensional, corresponding to C1=150mm) and 5 second-gated fusion features of 96 dimensions (H1'=96-dimensional, corresponding to G1=20.05mm) through a dimensionality optimization layer (128→96-dimensional, removing redundant features).

[0082] The second gated feature processor (GP2) currently processes the F1'-F5' (96-dimensional) and H1'-H5' (96-dimensional) outputs from GP1. GP2 focuses on strengthening cross-location feature dependencies (such as the impact of the length correlation between the first and third coating segments on the overall coating uniformity). It uses a bidirectional long short-term memory network (Bi-LSTM, 96-dimensional hidden layers) to perform sequence modeling on F1'-F5' and H1'-H5' respectively, extracting temporal features (such as the cumulative effect of thickness compensation corresponding to the increasing length trend of the first three coating segments). Then, through a dimension optimization layer (96→64-dimensional), it outputs five 64-dimensional first-gated fusion features (F1''=64-dimensional) and five 64-dimensional second-gated fusion features (H1''=64-dimensional).

[0083] The third gated feature processor (GP3, the last one): currently processes the F1''-F5'' (64-dimensional) and H1''-H5'' (64-dimensional) outputs from GP2. GP3 focuses on core control features (such as the first 32 dimensions in F1'' directly related to thickness deviation, and the 20th-45th dimensions in H1'' related to gap die-cutting adaptation) through a self-attention mechanism (4 heads, 16 dimensions each), and fuses the original features output by GP2 through residual connections, ultimately outputting five 64-dimensional first gated fused features (F1... final =64-dimensional, including a thickness compensation coefficient of 0.02MPa / mm for a 150mm coating section) and 5 64-dimensional second-gated fusion features (H1) final =64 dimensions, including a speed adjustment coefficient of -0.01m / s / mm with a 20.05mm gap.

[0084] The server will output F1 from GP3. final -F5 final (5×64 dimensions) and H1 final -H5 final(5×64 dimensions) are merged into a 640-dimensional feature vector, which is then used as input for coating parameters to generate the model. The model is based on F1... final The thickness compensation coefficient (0.02MPa / mm) is used to calculate the pressure adjustment for a 150mm coating section: 150mm × 0.02MPa / mm = 3MPa. This is superimposed with a base pressure of 0.3MPa, generating a coating head pressure command of 3.3MPa. Based on H1... final The speed adjustment coefficient (-0.01 m / s / mm) is used to calculate the speed adjustment for a 20.05 mm gap (standard 20 mm): (20.05-20) mm × (-0.01 m / s / mm) = -0.0005 m / s. Adding the base speed of 0.5 m / s, the resulting substrate speed command is 0.4995 m / s ≈ 0.5 m / s (fine-tuned to 0.49 m / s to allow for margin). For F2... final -F5 final and H2 final -H5 final Similar calculations are performed sequentially to generate 5 sets of pressure / flow parameters for the coating section and 5 sets of speed / tension parameters for the gap section. These parameters are then sent to the coating machine PLC via the industrial bus to achieve precise closed-loop control of multi-segment intermittent coating.

[0085] In this embodiment of the invention, the multi-segment coating control model includes a coating parameter generation model; the multi-segment intermittent coating control of the target substrate based on the acquired first gating fusion features and second gating fusion features can be implemented through the following examples.

[0086] After merging the first gated fusion features and the second gated fusion features, they are converted into coating data features that are located in the same feature domain as the coating process parameters of the target substrate.

[0087] The coating process description features and coating control parameters of the target substrate are used as coating parameter guidance information, and the coating data features are combined with the coating parameter generation model to generate multi-segment intermittent coating control instructions for the target substrate.

[0088] In an embodiment of the invention, for example, the multi-segment coating control model configured on the server includes a coating parameter generation model (a sequence generation network based on the Transformer architecture) used to convert gating fusion features into actual coating control commands. The target substrate is aluminum foil with a base coating (used for the positive electrode sheet of a lithium-ion battery). The last processor (GP3) of the gating feature encoder outputs five 64-dimensional first gating fusion features (F1). final -F5 final (corresponding to 5 coating segments) and 5 64-dimensional second-gated fusion features (H1) final -H5 final(corresponding to 5 gap segments), the server performs the following steps based on these characteristics:

