Batch colorization assistant based on artificial intelligence, ai
By using an AI-based liquid coloring system, the colorant in liquid batches can be automatically measured and predicted, solving the consistency and efficiency problems in the traditional liquid coloring process and achieving efficient and accurate color matching and batch management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SCHNEIDER ELECTRIC SYSTEMS USA INC
- Filing Date
- 2025-05-09
- Publication Date
- 2026-06-09
AI Technical Summary
Traditional liquid coloring processes rely on expert knowledge, resulting in inconsistent coloring results and low efficiency. Conventional measuring tools have errors, affecting batch-to-batch consistency and production efficiency.
An AI-based liquid coloring system is employed, which uses a color measurement unit, a color digitization processor, and an AI coloring prediction engine to achieve automated measurement of liquid batch samples and accurate prediction of colorants. The system combines historical information and target color data to create a model and optimize the coloring process.
It improves the efficiency and consistency of the coloring process, reduces the time required to make physical test panels, lowers production costs, and provides real-time factory management support, ensuring color consistency between batches.
Smart Images
Figure CN122176148A_ABST
Abstract
Description
[0001] Cross-reference of related applications
[0002] This application claims the benefit of Indian Patent Application No. 202411097083, filed on December 9, 2024, the entire disclosure of which is incorporated herein by reference. Background Technology
[0003] Industrial tinting is the process of adjusting the color of a liquid product to match a specific objective for its intended use. Conventional tinting processes are lengthy, requiring multiple iterations of testing samples and adding colorant to achieve the desired color. In traditional paint tinting, a sample from a liquid batch is applied to a test panel, which is then cured to reveal the actual color. This test panel is compared to a reference panel of the target color to evaluate the match. Tinting experts from quality control evaluate the colors of the sample and the target panel to determine the appropriate colorant and its quantity to be added to the liquid batch to achieve the target color within a specified tolerance. Therefore, this process often results in significant time consumption and waste, as a new test panel must be created for each evaluation.
[0004] The coloring process relies heavily on the expertise of coloring specialists to achieve the desired results. A thorough understanding of the product color, colorants, and their respective strengths is crucial for successful coloring. However, plant managers and operators often lack insight into the coloring process, including the duration of operations and the number of iterations required. This reliance on individual expertise can lead to variability in coloring results, with outcomes differing between experts. Therefore, inconsistencies in the coloring process can affect batch-to-batch consistency, making it challenging for plant managers and operators to effectively plan batch scheduling.
[0005] Current conventional tools for measuring liquid color have limitations that hinder accurate color measurement. While spectrophotometers can determine color measurements, measuring liquids such as paint in existing systems introduces several potential sources of error. For example, the meniscus of a liquid can affect measurement accuracy, and newly manufactured high-viscosity liquids (such as paint) often contain air bubbles, which further impairs the accuracy of readings. Summary of the Invention
[0006] This disclosure provides an artificial intelligence (AI)-based liquid coloring solution that integrates systems for accurate measurement, historical coloring information, and colorant dosage estimation to provide an efficient coloring process within industrial mixing systems.
[0007] In one aspect, a method for coloring a liquid in an industrial mixing system includes collecting a liquid batch sample from a batch of liquid in the industrial mixing system and measuring the liquid batch sample using a sensor to generate a batch color measurement associated with the liquid batch sample. The method also includes transmitting the batch color measurement to a color digitization processor, whereby the color digitization processor executes an AI color prediction engine. Executing the AI color prediction engine includes storing the batch color measurement in a color digitization database, wherein the database includes color information of multiple historical batch samples, color measurements of multiple target standard samples, and colorant information, wherein the colorant information includes colorant values. Executing the AI color prediction engine also includes defining an incremental error between the batch color measurement and the target standard sample color measurement. Executing the AI color prediction engine also includes modeling the coloring of the liquid batch by weighting the incremental error, the batch color measurement, the batch volume, the color information of multiple historical batch samples, and the colorant information to generate at least one predicted colorant and the mass of the colorant based on the multiple colorant information, to achieve the color measurement of the target standard samples. The method also includes receiving a predicted colorant and the mass of the colorant by an industrial mixing system, and mixing one or more colorants by the industrial mixing system based on the predicted colorant and the mass of the colorant.
[0008] In another aspect, a liquid coloring system includes: a mixing system for mixing a liquid; a color measurement unit for receiving and measuring a liquid sample collected from the mixing system to generate liquid color information; a color digitization processor coupled to the color measurement unit; a color digitization database storing product information and colorant information, and coupled to the color digitization processor; and a memory storing computer-executable instructions. When executed by the color digitization processor, the instructions configure the coloring system to: receive liquid color information from the color measurement unit, identify a product identifier based on the liquid color information and product information, and execute an artificial intelligence (AI) color prediction engine to predict colorants. Executing the AI color prediction engine includes defining an incremental error based on the liquid color information and the product identifier. Executing the AI color prediction engine also includes modeling the coloring of the liquid sample by predicting one or more colorants and colorant masses based on the incremental error, liquid color information, product identifier, and colorant information through one or more models. The liquid coloring system also includes transferring the colorant and colorant mass to the mixing system.
[0009] In another aspect, a system for measuring the color of a liquid includes a test container. The test container includes a base, an internal mounting element, and an internal compartment, wherein the internal mounting element is coupled to the base. The system for measuring the liquid also includes a test holder, wherein the test holder is mounted within the internal compartment. The test holder includes a transparent inner surface configured to receive a test sample. The system for measuring the color of the liquid also includes a spectrophotometer coupled to the internal mounting element below the transparent inner surface of the test holder, wherein the spectrophotometer is oriented such that its sensor end points toward the transparent inner surface.
[0010] Other objects and features of the invention will be apparent in part and are set forth in part herein. Attached Figure Description
[0011] Figure 1 An artificial intelligence (AI)-based coloring assistant system according to an embodiment is shown.
[0012] Figure 2A The sample display generated by the insight processor shows the color iterations required for product clustering, where clusters are created based on color.
[0013] Figure 2B An embodiment of software is shown that enables the selection of colorants to generate a prediction of the color weight of the colorant.
[0014] Figure 3 The process for using an artificial intelligence (AI) color prediction engine to predict colorants for coloring is shown.
[0015] Figure 4 It shows the use of in Figure 3 An example of the AI color prediction engine used in the process.
[0016] Figure 5 A method for digitizing information about the coloring process and training an AI coloring prediction engine is shown.
[0017] Figure 6A An isometric view of the color measurement unit is shown.
[0018] Figure 6B The interior of the color measuring unit is shown during liquid batch measurements.
[0019] In all the accompanying drawings, the corresponding reference symbols denote the corresponding parts. Detailed Implementation
[0020] The features and other details of the concepts, systems, and techniques sought to be protected herein will now be described in more detail. It will be understood that any specific embodiments described herein are shown by way of illustration and not as a limitation on the disclosure and concepts described herein. Features of the subject matter described herein may be employed in various embodiments without departing from the scope of the sought-protected concepts.
[0021] Referring to the following description and figures, a system 100 for coloring liquids based on artificial intelligence (AI) such as machine learning (ML) is disclosed. Figure 1 This is a block diagram illustrating a system 100 performing example processes embodying aspects of the present disclosure. In one embodiment, the industrial mixing system 102 contains a liquid product to be colored, such as paint. In other embodiments, other liquid products may be colored. In some embodiments, the liquid product begins as colored or tinted. Liquid batch samples are collected from the industrial mixing system 102 and conveyed to a color measurement unit 104. The color measurement unit 104, further described below, measures the liquid batch samples to determine color measurements of the batch samples. As mentioned above, conventional coloring processes require the use of liquid batch samples to create batch test panels for comparison with a target reference panel, rather than for evaluating the liquid product itself. Aspects of the present disclosure improve production efficiency because they overcome the need to repeatedly manufacture test panels, thus eliminating the time required to create and cure multiple test panels. Furthermore, the liquid color measurement system 100 reduces batch production costs because the system does not require the creation of physical batch test panels or product sample panels.
[0022] Color digitization processor 106 receives liquid batch measurement results and coordinates the coloring process. In some embodiments, color digitization processor 106 is coupled to a display for showing graphical information related to the coloring process and an input device for configuring or updating the coloring process. In embodiments, an operator of the coloring process can perform actions such as monitoring the coloring process, modifying aspects of the process, checking the current coloring process, or checking historical coloring processes via interactive software. The interactive software can be standalone or a component of coloring process management software. Color digitization processor 106 is connected to both color digitization database 108 and AI coloring prediction engine 110. In some embodiments, AI coloring prediction engine 110 is executed by color digitization processor 106.