[0089] The server first sends 5 F1 final -F5 final (Total dimensions 5 × 64 = 320) and 5 H1s final -H5 final (Total dimensions 5 × 64 = 320) Segmented by position (F1) final With H1 final Corresponding to the first coating segment and the gap sub-segment, and so on, a 640-dimensional joint feature vector is formed. This vector is transformed into a 128-dimensional coating data feature through a pre-trained feature mapping layer (linear transformation matrix W = 128 × 640 dimensions, bias vector b = 128 dimensions). This feature is located in the same feature domain as the coating process parameters of the target substrate (coating head pressure, slurry flow rate, substrate velocity, etc.). For example, dimensions 1-32 of the coating data feature correspond to "coating segment thickness compensation" (such as F1). final The extracted 150mm coating section requires a pressure compensation coefficient of 0.02MPa / mm. Dimensions 33-64 correspond to "gap section speed adjustment" (e.g., H1). final The 20.05mm gap requires a speed adjustment coefficient of -0.01m / s / mm. Dimensions 65-128 correspond to "cross-segment coordination parameters" (such as the flow accumulation adjustment factor corresponding to the increasing trend of the first 3 coating segment lengths).

[0090] The server merges the coating process description features of the target substrate and the current coating control parameters into coating parameter guidance information. The process description features are "lithium iron phosphate cathode coating, target thickness 80μm±0.5μm, substrate width 300mm, primer thickness 2μm, standard gap 20mm±0.2mm", which is processed into a 128-dimensional vector (including the value 0.65 corresponding to the target thickness of 80μm, the value 0.42 corresponding to the standard gap of 20mm, etc.) after text embedding (Word2Vec) and normalization. The current coating control parameters are real-time collected equipment operation data: coating head pressure 0.3MPa (normalized 0.3), slurry flow rate 50ml / min (normalized 0.5), substrate transfer speed 0.5m / s (normalized 0.5), substrate tension 30N (normalized 0.3), which is also encoded into a 128-dimensional vector. The two are concatenated into 256-dimensional coating parameter guidance information, which serves as the "control target anchor point" of the model.

[0091] The server concatenates the 128-dimensional coating data features with the 256-dimensional coating parameter guidance information into a 384-dimensional input vector, which is then input into the coating parameter generation model. The model captures feature correlations through a self-attention mechanism (8 heads, 48 ​​dimensions per head): for example, the matching relationship between "0.02MPa / mm pressure compensation coefficient" in the coating data features and "target thickness 80μm" in the guidance information, calculates the pressure adjustment amount of the first coating segment (150mm) = 150mm × 0.02MPa / mm = 3MPa (base pressure 0.3MPa + adjustment amount 0.02MPa = 0.32MPa); through the correlation between "-0.01m / s / mm speed adjustment coefficient" and "standard gap 20mm", calculates the speed adjustment amount of the first gap sub-segment (20.05mm) = (20.05-20)mm × (-0.01m / s / mm) = -0.0005m / s (current speed 0.5m / s + adjustment amount ≈ 0.49m / s). The model sequentially calculates parameters for five coating sections and five gap sub-sections, generating control commands: for the first coating section, "pressure 0.32MPa, flow rate 50.5ml / min"; for the first gap sub-section, "velocity 0.49m / s, tension 30.5N". Subsequent section parameters are optimized according to this logic. The server sends the commands to the coating machine PLC via the Profinet bus, achieving precise control. The final coating thickness deviation is ≤0.4μm, and the gap length deviation is ≤0.08mm, meeting the requirements for electrode production.

[0092] In this embodiment of the invention, each gap distribution feature corresponds to a gap sub-segment in the coating segment; the dimensionality reduction processing of the obtained coating distribution feature set and gap distribution feature set based on feature similarity to obtain a coating distribution feature array and a gap distribution feature array with a set control length can be implemented through the following example.

[0093] For each coating segment collected, the coating distribution characteristics of the coating segment are written into the global coating storage unit, and the gap distribution characteristics of the coating segment are written into the gap local storage unit.

[0094] When the number of coating distribution features stored in the global coating storage unit reaches the set control length, based on feature similarity, feature aggregation is performed on the coating distribution features of at least two coating segments in the global coating storage unit, and the coating distribution features to be written are written into the global coating storage unit.

[0095] Furthermore, for each gap segment, when the number of corresponding gap distribution features stored in the gap local storage unit reaches the set control length, based on feature similarity, feature aggregation is performed on the corresponding gap distribution features of at least two coating segments in the gap local storage unit, and the gap distribution features to be written are written into the gap local storage unit.