[0023] The color digitization database 108 stores information about the current coloring process, historical coloring processes, stock keeping units (SKUs) of the target standard liquid sample, and information about the colorants added to color the liquid. In some embodiments, the SKU information includes the SKU identifier, the SKU end application, the color measurement value of the SKU, the measurement date, and the acceptable incremental error. Historical batch coloring information may include information about the number of coloring operations, the required colorant, the colorant quality, the colorant mixing time, the batch size, the batch date, whether the batch was accepted, and / or the color measurement value for each product identifier, as well as information related to mapping colors to product identifiers. The colorant information stored in the color digitization database 108 includes one or more colorant identifiers and intensity measurements.
[0024] In this embodiment, the AI color prediction engine 110 includes computer-executable instructions executed by the color digitization processor 106. As will be described in further detail below, the AI color prediction engine 110 receives color measurements, compares them to corresponding values for a target color, and predicts which colorant to add and its quality to achieve the desired result matching the target color. The target color may be determined by a selected input color or automatically generated within the AI color prediction engine 110 based on color values from an earlier received batch of products.
[0025] Color digitizer 106 transmits prediction information to insight processor 112. In some embodiments, insight processor 112 operates as a component within color digitizer 106. In other embodiments, insight processor 112 operates independently via a network connection. In some embodiments, insight processor 112 is connected to color digitizer 106 via a wireless network such as Wi-Fi or Bluetooth. Insight processor 112 executes instructions for transmitting prediction information to an external display.
[0026] The Insight Processor 112 allows plant managers or others supervising the coloring process to access details about the coloring process, including expected processing time and the predicted and / or actual number of iterations required to complete the coloring process. The Insight Processor 112 provides real-time updates on the coloring process to plant managers or any other authorized plant personnel. Advantageously, up-to-date coloring process information enables operators to accurately plan based on coloring process execution time, thereby improving plant efficiency. Furthermore, the Insight Processor 112 combines interactive software to analyze historical coloring information for display to operators. The Insight Processor 112 evaluates historical coloring information to assess products and provide feedback on product types requiring more coloring iterations. For example, Figure 2AThe illustration shows the coloring iterations required by product clustering according to an embodiment. The insight processor 112 and interactive software provide insights into the coloring process by clustering information from the historical coloring process to inform future coloring process planning.
[0027] In some embodiments, the interactive software enables the operator to generate predictions about the outcome of the coloring process. For example, by Figure 2B As shown, in one embodiment, the operator inputs the selected colorant and target standard sample color to generate a prediction. The AI coloring prediction engine 110 then predicts the quality of the colorant to be added to achieve the desired target standard sample. Next, the operator performs coloring via the industrial mixing system 102 based on the prediction, or the operator can override the prediction to change the colorant quality. If the operator overrides the prediction, the changed value is transmitted to the color digitization database 108 to be incorporated into the retrained AI coloring prediction engine 110. In another embodiment, the user inputs colorant data into interactive software. Then, after selecting the colorant to be used and its quality, the AI coloring prediction engine 110 generates a prediction of the incremental effect on the liquid batch. The operator can adjust the added colorant and its quality to find the desired outcome of the coloring process.
[0028] Alongside the transmission to the insight processor 112, the color digitization processor 106 also transmits the predicted colorant and its quality to the industrial mixing system 102 to complete the coloring process. The industrial mixing system 102 then adds the required colorant to the mixer to update the liquid batch. By collecting subsequent measurements, the resulting liquid batch can be processed multiple times to achieve the desired color.
[0029] Now refer to Figure 3 An example coloring process embodying various aspects of this disclosure is illustrated. The process begins with the digitization of coloring information at 302. Digitizing batch information includes storing SKU liquid sample measurements, colorant liquid sample measurements, and previous batch sample measurements in a color digitization database 108. The SKU liquid sample measurements correspond to potential target values for the coloring process. The process continues by receiving measurements of the batch sample color at 304. In some embodiments, such as... Figure 6A As shown, the color of a batch of samples is determined by a color measurement unit. In some embodiments, the color measurements received by the color digitization processor 106 include measurements in any of a variety of color spaces, such as CIE XYZ, CIE L*a*b*, CIE L*C*h*, or CIELUV. After receiving the color measurements, at 306, the operator enters the stock holding unit (SKU) or product identifier for the batch color.