[0096] In an embodiment of the invention, for example, the server performs dynamic feature management for a multi-segment intermittent coating scenario on an aluminum foil substrate with a base coating (targeting lithium-ion battery electrodes). This is achieved through a global coating storage unit (storing the coating distribution characteristics of all coating segments) and a local gap storage unit (independently storing the corresponding gap distribution characteristics for each gap sub-segment). The feature set is dimensionality-reduced based on feature similarity to ensure that the number of stored features remains at a set control length of 5. The coating segments of the target substrate are continuously generated at a speed of 0.5 m / s, with each coating segment corresponding to a gap sub-segment. The server collects the "coating length characteristics" (coating distribution characteristics) of the coating segments and the "blank length characteristics" (gap distribution characteristics) of the gap sub-segments using a high-precision color mark sensor. The specific processing flow is as follows:

[0097] Real-time writing process: Each time the server collects a coating segment, it immediately writes its coating distribution characteristics (coating length) into the coating's global storage unit (initially empty). For example:

[0098] The first coating section has a coating length of 150mm, and the post-writing memory cell feature set S is... coat =

[150] (Quantity = 1 < 5, dimensionality reduction is not triggered);

[0099] The second coating section is 152mm long; after writing, S... coat =[150,152] (Quantity = 2 < 5);

[0100] The third coating section is 149mm long; after writing, S... coat =[150,152,149] (Quantity = 3 < 5);

[0101] The fourth coating section is 151mm long; after writing, S... coat =[150,152,149,151] (Quantity = 4 < 5);

[0102] The fifth coating section is 153mm long; after writing, S... coat =[150,152,149,151,153] (Quantity = 5 = Sets the control length, triggers dimensionality reduction).

[0103] Feature similarity calculation and aggregation: The server calculates S using Euclidean distance. coat Similarity of all feature pairs (the smaller the distance, the higher the similarity):

[0104] Distance between 150 and 152 = =2; the distance between 152 and 149 is 3; the distance between 149 and 151 is 2; the distance between 151 and 153 is 2; the distance between 150 and 149 is 1 (minimum). Therefore, 150mm (first coating segment) and 149mm (third coating segment) are determined to be the two features with the highest similarity.

[0105] Perform mean aggregation on the two: (150+149) / 2=149.5mm, delete the original 150 and 149, and update the storage unit to S. coat =[149.5,152,151,153] (Quantity = 4 < 5).

[0106] New feature written: The coating length of the 6th coating segment is 150.5 mm. After writing, S coat =[149.5,152,151,153,150.5] (Quantity = 5 = Set adjustment length), complete the dimensionality reduction of the coating distribution feature array, the array is [149.5,152,151,153,150.5]. For each subsequent new coating segment, repeat the "write - full 5 - aggregate - write new feature" process, always maintaining 5 features.

[0107] Each coating segment is followed by a gap segment (e.g., the first coating segment corresponds to the first gap segment, the second coating segment corresponds to the second gap segment, and so on). The server configures an independent gap local storage unit (denoted as LS1, LS2, ..., LS) for each gap segment. n (where n is the gap segment number), only the blank length feature of the corresponding segment is stored, and dimensionality reduction is performed independently. Taking the first gap segment (LS1) as an example:

[0108] Real-time writing process: After the server completes the coating of the m-th coating segment, it immediately writes the blank length feature of the corresponding gap segment (m-th gap segment) into LS. m For LS1 (which stores all the blank lengths of the first gap segment):

[0109] After the first coating, leave a blank length of 20mm in the first gap segment and write LS1=

[20] (quantity=1<5);

[0110] After the second coating, the blank length of the first gap segment (which appears repeatedly due to substrate cyclic coating) is 19.8 mm, and is written in LS1=[20,19.8] (quantity=2<5).

[0111] After the third coating, the first gap segment is 19.9mm, and LS1=[20,19.8,19.9] (quantity=3<5) is written.

[0112] After the fourth coating, the first gap segment is 20.1mm, and LS1=[20,19.8,19.9,20.1] (quantity=4<5) is written.

[0113] After the 5th coating, the first gap segment is 20.2mm, and LS1=[20,19.8,19.9,20.1,20.2] is written (quantity=5=set adjustment length, triggering dimensionality reduction).