[0030] The coloring process continues at 308 by comparing batch sample color measurements with a standard reference measurement of the target color. In some embodiments, comparing batch sample color measurements with the standard reference measurement includes calculating the difference between each coordinate in the color space. For example, calculating the difference between each of the L*, a*, and b* coordinates in the CIELAB color space. In some embodiments, after calculating the difference between each coordinate, the color digitizer 106 or the AI coloring prediction engine 110 determines one or more incremental errors at step 310. In one embodiment, the incremental error in the CIELAB color space is determined by calculating the root mean square of all coordinate differences. However, other methods for determining the incremental error can be applied. The AI prediction engine 110 then evaluates the status based on the incremental error at step 312. In one embodiment, the AI prediction engine 110 predicts an acceptable tolerance for the incremental error of the target standard sample value based on historical batch information, including whether a given batch of the same SKU was accepted or rejected. If the incremental error is within the acceptable tolerance of the target standard sample value, the coloring process is complete. The color digitizer 106 then transmits this status to the insight processor 112. In some embodiments, the color measurement unit captures another color measurement value of the obtained batch to update the color digitization database using further information.
[0031] Table I below shows batch information and an example of calculating incremental errors within the CIELAB color space:
[0032] Table I
[0033]
[0034] If the incremental error is outside the acceptable tolerance, the coloring process continues with further coloring. At step 316, the AI coloring prediction engine predicts the amount of colorant and the mass of colorant required to obtain the target standard sample measurements by modeling the coloring process. This process... Figure 4As shown in the diagram, and further described below. After predicting the colorant to be added to the liquid batch, at step 318, the color digitizer 106 transmits the colorant and the desired mass to the industrial mixing system 102. In some embodiments, the operator may override the predicted mass of the colorant. For example, the operator selects a mass at 70% of the predicted mass for testing after the next iteration of mixing. The mixing system then continues to color the liquid batch with the colorant having the predicted mass or the mass input by the operator. After coloring the liquid, at step 320, the color digitizer 106 transmits the predicted colorant along with information about the expected coloring operations and / or time required to complete the coloring process to the insight processor 112. Then, after mixing is completed by the industrial mixing system 102, coloring can be restarted at step 304 by collecting a second batch of samples from the resulting batch.
[0035] Now for reference Figure 4 The AI color prediction engine includes one or more models for a coloring process that generate one or more color predictions. The AI color prediction engine 110 begins by receiving color measurement information 402. In one embodiment, the color measurement information 402 includes color measurements such as CIELAB coordinates, calculation of incremental errors, batch volume, and target color information including target color measurements. The one or more models then weight the color information to predict which colorant and the required mass of each colorant to add to the liquid batch to achieve a match with the target color after application and drying. Figure 4 An embodiment is shown in which the AI prediction engine 110 feeds color information into multiple models. The prediction generation models include any of a variety of machine learning models. This embodiment illustrates the use of one or more of a first-principles nonlinear model 404, a first-principles linear model 406, or a smart shading model 408. The smart shading model 408 is a machine learning model that has been trained based on historical shading information and colorant information, as described below, see [link to relevant documentation]. Figure 5 .
[0036] Each model generates a prediction of the weights used for coloring. Each model generates a prediction when it receives color measurement information 402. Each model can be trained separately, with different weights assigned to each attribute considered based on the model. In one embodiment, the models are weighted based on an intensity factor. Color measurements in digital space may not accurately represent color representations in physical space. Thus, one or more prediction models can be weighted to update colors after the initial predictions, thereby better reflecting colors in physical space. The predictions from each model are then fed into a dynamic model selector 410.
[0037] The dynamic model selector 410 generates a final weighted prediction 412 for the coloring operation. In some embodiments, the dynamic model selector 410 selects a single prediction for the input model based on historical coloring information and the target color. In some embodiments, the dynamic model selector 410 generates composite weights based on the input model predictions. In other embodiments, the dynamic model selector 410 acts as a second layer of AI prediction by generating weights for each model prediction to generate single weights through a machine learning model.