[0114] Feature similarity calculation and aggregation: The server calculates the Euclidean distance between features in LS1:

[0115] The distance between 20 and 19.8 is 0.2, the distance between 19.8 and 19.9 is 0.1, the distance between 19.9 and 20.1 is 0.2, the distance between 20.1 and 20.2 is 0.1, and the distance between 20 and 20.1 is 0.1 (minimum). Therefore, 20mm (first time) and 20.1mm (fourth time) are determined to be the two features with the highest similarity.

[0116] Perform mean aggregation on the two: (20+20.1) / 2=20.05mm, delete the original 20 and 20.1, and update LS1 to [20.05,19.8,19.9,20.2] (quantity=4<5).

[0117] Write new features: After the 6th coating, the blank length of the first gap segment is 19.7mm. Write LS1=[20.05,19.8,19.9,20.2,19.7] (quantity=5=set adjustment length) to complete the dimensionality reduction of the gap distribution feature array of the first gap segment, and the array is [20.05,19.8,19.9,20.2,19.7]. Other gap segments (LS2,LS3…) are processed independently according to the same logic to ensure that the length of the gap distribution feature array of each segment is always 5.

[0118] Through the above processing, the server dynamically maintains the length of the coating distribution feature array and the gap distribution feature array of each gap segment at the set adjustable length of 5. This not only preserves the key feature information but also avoids the expansion of the number of features, which would lead to a decrease in the efficiency of subsequent gating fusion. This lays the data foundation for accurately generating coating control instructions.

[0119] In this embodiment of the invention, before employing a dual-feature gating mechanism to extract the first gating fusion feature between the polymerization characteristic and each coating distribution feature in the coating distribution feature array, and the second gating fusion feature between the polymerization characteristic and each gap distribution feature in the gap distribution feature array, the following implementation method is also provided.

[0120] Once the set control node is determined to be reached, the coating distribution feature array stored in the global coating storage unit and the gap distribution feature array stored in the gap local storage unit are read.

[0121] In this embodiment of the invention, for example, before executing the dual-feature gating mechanism, the server needs to trigger the reading of the feature array through a preset control node to ensure that the feature data input to the gating unit meets the set control length and is stable and effective. The target substrate is an aluminum foil with a base coating, the set control length is 5, and the triggering condition of the control node is dynamically determined based on the feature state of the coating global storage unit and the gap local storage unit. The specific process is as follows:

[0122] The server's configuration for determining the control node is as follows: "When the global storage unit of the coating completes the k-th feature aggregation and writes a new feature, the number of stored coating distribution features stabilizes at the set control length of 5, and all enabled gap local storage units (corresponding to the gap segments that have appeared) have completed at least one feature aggregation and stably maintained 5 features, the control node is triggered." When a gap segment appears ≥3 times, the server determines it as a 'persistently existing segment' and enables the corresponding gap local storage unit, ensuring that the node triggering conditions are quantifiable.

[0123] Taking the target substrate coating process as an example: After the server completes the coating of the 6th coating segment, the global storage unit of the coating is "written with the feature of the 6th coating segment (150.5mm) → the number of features in the storage unit is restored to 5 ([149.5,152,151,153,150.5])", and the local storage units (LS1 to LS5) of the first 5 gap segments (gap segments 1 to 5) have all completed the first feature aggregation (e.g., LS1 stores [20.05,19.8,19.9,20.2,19.7], LS2 stores [19.9,20.1,19.7,20.0,19.8], etc., all of which are 5 features). At this time, the server confirms through the status detection module that: the number of features in the global storage unit of the coating = 5, the number of features in all gap local storage units = 5, and there are no temporary features that have not been reduced in dimensionality (e.g., redundant features to be aggregated), and determines that the set control node has been reached.

[0124] After the server triggers the control node, it synchronously reads the feature arrays of the coating's global storage unit and each gap's local storage unit through the feature reading interface:

[0125] Read the coating distribution feature array: The coating global storage unit outputs the currently stored 5 coating distribution features to the input buffer of the gated feature encoder through the high-speed cache interface (transmission rate 1GB / s). The array is [149.5,152,151,153,150.5] mm (including the aggregated 1 / 3 coating segment feature 149.5 mm, the original 2nd coating segment 152 mm, etc.).