[0038] After generating the final weighted prediction 412, the model 414 is retrained using batch results. In some embodiments, the industrial mixing system 102 uses the final weighted prediction 412 to add colorant to the liquid batch. Then, at 402, color measurement information generated as a result is fed back into the AI color prediction engine for retraining. When testing the model, the color measurement information 302 is derived from the results of historical coloring data, rather than from new samples measured after the coloring process. Each model undergoes reweighting of its various inputs, including the dynamic model selector 412 in embodiments where the dynamic model selector is the machine learning model itself. Aspects of this disclosure improve the consistency of the coloring process. Conventional coloring models require experts in the coloring process to evaluate batch sample plates and standard reference plates to predict one or more colorants required for the liquid batch. As a result, the process can vary based on expert knowledge and subjective observation. However, the process disclosed herein improves upon this by ensuring consistent predictions for a given target color through the digitization of coloring expert knowledge and consistent, standardized modeling via the AI color prediction engine 110.
[0039] Now for reference Figure 5 The AI color prediction engine 110 includes a color model trained based on historical coloring information. First, at step 502, the product is digitized within the color digitization database 108. Product digitization includes storing information such as a product identifier or SKU, color measurements for liquid products, color measurements for dried products, and previous coloring information. The previous coloring information also includes color measurements for liquid products at each step of the coloring process, the colorant added at each coloring step, the mass of the added colorant, and any available dried color measurements at each step.
[0040] Next, at step 504, the colorant information is digitized for use within the AI color prediction engine 110. Digitizing the colorant information includes storing information such as the colorant identifier, any manufacturing information (if available), and the colorant strength that can be measured, predicted, or provided by the manufacturer. Colorant strength represents the effectiveness of a colorant in coloring liquid products. A colorant with high colorant strength requires a smaller mass of colorant to be added to the liquid batch to achieve the desired result, while a colorant with low colorant strength requires a larger mass of colorant to achieve the desired result. The colorant information also includes historical information about the use of the colorant, such as which products include the use of the colorant and the quality of the colorant used. In some embodiments, the AI color prediction engine 110 predicts colorant strength by weighting previous historical coloring process information based on the quality of the colorant used.
[0041] Model training begins at step 506 by ingesting historical coloring information. A curated set of historical coloring information is synthesized to train the model. Information is selected based on various criteria, such as the coloring result, the reliability of the information, and the sufficiency of detailed process information for each step of the coloring process. Similarly, colorant information is also curated to optimize model training. As part of curating the training information set, a curated test information set is also generated. The model is then trained based on the curated historical coloring information and the curated colorant information.
[0042] After initial training, the model is tested at step 508. Testing the model involves feeding the model with compiled historical coloring process information from the test set to determine predictions. The predictions are then compared to the historical process results to determine the accuracy of the predictions. The weights of the colorant information or product information are then updated based on the accuracy of the predictions. Finally, the model is further retrained at step 510 when the coloring process generates new coloring information to be fed back into the model. As a result, the model continuously learns from the generated predicted colorants and the resulting color measurements.
[0043] Now for reference Figure 6AIn some embodiments, the system for coloring liquids implements a color measurement unit 104. The color measurement unit 104 provides accurate color measurements of liquids, including highly viscous liquids. In the illustrated embodiment, the color measurement unit 104 includes a container 602 within an inner region 604 in which the liquid is measured. The container 602 can be any opaque material to prevent external light from interfering with the measurement. In some embodiments, the color measurement unit 104 also includes a hinged door sealed to the outside of the container 602 to prevent all light from the inner region 604. The color measurement unit 104 also includes an inner region 604 within the container 602, which is designed for testing batch samples of the liquid. Within the inner region 604, a test holder 606 is mounted inside the container 602. In this embodiment, the test holder 606 includes a transparent test surface 610 formed therein. In some embodiments, the test holder 606 includes only the transparent test surface 610. The transparent test surface 610 can be made of a light-transmitting material, such as glass or plastic, thereby enabling testing of batch samples of the liquid from below the test holder 606.