[0126] Read the gap distribution feature array: Each gap local storage unit (LS1 to LS5) outputs the feature array sequentially according to the segment number. For example, LS1 outputs [20.05, 19.8, 19.9, 20.2, 19.7] mm for the first gap segment, and LS2 outputs [19.9, 20.1, 19.7, 20.0, 19.8] mm for the second gap segment. All arrays are verified (feature dimension 5×1, value range within 19.5-20.5 mm) and then written to the buffer.

[0127] After the reading is complete, the server sends a "feature ready" signal to the gated feature encoder, triggering the dual-feature gating mechanism to ensure that subsequent feature fusion is performed based on the latest and stable feature array, providing a data foundation for the accurate generation of coating control instructions.

[0128] In this embodiment of the invention, feature aggregation is performed on the features to be aggregated in at least two coating segments using any of the following methods:

[0129] Based on feature similarity, the average value of the features to be aggregated in the continuous coating segments with the closest parameters in the storage unit is calculated.

[0130] Based on feature similarity, the features to be written and aggregated are grouped and classified with the features to be written and aggregated in the storage unit, and the average value of the features to be aggregated in each grouped and classified coating segment is calculated.

[0131] Wherein, the feature to be aggregated is a coating distribution feature or a gap distribution feature, and correspondingly, the storage unit is a coating global storage unit or a gap local storage unit.

[0132] In this embodiment of the invention, for example, when the server aggregates the features to be aggregated (coating distribution features or gap distribution features) in the global storage unit of the coating or the local storage unit of the gap, two typical methods are used based on feature similarity to ensure that the aggregated features retain the core control information and that the dimension conforms to the set control length. The following is an explanation in conjunction with a specific scenario:

[0133] Application Scenario: The global storage unit for the coating stores the coating distribution characteristics (coating length) of consecutive coating segments. It is necessary to aggregate the consecutive features with the closest parameters. For example, after the 5th coating, the global storage unit stores 5 consecutive coating segment features: [150, 152, 151, 153, 154] (unit: mm, all are coating lengths of consecutive segments). The server calculates the Euclidean distance (parameter proximity) between all consecutive feature pairs: 150 and 152 distance = 2, 152 and 151 distance = 1 (minimum), 151 and 153 distance = 2, 153 and 154 distance = 1. The consecutive coating segments with the closest parameters are the 2nd and 3rd coating segments (152mm and 151mm). The server performs mean aggregation on the two: (152+151) / 2=151.5mm, deletes the original 152 and 151, updates the storage unit to [150,151.5,153,154], and then writes the new coating segment feature 150.5mm, restoring the 5 feature arrays [150,151.5,153,154,150.5].

[0134] Application Scenario: Gap local storage units (such as LS1 in the first gap segment) store the gap distribution features (blank length) to be written, which need to be grouped and aggregated with the already written features. For example, LS1 has stored 4 features: [19.8, 20.0, 20.2, 19.9] (unit: mm, all are blank lengths of the first gap segment), and a new feature of 20.1 mm needs to be written. The server presets 3 feature grouping intervals: [19.7-19.9] (low interval), [19.9-20.1] (middle interval), and [20.1-20.3] (high interval). Among the features already written, 19.8 and 19.9 belong to the low interval (mean 19.85), 20.0 belongs to the middle interval, and 20.2 belongs to the high interval. The 20.1mm to be written belongs to the middle interval and is grouped with the already written 20.0mm. The mean aggregation is performed on this group: (20.0+20.1) / 2=20.05mm, replacing the original 20.0mm, and the storage unit is updated to [19.85,20.05,20.2]. Then, the new feature 19.7mm (low interval) is written, and the low interval features become [19.85,19.7], with a mean of (19.85+19.7) / 2=19.775mm. The final LS1 feature array is [19.775,20.05,20.2]. After supplementing the new features, 5 features are maintained, and the grouping, classification, and aggregation are completed.

[0135] In this embodiment of the invention, if there are multiple sets of consecutive coating segments with the closest parameters, the step of calculating the average value of the features to be aggregated of the consecutive coating segments with the closest parameters in the storage unit can be implemented through the following example.

[0136] The average value of the features to be aggregated in the first written and closest consecutive coating segments in the storage unit is calculated; or

[0137] The average value of the features to be aggregated for each group of consecutive coated segments in the storage unit with the closest parameters is calculated.

[0138] In an embodiment of the present invention, for example, when the server has multiple sets of consecutive coating segments with the closest parameters to be aggregated in the storage unit, it performs mean calculation through two strategies to ensure that the aggregated features retain temporal correlation and meet the set control length. The following description is based on the scenario of the global storage unit of the coating layer:

[0139] The global coating storage unit stores the coating distribution characteristics (coating length) of 5 consecutive coating segments: [150, 151, 152, 151, 150] mm (arranged in writing order, with parameters representing the 1st to 5th coating segments respectively). The server calculates the Euclidean distance (parameter closeness) of all consecutive feature pairs: (150, 151) = 1, (151, 152) = 1, (152, 151) = 1, (151, 150) = 1, with 4 sets of parameters being the closest (each with a distance of 1). At this point, following the "first write" principle, the first group of continuous coating segments (the first and second coating segments, 150mm and 151mm) is selected and mean aggregation is performed: (150+151) / 2=150.5mm. The original 150 and 151 are deleted, and the storage unit is updated to [150.5,152,151,150]. Then, the new feature 153mm is written, restoring the 5 feature arrays [150.5,152,151,150,153].

[0140] The same storage unit has a feature array [150, 151, 152, 151, 150] mm, with four consecutive feature pairs having a distance of 1 (the closest parameters). The server calculates the mean independently for each group of consecutive coated segments according to the "aggregate separately" principle: the mean for group 1 (150, 151) is 150.5 mm, the mean for group 2 (151, 152) is 151.5 mm, the mean for group 3 (152, 151) is 151.5 mm, and the mean for group 4 (151, 150) is 150.5 mm. The original five features are deleted, leaving the four aggregated features [150.5, 151.5, 151.5, 150.5]. After writing the new feature 153 mm, the storage unit feature array becomes [150.5, 151.5, 151.5, 150.5, 153] (length 5), completing the aggregation of multiple groups.

[0141] In this embodiment of the invention, the target substrate is a foil substrate with a primer coating; the step of collecting data on each coated segment of the target substrate to obtain corresponding coating distribution characteristics and gap distribution characteristics includes:

[0142] A high-precision color mark sensor is used to detect three different color areas on the foil substrate: the empty foil area, the base coating area, and the coating area, to collect the base coating data and blanking data of the current coating section.

[0143] Based on the base coating data and the blanking data, the coating length characteristics of the coating segment and the blanking length characteristics of the gap segment are obtained, and the coating length characteristics are used as the coating distribution characteristics, and the blanking length characteristics are used as the gap distribution characteristics.

[0144] In this embodiment of the invention, for example, the server, targeting a foil substrate with a base coating (such as aluminum foil for lithium-ion battery electrodes, 12μm thick, with a carbon black conductive base coating), uses a high-precision color mark sensor to detect the boundaries of three color regions: empty foil, base coating, and coating layer. Key data is collected, and coating distribution characteristics and gap distribution characteristics are extracted. The specific process is as follows: The server controls the coating machine to transport the foil substrate at a speed of 0.5m / s, simultaneously activating the high-precision color mark sensor (resolution 0.01mm, sampling frequency 1kHz, spectral response range 400-700nm) deployed at the detection station. This sensor performs a line scan of the substrate surface, distinguishing three regions through RGB color values: the empty foil region (natural aluminum foil color, silver-white) outputs RGB values ​​(255, 255, 240); the base coating region (carbon black layer, black) outputs (30, 30, 30); and the coating region (active material, dark gray) outputs (100, 100, 100). The sensor outputs the RGB value of one sampling point every 1ms, which the server receives and stores in real time through a data acquisition module. For example, when the substrate passes through the sensor detection area, the sampling sequence appears sequentially: (30,30,30) (base coat) → (100,100,100) (coating start point) → (100,100,100) (coating area) → (30,30,30) (coating end point / gap start point) → (30,30,30) (gap area) → (100,100,100) (next coating start point). The server identifies the boundaries of "base coat-coating" and "coating-base coat" based on color transition points, corresponding to the start and end points of the coating segment, respectively; the boundaries of "coating-base coat" and "base coat-coating" correspond to the start and end points of the gap segment, and accordingly collects the base coat data (length of the base coat area before the coating start point) and blanking data (length of the base coat area in the gap segment) of the current coating segment. The server calculates the length of the collected primer and blank data, generating coating distribution characteristics and gap distribution characteristics: Coating length characteristics (coating distribution characteristics): The starting point of the coating segment is the color transition point from "primer to coating" (RGB changes from (30,30,30) to (100,100,100)), and the ending point is the transition point from "coating to primer" (RGB changes from (100,100,100) to (30,30,30)). The server calculates the distance between the two points based on the sampling frequency and substrate velocity: For example, the starting sampling time t1 = 100ms, the ending sampling time t2 = 400ms, the sampling interval Δt = 300ms, the substrate velocity v = 0.5m / s = 500mm / s, and the coating length L = v × Δt = 500mm / s × 0.3s = 150mm. This value is the coating length characteristic (coating distribution characteristic) of the current coating segment.Blank space length characteristic (gap distribution characteristic): The starting point of the gap segment is the end point of the coating segment (t2=400ms), and the end point is the starting point of the next coating segment (t3=440ms). The sampling interval Δt=40ms, and the blank space length L=500mm / s×0.04s=20mm, is the blank space length characteristic (gap distribution characteristic) of the current gap segment. For example, the "base coat → coating layer" transition point of the second coating segment is at t=500ms, and the "coating layer → base coat" transition point is at t=804ms, Δt=304ms, and the coating length=500×0.304=152mm; the corresponding gap segment start point t=804ms, end point t=843.6ms, Δt=39.6ms, and blank space length=500×0.0396=19.8mm. The server stores 152mm as the coating distribution characteristic of the second coating segment and 19.8mm as the gap distribution characteristic of the corresponding gap sub-segment, and stores them sequentially in the corresponding storage units.