[0044] Now for reference Figure 6B Below the test fixture 606 coupled to the base of the container is a mounting bracket 612 for attaching a sensor 614. In some embodiments, the mounting bracket 612 is adjustable, allowing the sensor 614 to be moved closer to or further away from the transparent test surface 610 for more accurate testing. In some embodiments, the sensor 614 is a spectrophotometer. The spectrophotometer may be coupled to a controller or another device for transmitting data captured by the spectrophotometer to the color digitization processor 106. An example of a suitable spectrophotometer is the Micro-Epsilon CFS2-M20-E-2400. Similarly, a suitable controller for transmitting sensor data is the Micro-Epsilon ColorSENSOR CFO200.
[0045] In some respects, the position of sensor 614 enables rapid and accurate testing of liquid samples. A batch of samples within test container 618 is placed on a transparent inner test surface 610. Sensor 618 then captures color measurements of the batch sample, which can be processed by a separate controller or color digitization processor 106. As previously mentioned, conventional methods of color detection involve orienting the spectrophotometer above the test sample. With conventional arrangements, the meniscus of the liquid sample and / or air bubbles within highly viscous liquids are causes of color measurement errors. In this embodiment, the sample can be rapidly placed within the inner test surface 610 and measured using sensor 614 without significant reconfiguration to adjust the liquid's meniscus or delay to allow air bubbles to escape from the liquid. Furthermore, color measurement unit 104 can utilize… Figure 6A and Figure 6B The test configuration captures color measurements from liquid samples, thus avoiding the drying time associated with conventional color measurements.
[0046] In some embodiments, the color measurement unit 104 is coupled to the color digitization processor 106. This direct connection between the color measurement unit 104 and the color digitization processor 106 enables direct management of the sensor from the color digitization processor 106. Furthermore, the direct link ensures secure and rapid storage of liquid batch measurements without reliance on an external network. In some embodiments, the color digitization processor 106 also includes an insight processor 112 and an AI color prediction engine 110 as sub-components of the coloring process software. In such embodiments, the color digitization processor 106 operates the entire coloring and prediction process and displays all information about the current and historical coloring processes. In some embodiments, the color measurement unit 104 and the color digitization processor 106 are coupled via a wireless network, such as Wi-Fi or Bluetooth.
[0047] Embodiments of this disclosure may include a dedicated computer, which includes various computer hardware as described in more detail herein.
[0048] For illustrative purposes, programs and other executable program components may be shown as discrete blocks. However, it should be recognized that such programs and components reside in different storage components of the computing device at different times and are executed by the device's data processor.
[0049] Although described in conjunction with an example computing system environment, embodiments of the various aspects of the invention may operate with other dedicated computing system environments or configurations. The computing system environment is not intended to impose any limitation on the scope or functionality of any aspect of the invention. Furthermore, the computing system environment should not be construed as having any dependency or requirement associated with any component or combination of components shown in the example operating environment. Examples of computing systems, environments, and / or configurations that may be suitable for use with the aspects of the invention include, but are not limited to, personal computers, server computers, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, mobile phones, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the foregoing systems or devices, etc.
[0050] Embodiments of various aspects of this disclosure can be described in the general context of data and / or processor-executable instructions, such as program modules stored in one or more tangible, non-transient storage media and executed by one or more processors or other devices. Typically, program modules include, but are not limited to, routines, programs, objects, components, and data structures that perform a particular task or implement a particular abstract data type. Various aspects of this disclosure can also be practiced in distributed computing environments where tasks are performed by remote processing devices linked via a communication network. In a distributed computing environment, program modules can reside in both local and remote storage media, including memory storage devices.
[0051] In operation, processors, computers, and / or servers can execute processor-executable instructions (e.g., software, firmware, and / or hardware) such as those shown herein to implement aspects of the present invention.
[0052] Embodiments may be implemented using processor-executable instructions. These processor-executable instructions may be organized into one or more processor-executable components or modules on a tangible processor-readable storage medium. Furthermore, embodiments may be implemented using any number and organization of such components or modules. For example, aspects of this disclosure are not limited to the specific processor-executable instructions or specific components or modules shown in the drawings and described herein. Other embodiments may include different processor-executable instructions or components having more or fewer functions than those shown and described herein.