[0145] This invention provides a computer device 100, which includes a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device 100 performs the aforementioned automated coating method. Figure 2 As shown, Figure 2 This is a structural block diagram of a computer device 100 provided in an embodiment of the present invention. The computer device 100 includes a memory 111, a processor 112, and a communication unit 113. To enable data transmission or interaction, the memory 111, processor 112, and communication unit 113 are electrically connected to each other directly or indirectly. For example, these components can be electrically connected to each other through one or more communication buses or signal lines.

[0146] For illustrative purposes, the foregoing description has been made with reference to specific embodiments. However, the foregoing illustrative discussions are not intended to be exhaustive or to limit the present disclosure to the precise forms disclosed. Numerous modifications and variations are possible in accordance with the foregoing teachings. These embodiments were chosen and described in order to best illustrate the principles of the present disclosure and its practical application, thereby enabling those skilled in the art to best utilize the disclosure and to employ various embodiments with different modifications to suit a particular intended application.

Claims

1. An automated coating method, characterized in that, include: Based on the multi-segment coating control model, data is collected from each coating segment in the target substrate to obtain the corresponding coating distribution features and gap distribution features. Based on feature similarity, the obtained coating distribution feature set and gap distribution feature set are dimensionality reduced to obtain coating distribution feature array and gap distribution feature array with set control length. After merging the pre-trained coating retrieval features and gap retrieval features, a cohesive mechanism is used to extract the corresponding aggregation characteristics. The coating retrieval features and the gap retrieval features are obtained by training the corresponding adjustable retrieval features based on typical coating samples. The aggregation characteristics integrate the core elements of the coating retrieval features and the gap retrieval features. A dual-feature gating mechanism is adopted to extract the first gating fusion feature between the polymerization characteristic and each coating distribution feature in the coating distribution feature array, and to extract the second gating fusion feature between the polymerization characteristic and each gap distribution feature in the gap distribution feature array. Based on the acquired first-gated fusion features and second-gated fusion features, the multi-segment intermittent coating control of the target substrate is controlled.

2. The method according to claim 1, characterized in that, The multi-segment coating control model includes a gated feature encoder; the adoption of a dual-feature gating mechanism, extracting the first gated fusion feature between the aggregation characteristic and each coating distribution feature in the coating distribution feature array, and extracting the second gated fusion feature between the aggregation characteristic and each gap distribution feature in the gap distribution feature array, includes: The polymerization characteristics are broken down into a first polymerization characteristic corresponding to the coating distribution characteristics and a second polymerization characteristic corresponding to the gap distribution characteristics; The first gating unit in the gated feature encoder extracts the first aggregated feature and the first gating fusion feature between each coating distribution feature in the coating distribution feature array based on the dual feature gating mechanism. The second gating unit in the gated feature encoder extracts the second aggregated feature and the second gating fusion feature between each gap distribution feature in the gap distribution feature array based on the dual feature gating mechanism.