[0053] Unless otherwise specified, the order in which operations of the aspects of this disclosure shown and described herein are performed or carried out is not required. That is, unless otherwise specified, operations may be performed in any order, and embodiments may include more or fewer operations than those disclosed herein. For example, it is conceivable that performing or carrying out a particular operation before, simultaneously with, or after another operation is within the scope of this invention.
[0054] When describing elements of the invention or embodiments thereof, the articles “a,” “an,” “the,” and “the” are intended to indicate the presence of one or more elements. The terms “comprising,” “including,” and “having” are intended to be inclusive and indicate that additional elements may be present in addition to those listed.
[0055] Not all of the components shown or described may be required. Additionally, some implementations and embodiments may include additional components. Variations in the arrangement and type of components may be made without departing from the spirit or scope of the claims set forth herein. Furthermore, different or fewer components may be provided, and components may be combined. Alternatively or additionally, a component may be implemented by several components.
[0056] The above description illustrates embodiments by way of example and not limitation. This specification enables those skilled in the art to make and use aspects of the invention, and describes numerous embodiments, modifications, variations, alternatives, and uses of aspects of the invention, including modes currently considered best for carrying out aspects of the invention. Furthermore, it should be understood that aspects of the invention, in their application, are not limited to the details of the construction and arrangement of components set forth in the following description or shown in the accompanying drawings. Aspects of the invention can have other embodiments and can be practiced or performed in various ways. Moreover, it should be understood that the wording and terminology used herein are for descriptive purposes and should not be considered limiting.
[0057] It will be apparent that modifications and variations can be made without departing from the scope of the invention as defined in the appended claims. Since various changes can be made to the above constructions and methods without departing from the scope of the invention, it is intended that all content contained in the above description and shown in the accompanying drawings should be interpreted as illustrative rather than restrictive.
[0058] In view of the above, it can be seen that several advantages of various aspects of the present invention have been achieved, and other advantageous results have been obtained.
[0059] An abstract and a summary of the invention are provided to help the reader quickly determine the nature of the technical disclosure. They are submitted with the understanding that they will not be used to interpret or limit the scope or meaning of the claims. The summary of the invention is provided to introduce, in a simplified form, a series of concepts that will be further described in the detailed description. The summary of the invention is not intended to identify key or essential features of the claimed subject matter, nor is it intended to help determine the claimed subject matter.
Claims
1. A method for coloring a liquid in an industrial mixing system, the method comprising: Collecting liquid batch samples from liquid batches in industrial mixing systems; The liquid batch sample is measured using a sensor to generate batch color measurement values associated with the liquid batch sample; The batch color measurement values are transmitted to the color digitization processor; The color digitization processor executes an artificial intelligence (AI) color prediction engine, wherein executing the AI color prediction engine includes: The batch color measurement values are stored in a color digitization database, which includes color information of multiple historical batch samples, color measurement values of multiple target standard samples, and colorant information, wherein the colorant information includes colorant values; Define the incremental error between the batch color measurement value and the target standard sample color measurement value; and By weighting one or more of the incremental error, the batch color measurement value, the batch volume, the color information of the multiple historical batch samples, and the information of the multiple colorants, the coloring of the liquid batch is modeled, so as to generate one or more predicted colorants and their masses based on the information of the multiple colorants, so as to achieve the color measurement of the target standard sample. The predicted colorant and its mass are received by the industrial mixing system; and The industrial mixing system mixes one or more colorants into the liquid batch based on the predicted colorant and its quality.
2. The method according to claim 1, wherein, The AI color prediction engine also includes: Generate prediction information for the completion of the coloring process of the liquid batch, including the expected number of coloring operations and the expected coloring time; The prediction information for colorization completion is sent to an insight processor, which is configured to transmit the graphic data of the colorization prediction to a display.
3. The method according to claim 1 or 2, wherein, The method further includes: After mixing the one or more colorants, the resulting liquid sample is collected from the industrial mixing system; The obtained liquid sample is analyzed by the sensor to generate a obtained color measurement value associated with the obtained liquid sample; The obtained color measurement values are transmitted to the color digitization processor; The obtained color measurement values are used to update the color digitization database; and Based on the batch color measurement values, the predicted colorants, the mass of each of the predicted colorants, the target standard sample color measurement values, and the obtained color measurement values, the color information of the multiple historical batch samples and the information of the multiple colorants are reweighted.
4. The method according to any one of claims 1 to 3, wherein, The sensor includes a spectrophotometer, and the batch color measurements include CIELAB coordinates.
5. The method according to any one of claims 1 to 4, wherein, The execution of the AI color prediction engine also includes: determining the target standard sample incremental error tolerance based on the color measurements of the plurality of historical batches of samples and the color measurements of the target standard sample before defining the incremental error.
6. The method according to any one of claims 1 to 5, wherein, The colorant information also includes colorant intensity, and wherein executing the AI color prediction engine further includes: After storing the batch sample colors, the colorant intensity of each of the multiple colorant information is generated based on the multiple historical batch sample color information and the multiple colorant information.
7. The method according to any one of claims 1 to 6, wherein, The method further includes: Training the model before executing the AI colorization prediction engine, wherein training the model includes: Multiple sets of standard color information are created for the artificial color prediction engine. The standard color information includes product identifiers, standard liquid colors, standard drying colors, and historical coloring data. Multiple sets of organized colorant information are created for the artificial coloring prediction engine. The organized colorant information includes colorant identifiers, historical coloring data, and colorant liquid colors. Determine the colorant strength for each of the compiled colorant information; The model is trained by generating multiple weights based on the colorant intensity of each of the sorted standard color information, the sorted colorant information, and the sorted colorant information. Generate one or more predicted test colorants based on the color of the test liquid; The weights are regenerated based on the predicted test colorant and one or more expected colorants.
8. A liquid coloring system, comprising: A mixing system used for mixing liquids; A color measurement unit is used to receive and measure liquid samples collected from the mixing system to generate liquid color information; A color digitization processor is coupled to the color measurement unit; A color digitization database is coupled to the color digitization processor, and the color digitization database stores product information and colorant information; as well as The memory stores computer-executable instructions that, when executed by the color digitizer, configure the color coloring system to: Receive the liquid color information from the color measurement unit; Product identifiers are identified based on the liquid color information and the product information; as well as Execute an artificial intelligence (AI) color prediction engine to predict colorants, wherein executing the AI color prediction engine includes: Incremental error is defined based on the liquid color information and the product identifier; The coloration of the liquid sample is modeled using one or more models by predicting one or more colorants and their masses based on the incremental error, the liquid color information, the product identifier, and the colorant information; and The colorant and colorant mass are transferred to the mixing system.
9. The liquid coloring system according to claim 8, wherein, The prediction of the colorant and colorant quality is also based on multiple colorant intensities, and wherein the color prediction processor is further configured to: Following the identification step, the plurality of colorant intensities are generated to obtain the colorant information.
10. The liquid coloring system according to claim 8 or 9, wherein, The system also includes: An insight processor, coupled via a network to the color digitization processor, is configured to transmit graphic information based on the product information and colorant information to a display.
11. The liquid coloring system according to any one of claims 8 to 10, wherein, The instructions in memory also include predictions of one or more colorants and colorant quality: Estimate the number of iterations required for the coloring process and the expected coloring time; The number of iterations, the expected coloring time, and the colorant and colorant quality are transmitted to the insight processor.
12. The liquid coloring system according to any one of claims 8 to 11, wherein, The AI color prediction engine also includes: Receives input representing the colorant; and Before defining the incremental error, the obtained color is predicted based on the input representing the colorant, the liquid color information, and the colorant information.
13. The liquid coloring system according to any one of claims 8 to 12, wherein, The one or more models include first-principles nonlinear models and first-principles linear models.
14. A system for measuring the color of a liquid, comprising: A test container, the test container including a base, an internal mounting component and an internal compartment, wherein the internal mounting component is coupled to the base; A test fixture, wherein the test fixture is mounted within the internal compartment, and the test fixture includes a transparent inner surface configured to receive a test sample; and A spectrophotometer is coupled to the internal mounting below the transparent inner surface of the test fixture, wherein the spectrophotometer is oriented such that the sensor end of the spectrophotometer points toward the transparent inner surface.
15. The system for measuring the color of a liquid according to claim 14, further comprising: The color digitization processor is connected to the spectrophotometer via a network.
16. The system for measuring the color of a liquid according to claim 14 or 15, wherein, The internal mounting includes an adjustable mounting that allows the sensor to move closer to and away from the transparent inner surface.