3. The method according to claim 2, characterized in that, The gated feature encoder includes gated feature processors connected in sequence; for the first gated feature processor, the currently processed features include the coating distribution feature array and the gap distribution feature array; for other gated feature processors, the currently processed features include each first gated fusion feature and each second gated fusion feature output by the previous gated feature processor. The control of multi-segment intermittent coating of the target substrate based on the acquired first-gated fusion features and second-gated fusion features includes: Based on the first gated fusion features and the second gated fusion features output by the last gated feature processor, the multi-segment intermittent coating control of the target substrate is controlled.

4. The method according to claim 1, characterized in that, The multi-segment coating control model includes a coating parameter generation model; the multi-segment intermittent coating control of the target substrate based on the acquired first-gated fusion features and second-gated fusion features includes: After merging the first gated fusion features and the second gated fusion features, they are converted into coating data features that are located in the same feature domain as the coating process parameters of the target substrate. The coating process description features and coating control parameters of the target substrate are used as coating parameter guidance information, and the coating data features are combined with the coating parameter generation model to generate multi-segment intermittent coating control instructions for the target substrate.

5. The method according to any one of claims 1 to 4, characterized in that, Each gap distribution feature corresponds to a gap sub-segment in the coating segment; based on feature similarity, the obtained coating distribution feature set and gap distribution feature set are respectively subjected to dimensionality reduction processing to obtain a coating distribution feature array and a gap distribution feature array with a set adjustable length, including: For each coating segment collected, the coating distribution characteristics of the coating segment are written into the global coating storage unit, and the gap distribution characteristics of the coating segment are written into the gap local storage unit. When the number of coating distribution features stored in the global coating storage unit reaches the set control length, based on feature similarity, feature aggregation is performed on the coating distribution features of at least two coating segments in the global coating storage unit, and the coating distribution features to be written are written into the global coating storage unit. Furthermore, for each gap segment, when the number of corresponding gap distribution features stored in the gap local storage unit reaches the set control length, based on feature similarity, feature aggregation is performed on the corresponding gap distribution features of at least two coating segments in the gap local storage unit, and the gap distribution features to be written are written into the gap local storage unit.

6. The method according to claim 5, characterized in that, Before employing a dual-feature gating mechanism to extract the first gated fusion feature between the aggregation characteristic and each coating distribution feature in the coating distribution feature array, and before extracting the second gated fusion feature between the aggregation characteristic and each gap distribution feature in the gap distribution feature array, the method further includes: Once the set control node is determined to be reached, the coating distribution feature array stored in the global coating storage unit and the gap distribution feature array stored in the gap local storage unit are read.

7. The method according to claim 5, characterized in that, Feature aggregation is performed on the features to be aggregated from at least two coating segments using any of the following methods: Based on feature similarity, the average value of the features to be aggregated in the continuous coating segments with the closest parameters in the storage unit is calculated. Based on feature similarity, the features to be written and aggregated are grouped and classified with the features to be written and aggregated in the storage unit, and the average value of the features to be aggregated in each grouped and classified coating segment is calculated. Wherein, the feature to be aggregated is a coating distribution feature or a gap distribution feature, and correspondingly, the storage unit is a coating global storage unit or a gap local storage unit.

8. The method according to claim 7, characterized in that, If multiple sets of consecutive coated segments with the closest parameters exist, the step of averaging the features to be aggregated in the storage unit for the consecutive coated segments with the closest parameters includes: The average value of the features to be aggregated in the first written and closest consecutive coating segments in the storage unit is calculated; or The average value of the features to be aggregated for each group of consecutive coated segments in the storage unit with the closest parameters is calculated.

9. The method according to claim 1, characterized in that, The target substrate is a foil substrate with a primer coating; the data acquisition of each coated segment in the target substrate to obtain the corresponding coating distribution characteristics and gap distribution characteristics includes: A high-precision color mark sensor is used to detect three different color areas on the foil substrate: the empty foil area, the base coating area, and the coating area, to collect the base coating data and blanking data of the current coating section. Based on the base coating data and the blanking data, the coating length characteristics of the coating segment and the blanking length characteristics of the gap segment are obtained, and the coating length characteristics are used as the coating distribution characteristics, and the blanking length characteristics are used as the gap distribution characteristics.

10. A server system, characterized in that, Includes a server, the server being used to perform the method according to any one of claims 1-9.

Images